Last week’s Loh-Down was a deep dive into The Fed’s Operation Twist programs and the Treasury Department’s buyback operations.

This week, in honor of the long weekend, I plan to spare you the pain of having to read another long essay about the history and mechanics of specific Fed and Treasury policies.

That doesn’t mean you get to avoid hearing about the Fed or Treasury this week, though. No, it’s important we follow up on the topic a bit because the September Federal Open Market Committee (FOMC) meeting is right around the corner. I am still of the opinion that the Fed will try to avoid rate hikes completely if they can, or at least push them until after the election if they can’t, but we should still spend a few minutes discussing what could make them hike sooner than expected.

It’s Oil Prices, Stupid!

The famous slogan “It’s the economy, stupid” comes from Bill Clinton’s 1992 election campaign manager, James Carville. It’s a well-known phrase and repeatedly modified within politics and journalism to highlight our penchant for overcomplicating things. I’d be willing to bet Carville’s father, mother, or some other mentor in his life used the K.I.S.S. acronym regularly. No, not that KISS, I’m talking the ‘Keep it simple, stupid’ variety.

When you watch the Fed long enough, you learn that it’s not about what they should do, but about what they can and will do. Macroeconomists and the ‘online experts’ can tell you dozens of reasons the Fed should or shouldn’t hike rates in September, and they’ll do it with a bunch of smart-sounding data and charts. Anything I say today is not an argument for or against the soundness of their logic, but more of a critique on the unnecessary complications we make of things when we focus on should or shouldn’t rather than can/will and can’t/won’t.

Is oil the primary driver of inflation today? No, of course it isn’t, but it is the most salient issue for consumer pocketbooks. That and grocery prices, which have also been affected by fertilizer disruptions stemming from the War with Iran.

The Fed has plenty of justifications for looking through “temporary” inflation shocks, and it will use those arguments until the shock begins bleeding into broader prices, inflation expectations, or consumer behavior.

Granted, raising interest rates will not decrease oil prices. It will not bring down food prices. It will do little to fix today’s inflationary woes. If increased high enough and held there long enough, it will reduce aggregate demand, reduce bank lending, appease bond vigilantes, and keep U.S. yields competitive with other sovereign markets where our debt competes for capital. (If that last part doesn’t make complete sense to you, please refer to last week’s piece that linked to above.)

I am The Waller(us)

If you don’t believe that to be true, maybe you didn’t hear Chris Waller this week? He said he is still open to holding rates steady as long as August CPI inflation doesn’t come in hot. What is the most obvious candidate to make CPI come in hot? Ding ding ding, that would be oil prices. West Texas Intermediate oil hovers around $90 a barrel while gasoline and diesel prices were up again in August as fighting between the U.S. and Iran flared up yet again. Do the increases in oil, gas, and diesel flow through to core goods and services inflation in August, or does it stay limited to headline CPI? I do not have a crystal ball to tell you that, but if oil and gas prices do push inflation metrics up then you could see the Fed hike rates in September.

That could be especially true after today’s employment report. Nonfarm payrolls increased by 162K in August, with July’s -23K print being revised to +21K. That came with an increase to the labor force participation rate from 61.4% to 61.6%. That measure has oscillated between drifting lower and trending sideways for some time now. If there is no conclusion, or new off-ramp at least, to the War in Iran before the FOMC meeting, then the Fed can look Hawkish with a rate hike without having to worry as much about damaing the labor market.

Again, it is crucial to highlight that this would not fix our immediate inflationary issue.

Liberation Day 2: No Trade For You

Trump effectively threatened a Liberation Day redux this morning, stating that if the Fed didn’t give him his rate cuts he would stop all trade with any countries with which the U.S. runs a trade deficit. We have also seen the push for ejecting Lisa Cook picking up steam again, so nothing economic is cut and dried these days.

We are a country that has been running deficits too hot for about a decade now, that is no longer upholding the rules we previously set on the global stage, and is waging trade and kinetic wars with allies and enemies alike.

That backdrop makes forecasting more challenging than it already is. We also live in a world of financial repression at the moment, and that means many of the forecasting relationships we relied on before become less reliable under a regime of financial repression.

Right now we are experiencing a moment where human psychology and institutional incentives may matter more than what the data tells you.

• Do constant threats from the President cause some folks to make a decision of defiance?

• Does pain at the pump cause enough righteous anger that the Fed feels they have to be seen doing something…anything?

• Does financial repression win the day, and a Dovish tailwind follows?

I have no idea, but you can bet there will be a lot of ink spilled on this topic now and into the future.

Interesting Reads & Listens:

• Worse On Purpose Blog - On Quality

• Worse On Purpose Blog - Your Cookware Got Worse On Purpose

• Mauldin Economics: Thoughts from the Frontline - Inflation Conundrums

• The Knowledge Project Podcast - No. 1 Performance Psychologist: The Secret to High Performance and Excellence

• Plain English Podcast - A Philosopher’s One-Word Theory For Why the World Feels So Weird

Disclaimer: The author is not currently registered as an Investment Adviser or Investment Adviser Representative. The information contained in this publication reflects the author’s personal opinions and is provided solely for educational and informational purposes. Nothing contained herein should be construed as personalized investment, legal, tax, or financial advice, or as a recommendation to buy, sell, or hold any security.

Information obtained from third-party sources is believed to be reliable; however, the author makes no representation or warranty regarding its accuracy, completeness, or timeliness. Opinions expressed are current only as of the date of publication and are subject to change without notice.

Investing involves risk, including the possible loss of principal. The author makes no representation regarding the suitability of any investment for any particular person. Readers should conduct their own due diligence and consult an appropriately registered financial professional before making investment decisions.

To the fullest extent permitted by law, the author and The Loh-Down disclaim liability for investment decisions made based on information contained in this publication.

In 1961, the Federal Reserve under the Kennedy Administration launched Operation Twist. The program was designed for the Fed to buy long-duration Treasury bonds from the market, while selling off the short-term securities held on their balance sheet. The goal was to lower long-term interest rates, in order to stimulate economic investment, while holding or slightly increasing short-term interest rates to prevent gold and capital outflows from abroad.

Most studies after the fact deemed the twisting and shouting operation to be unsuccessful, having only ‘modest impact’ on the desired outcome. While those critics may be right about its success in terms of desired outcome, I’ll take the contrarian side and say it was successful in unintended ways… For instance, The Beatles released the wildly popular song ‘Twist and Shout’ in 1963, only two years after the operation began.

I can’t say for certain that the Beatles were inspired by the Federal Reserve, but I like to think they were trying to help the Fed Twist and Shout yields down in their own special way.

It was also successful in the fact that the experiment has spawned repeat attempts since then, like when the Federal Reserve revived Operation Twist in 2011 following the massive Quantitative Easing (QE) efforts after the 2008 Global Financial Crisis (GFC). This iteration was considered much more effective in lowering long-term interest rates, but it’s hard saying if it was Operation Twist specifically or simply the catastrophic meltdown of the global financial system and the trillion plus dollar QE program hoovering up Treasuries and Mortgage-Backed Securities (MBS) generally.

Whether that perceived success of the 2011 operation is accurate or not, it seems to have spurred the Treasury Department into trying their luck on a Quasi-Operation Twist. First with Janet Yellen’s Treasury Department in 2023/2024, while folks like Scott Bessent criticized the moves as a “Stealth Operation Twist.” This was of course followed up with Bessent doing the exact same thing as he took over as the new Treasury Secretary.

Which brings us to today as the impetus for this week’s Loh-Down, Bessent announcing Treasury buyback amounts for the 10-20 year and 20-30 year buckets would be doubled starting September 9, 2026 to $4 billion per auction. Similar to Yellen’s Treasury, the stated rationale was for “liquidity support,” but the markets see the move more as an attempt to contain the recent rise in long-duration Treasury yields. Yields went down at the time of the announcement, which must have pleased Bessent, but have since resumed their climb higher. That defiance by the bond market has also led to some further shouting from the Treasury Department that they might use their $950 billion Treasury General Account (TGA) to buy even more long-duration Treasuries if they feel so inclined. This is the bond vigilante story I was looking for in 2025, what a time to be alive!

This week, I will walk you through some of the mechanics of the initial Operation Twist, the Operation Twist redux, and why Janet Yellen and Scott Bessent’s latest buybacks really aren’t Operation Twist, but they sure rhyme with the Fed experiment.

What was Operation Twist, and why did we settle on that operation?

There was a lot of timeline and incomplete information above that might have left you scratching your head with some follow-up questions. That or you read the above and realized I was about to drone on for way too long again this week, and I don’t blame you for thinking that. Because that is exactly my plan.

For those of you that don’t have the patience to read something this long, I’ve used NotebookLM to make an audio version of this week’s piece.

Now away we go…By the time Operation Twist began in 1961, the U.S. was far along its reserve currency path established after World War II, and was facing the twin crises of a recession and a balance-of-payments deficit. America became very rich following the war by producing many of the goods for the world, and also leveraging U.S. fiscal resources to fund a rebuild of the European countries devastated by the war. By 1961, many European countries had shed the malaise and were growing strong. For the sake of time, I’ll commit a mental shortcut here and say part of the balance-of-payments problem was the downstream consequence of our own postwar successes with the Marshall Plan and private investment that followed. The timeline is much more complex and nuanced, but it is hard to argue at a broad level that our success in rebuilding Europe and Japan after World War II didn’t eventually play a big part in creating other attractive places for capital to go besides the United States.

Problematically, the world was operating on this USD as a reserve currency, but the U.S. was still technically on the gold standard following the Bretton Woods Accord. What does that mean exactly? Under Bretton Woods, foreign central banks could hold dollars as reserves because the U.S. committed to converting official dollar holdings into gold at $35 per ounce. This worked as long as foreign governments remained comfortable holding dollars. Persistent balance-of-payments deficits complicate this relationship, though. More dollars accumulate abroad when that happens, while America’s gold stock is finite and was not growing at the same pace. If too many of those dollar holders get spooked and decided they wanted gold instead of USD, then reserves could drain suddenly.

So, by 1961 the U.S. was staring down a recession, and concerns that other central banks might convert their USD holdings into gold. If that were to happen, the U.S. might have been forced to suddenly devalue their currency.

The Federal Reserve saw this issue and came up with the plan to buy long-duration Treasuries off the market, while simultaneously offloading the short-duration Treasury Bills (T-Bills) from their balance sheet. The goal was to lessen supply of long-duration bonds while increasing the supply of short-duration bonds without any change in total debt issuance. Bonds come with fixed maturities and/or fixed coupon payments, which gives you an investment yield. If a bond has a 5% coupon payment and you buy it at 100, your yield-to-maturity is 5%, because it will mature at $100. Your yield is simply the coupon rate in that example. If the bond has a 5% coupon rate, but you buy it in the open market for $99, your yield will be above the 5% coupon rate because one day it will mature at $100 when you paid $99. So you earned some price appreciation plus the 5% coupon rate over the life of the bond. If you paid more than $100 then your yield would be less than the 5% coupon rate. With bonds, price and yield have an inverse relationship.

So, when the Federal Reserve starts buying up the available supply of 10 or 30-year Treasuries in the market, it creates demand for that security. This pushes prices up, and therefore yields go down. If they are reducing supply of long-duration while increasing supply of short-duration then theoretically the expectation would be for yields to rise on the T-Bills while yields go down on long-duration Treasuries. This move was balance sheet neutral in that the Fed did not expand the balance sheet to hold more securities with this operation. All long-duration purchases were offset with ejecting shorter-term maturities from the balance sheet.

The goal was to lower long-term rates and incentivize corporate and mortgage lending, while bumping short-term rates up to make them more competitive with European countries. Why did that matter to the Fed? They felt that if our short-term rates were more competitive, then less creditor money would flee, and the U.S. wouldn’t experience a convertibility crisis.

The St. Louis Fed piece from the ‘90s I linked to at the beginning of this write-up calls Operation Twist mostly a failure and so did other papers before that. Those judgements may be a bit harsh as the Federal Reserve Bank of San Francisco noted in their 2011 paper. They found that the long-end of the yield curve did come down ~ 15 basis points (0.15%), which is fairly marginal in terms of inducing investment, but still technically had the intended impact. To me, the fact that there was no convertibility crisis during the period might be enough to say the operation was successful really. Muddling through some problems can be a lot less painful than an outright crisis.

2011 Twist Redux

Following the 2008 GFC, the Fed cut interest rates to nearly 0%, and began their QE program. QE is the process of the Fed buying up US Treasury securities and mortgage-backed securities and holding them on their balance sheet. This effectively reduces the supply of available debt for private or sovereign investors. From our bond lesson earlier we know a steady dose of demand will drive bond prices higher and yields lower, and that is exactly what the Fed wanted to do in order to revive the economy. The early QE programs were about expanding the Fed’s balance sheet and compressing yields broadly rather than targeting any specific maturity bucket. They wanted to incentivize risk-taking by investors again.

By 2011, the Fed Funds rate was pinned near 0%, but the long end of the yield curve remained higher than the Fed wanted.

Enter: Operation Twist 2 - Buy, Buy, Buy.

10-year yields peaked at 3.58% around February of that year, and then they began their descent lower. Operation Twist didn’t begin until September 2011, however, so the drop in rates can’t be attributed the Fed’s plans could it? Yes and no. The European Sovereign Debt Crisis was raging in 2010/2011, and as those fears scared bond investors there was a rush into less-risky debt like U.S. Treasuries. The “Flight-to-Quality” trade as it is called.

That had a meaningful effect on lowering long-term interest rates, but the Fed was ready to finish the job. You see, monetary policy doesn’t only work once a program starts, it starts once the market sees a clear signal from the Fed, and this was when the Fed really shifted its ‘forward guidance’ strategy into a higher gear. During the summer of 2011, rumors began flying that the Fed would soon implement Operation Twist, while becoming more explicit in their intention to keep rates lower for longer. The Bat Signal was shining bright for investors and speculators to front-run policy. Something we didn’t cover in our earlier bond lesson was that the longer the time-to-maturity a bond has, the more sensitive its price is to interest rate changes. It was time to go long duration before the Fed could, so you could reap a massive windfall on rising bond prices.

What kind of windfall are we talking about?

From July 2011, when rumors began swirling about Operation Twist, you could buy a 10-year Treasury for ~3% and a 30-year Treasury for ~4.2%. The chart above illustrates the theoretical 1-year return in July 2012 for purchasing those 10-year or 30-year Treasury securities. By then, the Fed’s operation had been in full-swing for months, and yields bottomed around 1.53% and 2.46%, respectively. Maybe you’re saying, “big deal, $10k grows to $14k over 1-year. That’s great, but not life-changing.” I can understand that response, but it it is important to contextualize that return.

First of all, a bond investor is buying that 30-year Treasury looking for a reasonable return, on an asset generally considered free of credit risk, not a grand slam. You’re purchasing a security that you expect to provide a 4.2% annualized yield-to-maturity, and within 1-year you can instead sell that security at a 40% return. You didn’t take on any new risk beyond the duration exposure you already expected to own, but suddenly found yourself sitting on a massive gain.

Second, the folks that were buying long-duration bonds on Operation Twist rumors were mostly institutional buyers that buy in the millions and billions, with some also applying massive leverage in the process. So the absolute return values are amplified when contextualized to the proper size.

Key Lesson from v1 to v2: Twist Harder, Shout Louder.

By most measures, the 2011 Operation Twist was materially more successful than the initial iteration. The decrease in yields was significant, bond investors that bought the rumor made out like bandits, the real estate market saw its first pockets of sunshine in years, and the stock market began its journey on what became a massive multi-year bull market. I’d willingly speculate that the Fed learned a thing or two from Operation Twist v1 and applied those lessons to v2: the program needed to be larger, and the communication around it needed to be louder and clearer.

We can see as a share of the public Treasury market and as a share of U.S. GDP, the operations were much larger. Fed communication became very vocal in this time period as well. There were the rumors before the program started, and constant, clear communication that they were going to massively buy duration at size. The market didn’t want to take the other side of that bet and get blown up trying to fight the Fed.

While this clearly provided clear results that demonstrated the Fed’s ability to bend yields to their will in a pinch, there are many perspectives folks could take to consider the operation unsuccessful still. GDP growth didn’t exactly ignite following bargain basement yields, the labor market improved at a steady snails-pace only, and in some pockets of the country conditions continued deteriorating.

Success, like beauty, is in the eye of the beholder I guess?

Janet and Scott—From Yellen to Bess Friends

Since we’re talking perspective, perhaps we expand that lesson a bit?

Like how outsiders tend to criticize insiders only to find themselves taking the same actions when they become the insider. That’s where Scott Bessent finds himself today after loudly criticizing Janet Yellen’s actions as Treasury Secretary to shorten the maturity composition of U.S. debt issuance and restart the Treasury buyback program for the first time since the early 2000s. Aren’t human quirks fun?

Let’s begin with the basics, while the buyback program hadn’t been used by the Treasury in years, Yellen was not the first to enact it. The tool itself has been around for some time to help improve market functioning. When the Treasury department auctions off debt securities, larger dealers will purchase debt to hold on their balance sheets and eventually sell to their clients. They are a middleman so the Treasury Department doesn’t need to waste its time finding all the individual buyers that might want some U.S. issued debt. Occasionally dealers might find their balance sheets a bit bloated as they couldn’t offload every dollar of a given maturity bucket. So when the next Treasury auction comes around, they can’t/don’t buy as much issuance. If the markets are functioning in a healthy manner that may not be a problem as some other dealer might want more at that time. When markets aren’t functioning as well, the buyback program is intended to improve dealer liquidity.

The Treasury offers to buy less-liquid, off-the-run securities in specific maturity buckets off their balance sheet, while auctioning debt at more liquid maturity terms. The dealer can offload those securities they were having trouble selling back to the Treasury Department and buy some newly issued securities that the market generally prefers. The buyback program by itself doesn’t eliminate the government’s financing needs, so any reduction in one maturity bucket ultimately is funded from cash or issuance elsewhere.

Where the criticisms of a “Stealth Operation Twist” came from was the one-two punch Yellen’s strategy consisted of. First, in late 2023 they began changing the composition of U.S. debt issuance to a heavy T-bill issuance schedule, while also reducing issuance of long-duration securities. Then the Treasury buyback program followed in May 2024 with a $2 billion per auction limit. The buyback program was technically open to purchases of all maturities, but the perception from critics was that they were focused on buying up the long-duration securities while issuing at the front-end. That they were being political and trying to put their thumb on the scale before the 2024 election.

I’ll say two things on that critique:

1. Given the size of the program enacted under Yellen it’s probably more appropriate to say she was putting a feather on the scale.

2. Everything done by the government is political, because the government is a political organization, but not everything the government does is partisan. We’d all sound a lot less dumb if we held space for more nuance in these conversations.

Anyway, the second Scott Bessent took over the Treasury, he turned from critic to advocate of the program. Today, the Treasury continues issuing heavily at the front-end of the yield curve, and as inflationary pressures and Japanese yen issues have pushed yields higher and higher, Bessent found himself in the same unenviable position. He needs to keep long-term yields from rising high enough to torpedo the economy.

That led to his August announcement that the long-end liquidity buyback cap of $2 billion would be expanded to at least $4 billion per auction. The initial market reaction was in line with what he hoped for, but when the sharks smell blood in the water, they swarm. Yields resumed their climb higher and Bessent shouted louder for those in the back to hear him. He wanted to make clear to the sharks circling and shorting the 10, 20, or 30-year Treasury space—$4 billion could just be his starting line. If he has to, he could use the full force of the TGA (~$950b) to buy back long-duration if push came to shove.

Nobody in the market really believes this right now because of exactly why the moves by Yellen and Bessent are not considered equivalent to Operation Twist. These purchase sizes are a drop in the bucket, and if the government wants to buy in size, they can only do so for a limited period of time. Yes, $950b is a lot of money, but you know what happens when they start drawing that down? They have to replenish it because they still need that money to fund the actual operations of the government.

If you move $1 from your left pocket to your right pocket, do you have $2 to spend or just $1?

The Fed’s Operation Twist works because they aren’t issuing the debt. Yes, they are part of the U.S. Government, but the Fed effectively can “print” money when they perform balance sheet expansion. The Treasury Department can’t. Say they do spend $250 billion over a couple of months buying long-duration securities and they issue short-term securities to do so. Those long-end yields come down, while front-end yields could push up and create headaches for the Fed. A Fed that has a new Chairman that (supposedly) hates how heavy-handed the Fed balance sheet usage has been.

Unless there is a coordinated effort with the Fed, which is certainly possible and we shouldn’t scoff at the idea of that happening, there’s only so much juice to squeeze.

Either Bessent’s moves are all shout and no twist, just like Yellen’s, or we are about to see another financial repression move by the Treasury and Fed.

Where Does That Leave Us Today?

If I had a definitive answer for you—well, I would probably be on an island I own and not writing this publication. Warsh has been a vocal hawk for many years before rejoining the Fed, but as I said earlier, the view from the inside sure looks a lot different from the outside. I am skeptical he is as hawkish as he says he is, and I don’t think it’s wise to discount the possibility of the Treasury and the Fed coordinating to prevent a run on U.S. debt.

The way Warsh spoke at his Jackson Hole speech today was interpreted as a Fed that may hike rates in September, but maybe he’s just shouting into the void with Bessent to see if shouting with only a hint of twist is enough to get them where they want?

I’d look for the bond vigilantes to test the resolve of both the Treasury and the Federal Reserve over the next few months. That could mean extra volatility in the markets with the AI fundamentals taking a backseat to macroeconomic drama.

If past is prologue, then it may prove wise to go long duration again soon, but I wouldn’t say that direction is clear today at all.

Interesting Reads & Listens:

• Ghosts of Electricity Substack - What Will Be Scarce?

• Apricitas Economics Substack - America’s Electricity Gap

• Noahpinion Substack - JD Vance’s Crusade Against GDP is Wrong and Bad

• Worse on Purpose - The World Cup Got Worse on Purpose

• The Rest is History Podcast - Ken Burns on Why America Really Declared Independence

• RenMac Offscript Podcast - Rotation Beneath a Calm VIX

Disclaimer: The author is not currently registered as an Investment Adviser or Investment Adviser Representative. The information contained in this publication reflects the author’s personal opinions and is provided solely for educational and informational purposes. Nothing contained herein should be construed as personalized investment, legal, tax, or financial advice, or as a recommendation to buy, sell, or hold any security.

Information obtained from third-party sources is believed to be reliable; however, the author makes no representation or warranty regarding its accuracy, completeness, or timeliness. Opinions expressed are current only as of the date of publication and are subject to change without notice.

Investing involves risk, including the possible loss of principal. The author makes no representation regarding the suitability of any investment for any particular person. Readers should conduct their own due diligence and consult an appropriately registered financial professional before making investment decisions.

To the fullest extent permitted by law, the author and The Loh-Down disclaim liability for investment decisions made based on information contained in this publication.

With last week’s post being as long as it was, I chose to leave out any discussion of Friday’s retail sales report. This week, you will not be so lucky as to escape more economic data talk.

Let’s start with the numbers, month-over-month (MoM) retail sales fell -0.6% while the MoM control group sales fell -0.4%. I always note the control group number in my data table because it contains the categories the Bureau of Economic Analysis (BEA) uses as an input when estimating consumption within Gross Domestic Product (GDP). That might make your eyes cross reading it. Essentially, total retail sales are not a clean proxy for the goods-consumption component of GDP. The BEA uses the narrower retail-control group as an input for estimating many Personal Consumption Expenditure (PCE) goods categories, while estimating the excluded categories through other data and methodologies.

So, when you see retail sales reported, the control group is often the more important number when interpreting the report’s GDP growth implications and the market reaction function. When there is a lot of volatility in the economy you can see the headline and control group numbers diverge from each other, and the market often focuses on the control group series rather than headline.

Before we discuss what July’s numbers indicate (or don’t indicate), and how it might impact Fed decision-making, let’s break down what exactly the retail sales control group is and isn’t made up of, and what happens with the excluded categories.

Who’s In Control?

First, it’s good to understand which four categories within the headline report are excluded from the control group. The four excluded categories are: gas stations, motor vehicles & parts, building materials/garden equipment, and food services/drinking places. All other categories from the headline report are included in the control group.

It is also helpful to understand how the retail sales report is derived. The Census Bureau surveys roughly 4,800 companies every month, and those firms report the nominal dollar value of their sales for the month. The Census then weights the responses to estimate national retail sales. The use of nominal dollar value of their sales for the month is important to follow here. For example, if a company sold 100 items at $10 per item they would report $1,000 one month, but if they sold 100 items at $11 per item for $1,100 the next month, the report will see that as a 10% gain in retail sales. Retail volume isn’t really growing in that example, things just got more expensive.

Price changes can influence every single category, however, most goods categories do not experience gasoline-like price swings from one month to the next, so changes in nominal sales often contain meaningful information about spending volumes.

When considering gasoline or other motor fuel sales, the BEA relies on more targeted data sources to estimate motor-fuel consumption separately, and that is the primary reason for exclusion from the control group. The practical reason most of the financial community focuses on is that gas prices are in fact quite volatile. Nominal dollar values reported at gas stations rise and fall regularly with the price of gas while volumes stay more or less static. The category could see a massive nominal dollar drop MoM, or for several months, due to a steep decline in gas prices. The volume of fuel being consumed may be the same or even more, but the price decrease offsets that and makes the number appear weaker than it is. This applies in reverse as well, if gas prices rise significantly.

Volatility isn’t the only reason the BEA might scrub a category from the headline report. The other categorical exclusions have more to do with the retail sales report lacking sufficient data for what the BEA needs to appropriately calculate GDP growth.

The BEA scrubs motor vehicles & parts from the control group because GDP growth needs to properly account for new vs. used vehicles. A new vehicle sale is a simple translation to GDP growth as it was produced then sold, and that vehicle represents current production and, when purchased by a household, current consumption. The used vehicle, however, was already counted at some point as produced and sold, and counting the full transaction value of a used vehicle would ultimately overstate GDP growth. No new goods were produced and consumed, just a used good resold to somebody else. Instead, the BEA looks at the dealer margin on used vehicles and incorporates that into GDP calculations.

Building-material dealers create a different measurement problem. A meaningful share of their sales goes to contractors and businesses rather than directly to households, so total store receipts are not a clean measure of consumer spending. Some spending ultimately belongs in PCE, while expenditures tied to residential construction and major improvements belong in residential investment. The BEA therefore relies on other data and benchmarking methods to sort those purchases into the appropriate GDP buckets.

Finally, we have food services and drinking places. When the BEA calculates GDP growth, they explicitly distinguish between goods consumption and services consumption. Consumption at restaurants/bars falls in the latter category, and must be excluded from the control group. Once those four categories are excluded, we are left with the narrower retail-control series that the BEA can work its statistical magic on.

The BEA then maps this data into the appropriate goods categories of its comprehensive PCE measurement that aims to capture both services and goods consumption. Since we’re talking about complex statistical measurements of an unbelievably complex economy, I have to be the bearer of “Are you kidding me?” news and tell you that retails sales control group data doesn’t map 1-for-1 and needs more adjustments before properly fitting the PCE categories. The good news is that after that mapping adjustment process, the BEA only has to face the final boss—inflation adjustments. Once they apply the appropriate price indexes they can finally derive a shiny data output estimate.

Please…Can We Just Talk About What the Data is Telling US Already?

I imagine my diving deep into the wonkiness of economic data drives you as crazy as it drives me while I am researching the process, but I do it from a place of love. Much like when your wife or girlfriend asks you if you’ll still love her if she were a snail, I just want to see if you’ll still read me if I talk about the absurd complexity of economic data?

Don’t worry I would much rather talk about what retail sales in July may or may not mean than how it’s derived. The simple read-through you might see out in the FinTwit sphere is that retail sales were negative, because high gas prices have finally derailed consumer spending. If you didn’t understand any of the numbers it might be understandable why you’d believe that.

Most categories were slightly positive for the month, but three components really drove the negative print:

• Non-store Retailers (ecommerce) - Non-store retail was destined to be a culprit when you factor in Amazon moving its Prime Day sale to a date in June from its usual July date. Non-store retailers saw a sizable move up in sales in June, and folks naturally took a breather from their retail therapy. If non-store retailers were to decline for several more reports, then we could be more certain that consumer strength has finally fallen, but making that call now would be premature.

• Gas - If you read the above instead of skipping straight to this section, then you’ll have already guessed that gasoline sales slowed in relation to the declining gas prices from their recent peaks. That’s a good reason to spend fewer dollars on gas. Oil started picking back up in late July and through August, but it hasn’t spiked quite enough to drive gas prices up the way they had from March to May. If gasoline sales show continued weakness as prices stay elevated, then that is the sign gas is pinching consumers in a material way. To be clear, by ‘material’ I mean consumers actually altering their driving or traveling decisions. The impact is obviously material for anybody having to pay more for gas at the pump rather than getting to spend it on something they’d prefer to spend that extra money on.

• Motor Vehicle & Parts - The motor-vehicle decline is harder to pin down, and the following explanations are more speculative. The easiest explanation is that there was a clear spike in June’s vehicle sales that was tough to repeat month-over-month. Why might June’s number have spiked so much? ‘Fleet sales’ appear to be one large factor. EV sales may have been another. They had trailed off a bit as federal tax credits expired September 2025, but after several months of high gas prices you could see their sales increasing once again. It also doesn’t hurt that several states have started offering their own EV incentive programs this year, and that the federal home charger tax credit was set to expire June 30th. Just like the other two categories, we would expect to see sales normalize in future months this year.

While the control group sales also printed negative MoM, they were less negative than headline. If the large declines in the categories above are more outlier than trend, then we should see some improvement in control group sales in August. If we instead see this decline become a trend, then we may be just a few short months away from retailers finally giving consumers their tariff refunds in the form of big holiday discounts.

Mark that one down, because weakening retail sales for several months is the only way consumers are getting any sort of tariff refund.

What About the Fed?

I know this is the subject we all wait breathlessly over. Does a weakening retail sales report in July stop the September rate hike? I wouldn’t be so bold as to say this report will absolutely prevent a rate hike in September, but when taken in conjunction with the last two months of inflation data and the weakening labor market reports, the perceived urgency for rate hikes could dissipate. Really, whether the September rate hike happens is more about the August data. Say we see a strong labor report, stronger retail sales, and renewed increases in CPI and PPI—that rate hike would become likely. Conversely, weakness in all of those reports could turn the tables and lead the market to project rate cuts in the future again, rather than a bias towards rate hikes.

That is not something I really see happening, but it is important to recognize how market perceptions could change in light of those four reports showing weakness.

In all likelihood, we see the labor market continue to be confusing, neither strong nor weak; we see inflation continue to negligibly drift down from its recent peaks; and retail sales improve from July without showing a major acceleration. Enough neutral news to stave off the rate hike again in September, but not enough to remove the hiking bias at the Fed.

There is plenty of time between now and the September meeting, and I am more than willing to change my mind once I actually see the data. Until then, I am still in the no rate hike camp.

Thanks for reading this week—next week or the week after I plan to talk a bit about Bessent’s increased Treasury Buyback program.

No, I will not be titling it Bess Buy, and I hope you’ll appreciate me more for that.

Interesting Reads & Listens:

• Noahpinion Substack - Stop Screeching About Immigration, and Get Smart About It

• Long Arc News Substack - Programmable Money Goes First

• Experimental History Substack - Stop Eating Lady Gaga’s Oreos

• RenMac Offscript Podcast - Productivity Boom Doesn’t Show Up

• Geopolitical Cousins Podcast - 16 Books That Shaped US

Disclaimer: The author is not currently registered as an Investment Adviser or Investment Adviser Representative. The information contained in this publication reflects the author’s personal opinions and is provided solely for educational and informational purposes. Nothing contained herein should be construed as personalized investment, legal, tax, or financial advice, or as a recommendation to buy, sell, or hold any security.

Information obtained from third-party sources is believed to be reliable; however, the author makes no representation or warranty regarding its accuracy, completeness, or timeliness. Opinions expressed are current only as of the date of publication and are subject to change without notice.

Investing involves risk, including the possible loss of principal. The author makes no representation regarding the suitability of any investment for any particular person. Readers should conduct their own due diligence and consult an appropriately registered financial professional before making investment decisions.

To the fullest extent permitted by law, the author and The Loh-Down disclaim liability for investment decisions made based on information contained in this publication.

The Loh-Down is so back. I hope my absence has only made the heart grow fonder, and if it didn’t—well don’t tell me.

We have a longer piece this week…A lot has happened since my last weekly write-up:

• Situational Awareness, a hedge fund geared specifically to the AI revolution, found itself over-leveraged and offside of a July semiconductor selloff and was forced to liquidate much of its public-equity book to stay alive and fight another day.

• The Japanese Yen continues to be a problem child among the big currencies. Japanese authorities’ support for the Yen has not been enough and the U.S. Treasury Department had to step in to help maintain stability. The ‘why’ and ‘how’ we intervened are the stories to watch in this space.

• The Fed didn’t hike rates in July and recent employment and inflation data are leading the market to believe September is off the table (for now) as well.

Is it too much to ask that the world slow down its machinations while I vacation? I’ll push that request up the flagpole, but for now let’s get the lowdown on each item.

Marginal Awareness

If you have not heard all the news surrounding Leopold Aschenbrenner and his hedge fund, Situational Awareness, then I congratulate you for having a life outside of finance. Leopold is a former OpenAI researcher, who wrote a paper on the coming AI buildout back in June 2024. He left his job there and began a hedge fund investing in the companies he saw as most important to that vision being realized. His hedge fund was making incredible money as that thesis played out, and those returns were turbocharged as his thesis gained notoriety. Why?

When everybody thinks you’re a stock-picking god they want to pile on to the same trades as you. Large institutional investment managers, including many hedge funds, have to file Form 13F disclosures showing certain U.S. listed holdings from the prior quarter. Investors then compare and contrast with prior filings to see what changed across the fund. You can find countless accounts on X waiting for his filing to drop to tail his trades, which is a bit of a fool’s errand considering the fund could have dramatically changed the positions between quarter-end and the day the report is released. Hedge funds can also obscure their actual positions with trades called total return swaps that give them long exposure without appearing to hold that long position.

That concept is important because these swaps can also be used to create substantial economic exposure without tying up the same amount of cash as an outright stock purchase. The opacity of these arrangements allows a nefarious person to make multiple over-leveraged bets with different counterparties using the same collateral, if they felt so inclined to skirt legal and ethical limits. Now, Leopold didn’t pull a Bill Hwang and enter into several similar swap positions with different large brokers using only enough collateral for one swap line with a single brokerage, but he was guilty of using excessive amounts of leverage in his fund. Reports and commentary have put his leverage around 4-5x. That means for every $1 in capital, he was investing $4-5 in positions. In a simplified 4-5x leveraged portfolio, a roughly 20–25% adverse move can wipe out the equity capital.

When brokerages let you borrow money on margin they will do what they need to do to protect the capital they lent you. That means they’ll issue a margin call before that 20-25% move can destroy your capital and leave them hanging on to a bad loan. The margin call requires you to up your cash collateral position so they know you’re good to pay off the loan. How do you increase your cash position if you don’t have excess liquidity lying around? You sell certain positions to raise the capital. This is where the perfect storm comes together.

Situational Awareness becomes massively popular and every momentum trader is tailing their positions. The stock prices massively outran any current fundamentals as the trade gets crowded. Your genius looks extra genius-like as the prices moon. Suddenly, though, a month like July comes around where various momentum trades begin to move against you. As the prices drop, the momentum traders rush out like cockroaches when you turn on the light, and this only serves to drop prices further. If you’ve employed the level of leverage that Leopold did, then your broker may issue a margin call and that only adds fuel to the unwinding process. What might be a 5-10% decline in the less leveraged areas of the market becomes a 40-70% drop in the momentum trades. From what we have been told, Leopold’s fund saw a -67% drawdown in July. That massive drawdown led to the margin call of all margin calls, the one where Ken Griffin of Citadel shows up at your door with a Cheshire Cat grin on his face asking if you’d like someone to take those securities off your hands for you.

Now, Leopold’s fund is still up 80% year-to-date according to the investor letter they shared publicly, but that is mostly because of their massive Anthropic holding and a few other private investments. Much of their public equity book is now owned by Citadel, and that cash infusion allows him to fight another day for his grand AI thesis. Leopold’s fund and several other over-leveraged degenerate gamblers are key culprits behind the crazy volatility we saw in the AI trade throughout July, and it’s a key reminder why leverage is an incredibly risky tool to employ. While most hedge funds employ leverage in their portfolios, it’s important to maintain a margin of awareness around the total risk you have in your books and then reduce that leverage before it snowballs into a forced liquidation to the smiling opportunist. This is especially true for a man whose thesis was set to play out over a decade-plus timeframe. Greed comes for us all.

If you want some more detailed explanations on how all of this played out, I would suggest this article or this one. Both are great reads.

Yentervention

Another story that’s been making waves the last few weeks has been a familiar one to markets over the years—an indebted Japan struggling with a weakening currency. The topic of Japan’s financial engineering of its economy needs a lot more than a few paragraphs in a weekly newsletter to really explain, so I’ll avoid too much backstory here. Suffice it to say that Japan is one of the most indebted countries in the world in terms of debt-to-GDP, and they have been fighting deflation for decades. That deflation has turned to inflation in recent years as Japan has slowly moved away from negative and zero interest rates, but the policy-rate gap with the United States remains enormous. That differential, along with fiscal and structural concerns, has kept heavy pressure on the yen despite BOJ tightening.

Japanese authorities have been intervening over the last year or two to prop up the currency when the selloff goes far enough to cause pain to its citizens. When a central bank intervenes to support its currency, they typically sell assets to raise U.S. dollars (USD) and buy its local currency—the Japanese yen in this case. Japan is one of the largest holders of U.S. Treasury debt and that is a prime asset to sell when you need to quickly raise USD. This is central bank background noise usually, “plumbing” as some call it, that can happen with any country in small ways.

The troubles start when central banks have to keep intervening because no amount of currency purchases from them seem to stop the sellers or speculators from further weakening the currency. Even then, that is usually a problem for the country dealing with the currency issue, not necessarily the U.S. It becomes our problem when that selling pressure coincides with other pressures that spark fears of contagion. What do I mean by that exactly?

The U.S. has struggled with inflation issues since the pandemic, and every time we seem to get it falling towards the Fed’s 2% target, something happens to disrupt progress. It’s unfair to label supply-chain disruptions as the culprit behind high inflation and I wouldn’t want you to leave this write-up with that conclusion. Monetary, fiscal, and other governmental policy decisions over many years have led us down this road, and supply-chain disruptions are just visible triggers worsening a problem. The War in Iran did not cause the inflation problem, but the supply-chain disruptions from it did rapidly and dramatically increase energy costs, which flow into the cost of many goods and services. Despite these inflationary issues, U.S. growth has held up well because of the AI boom. The AI boom has its own supply-chain bottleneck issues creating booming costs, and so up until June/July inflation data came out, it appeared we might be running the economy hot.

This has pushed 10, 20, and 30-year U.S. Treasury yields (interest rates) to their highest level in years. Higher yields on long bonds push the cost of borrowing up, especially for financial products like mortgages. Some people associate low long-term yields with a sign of your economy being strong and desirable to fixed income investors, but that isn’t really true. An investor wants to be compensated not only for the risks they are taking, but also for giving up other opportunities while their capital is locked up. If your economy is expected to grow rapidly, there is likely to be inflation alongside that as well, and the bond investor wants to get paid a reasonable rate of return for holding that bond over a multi-decade time horizon.

Higher long-term yields do not automatically mean investors expect default. They can reflect stronger expected growth, higher long-run inflation, greater Treasury supply, or simply a larger term premium demanded for locking up capital over decades. Whatever the mix, investors are demanding a higher rate of return to own long-duration Treasuries.

Now, just because investors may be demanding higher yields for somewhat positive reasons, doesn’t mean it can’t panic the folks at the Treasury Department. They are in charge of issuing debt for the country, and higher yields mean higher costs. It also means higher mortgage rates, which freezes housing markets further, and pushes the cost of ownership up unless housing prices were to drop dramatically. Exactly what you don’t want anytime, but especially not with midterms coming up. So when Scott Bessent started seeing Japan’s currency flailing despite their intervention, the alarm bells rang in his head. Bessent is an acolyte of George Soros, whose hedge fund famously “broke” the Bank of England back in the ‘90s. He did not want to see indiscriminate selling of the yen turn into a self-reinforcing feedback loop that spreads into other markets. The fear is that Japan would have to sell their huge holdings of Treasury securities and drive yields even higher for the U.S.

Fear of higher borrowing costs, a worsening housing market, and an eventual equity selloff explain why we intervened. The ‘how’ was equally interesting. In the past when we’ve intervened in another country’s currency troubles we would use USD to buy that currency. That obviously impacts the dollar and can impact yields, but the Treasury Department decided to sell Euros to buy the yen. Kind of funny from one perspective as we essentially roped Europe into the problem by selling off their currency to prop up the yen, but also potentially raises the ire of central bankers across the EU if they made a habit of doing so. Bessent’s immediate follow-up to this move, however, placated that potential problem. He encouraged Japan to use the Fed’s Foreign and International Monetary Authority (FIMA) Repo Facility program created at the height of pandemic fears.

The facility allows approved foreign monetary authorities to temporarily exchange Treasury securities held at the New York Fed for U.S. dollars through a repo transaction. That gives them access to dollar liquidity without having to dump their Treasury holdings into the open market. The facility was created to prevent a liquidity crisis when the world was shutting down from the pandemic, and while it was never shut down, it isn’t used often these days. Bessent encouraged Japan to leverage this, however, because it helped avoid the immediate possible contagion issue and because it stymies the speculators that might try to “break” the Bank of Japan. Sure they can keep applying selling pressure to the yen, but if Japan has a standing facility that can provide dollar liquidity against Treasury collateral, then you as a speculator might no longer want to fight that battle. Especially after Bessent went on the air to press the Fed to expand the current limit from the $60 billion per counterparty per day limit to something higher.

The Fed will have to approve an increase in that limit, so there is no guarantee Japan gets expanded capacity, but for now the facility is calming some fears.

If Bessent hadn’t pushed Japan to use this facility, it is very likely speculators would have kept putting pressure on the yen, and more Euros or USD would have needed to be sold to prop up the yen. It was a creative move on his part, but I think it speaks to the general debt tightrope we’ve been walking for the better part of a decade now.

These currency and debt issues are likely only beginning.

You’ve reached the Federal Reserve, Please Hold:

The Fed held rates steady at July’s FOMC meeting, with a 9-3 vote to hold the target range unchanged. Many folks across X and the rest of the financial media landscape were adamant that a rate hike was coming in July. While Kevin Warsh says he is committed to reaching the Fed’s 2% inflation target and the hawkish voices on the Fed seem louder, I explained in ‘Negative is Positive’ and ‘We’ve Got a Task Force For That’ the reasons against a rate hike until at least after the midterms. There is also still a case against even needing rate hikes as the year progresses, so it is still too early to lock in that view.

The weak June inflation data gave the Fed cover to avoid the hike in July, but in my view, the three dissents tell us that the Fed’s bias is to hold/hike rather than hold/cut. The likelihood of a September rate hike kept climbing after the meeting alongside climbing long-term yields. Enter last week’s employment report and this week’s CPI and PPI data.

Last week’s nonfarm payrolls (NFP) report showed a -23k job loss in July, along with a combined -103k revision to May and June. The three-month average for the NFP report through July is down to 20k jobs added per month. That is a negligible amount in an economy of our size, and puts a dent in the recent narrative of strengthening employment. The unemployment rate fell further to 4.1%, but that had to do with the continued drop in labor force participation. That is not the reason you want the unemployment rate to decline.

We know there has been a lot of noise around the NFP report the last few years, so it must be taken with a grain of salt. It does poke holes in the hawkish narrative either way. September rate hike forecasts decreased meaningfully after this report, and this week’s CPI and PPI data provided more momentum for that.

Year-over-year (YoY) CPI came in at 3.4%, while core CPI came in at 2.5%. Both numbers are lower than last month’s numbers of 3.5% and 2.6%, respectively. There were fears of energy inflation reigniting as tensions with Iran began to flare yet again, but oil prices are staying relatively contained compared to when this war first began. As long as that geopolitical constraint remains in place, or enough new supply outside the grips of the Strait of Hormuz come online, oil likely can’t fall much. It doesn’t have to increase dramatically either, though, and if it doesn’t then you at least see the inflation problem not getting worse.

That can be seen further in the PPI numbers. YoY PPI printed at 4.7% compared to June’s 5.5%, and Core PPI printed at 4.2% vs. June’s 4.7%. These numbers are still too elevated, but PPI reflects producer cost inputs. Any improvement in producer costs is good for eventual consumer inflation. There are several components within CPI and PPI that feed into the Fed’s preferred inflation measurement, PCE, and these inputs point toward a relatively subdued PCE print as well.

Between now and the September 15-16 FOMC meeting, we will see a July PCE report and August employment, CPI, and PPI reports. Barring a spike back to $100+ per barrel of oil, inflation isn’t likely to jump materially higher in August. It’s hard saying what the August employment will look like, but my bias is toward it not being relevant enough to put rate hikes on the table.

I would guess the FOMC keeps the Hawks caged in September.

TLDR

For the go-fast crew, a summary for what turned into a much longer piece than I expected:

• A hedge fund blew up, but survived to fight another day. Hopefully this time with less leverage and better risk management. Either way, Ken Griffin and his firm Citadel once again swooped in to capitalize on somebody else’s follies.

• The U.S. Treasury Department is helping the Japanese authorities prop up their currency to prevent contagion risk coming to our shores. Bessent brought a financial bazooka to the party so he didn’t face the ultimate irony of being on the receiving end of speculators breaking a central bank after being part of Soros’ team when they “broke” the Bank of England.

• No July rate hike, and inflation and employment data through July point to a September rate hike being unlikely as well. Watch out for an itchy Hawk trigger finger if oil spikes above $100 per barrel again, though.

Interesting Reads & Listens:

• Bahnsen Group Dividend Cafe - Is This the Dreaded Top?

• Rational Optimist Society - Unlocking the Ocean Economy

• Moontower Munchies Substack - DFW (David Foster Wallace)

• Kenyon College 2005 Graduation Speech - David Foster Wallace - This is Water

• Hidden Forces Podcast - Why History Can’t Prepare Us for What’s Coming Next

Disclaimer: The author is not currently registered as an Investment Adviser or Investment Adviser Representative. The information contained in this publication reflects the author’s personal opinions and is provided solely for educational and informational purposes. Nothing contained herein should be construed as personalized investment, legal, tax, or financial advice, or as a recommendation to buy, sell, or hold any security.

Information obtained from third-party sources is believed to be reliable; however, the author makes no representation or warranty regarding its accuracy, completeness, or timeliness. Opinions expressed are current only as of the date of publication and are subject to change without notice.

Investing involves risk, including the possible loss of principal. The author makes no representation regarding the suitability of any investment for any particular person. Readers should conduct their own due diligence and consult an appropriately registered financial professional before making investment decisions.

To the fullest extent permitted by law, the author and The Loh-Down disclaim liability for investment decisions made based on information contained in this publication.

Apologies for the irregular posting cadence this month. It has been a busy one for me. I will not be posting on a regular schedule again until about mid-August.

Q2 was a strong showing for the Nasdaq and my portfolios, with a 25.34% money-weighted return for the period compared to the Nasdaq’s time-weighted return of 21.57%. Portfolio returns are calculated on a money-weighted basis and therefore incorporate the timing and size of cash flows. Nasdaq returns reflect the index’s market performance and are not calculated using my portfolio cash flows. The comparison is provided for context but is not perfectly like-for-like. Either way, I am happy with the quarter, but it is just a snapshot in time. There’s still a lot of race left to run this year.

For example, a substantial portion of my Q2 return was attributable to my VCX shares that are still locked up. The stock found some new life as it briefly climbed back to $289 in late May. It has fallen precipitously since then, sitting around $41 a share at the time of writing—taking the portfolio return back down with it. Returns are essentially just a mirage until you realize them and use them to enrich your life in some way.

I covered VCX in great detail before, and have acknowledged I need to just ignore the price on that security since I cannot sell it anyway. It’s fun to watch the low-share float send the stock up randomly, but I can’t read too much into the good days unless I also want the subsequent drawdowns to be a constant raincloud hovering over me.

Best in Show

In Q2, the simple price return of eighteen (18) of my positions outperformed the aggregate portfolio return, with the top 10 price returns below.1

As you’ll notice in the screenshot from this period, instead of automatically excluding a name from showing in the list if I didn’t hold it the entire reference period, I have included it but left the note column indicating whether it met the full reference holding period or not.

The biggest price return for the period came from Penguin Solutions, but I did not buy my position early enough in the quarter to obtain that 04/01 price. My average cost is $38.73 a share, and I managed to sell 40% of my position near the recent highs around $70 a share prior to the July pullback.

I jumped into Penguin suddenly in Q2 as the AI memory bonanza was reaching a crescendo. It is well known in the AI buildout that memory and GPUs are large bottlenecks, and that drives their costs higher. Companies looking to deploy large AI solutions are doing anything they can to minimize excess GPU and/or memory purchases. Not in a way that would prevent them from adopting AI solutions they want to deploy, but it makes them look for creative solutions in the deployment process. Penguin is one of those solutions.

When enterprises use AI for repetitive workflows, models need to retain and repeatedly access context from prior interactions. They do this through a KV cache, which acts like a condensed set of notes the model can recall quickly. As the amount of context grows, however, the KV cache consumes more memory. Today, much of that data resides in high-bandwidth memory (“HBM”) located alongside the GPUs. HBM is fast, scarce, and expensive, so using too much of it for stored context leaves less available for the computational work it is best suited to perform.

Penguin’s MemoryAI server uses CXL-attached DDR5 memory to expand KV-cache capacity outside scarce GPU HBM. That allows customers to preserve expensive HBM for latency-sensitive computation while keeping more reusable context available for inference workloads. In simple terms, Penguin adds a lower-cost memory reservoir and software layer that helps AI systems decide what needs to remain close to the GPUs and what can be stored elsewhere without materially compromising performance.

The other big winner in that chart was Vishay Precision Group (VPG), which I did maintain a position in throughout the entire period. VPG makes high-precision sensors and measurement systems that help machines monitor weight, force, pressure, torque, and other physical variables. Its products serve industrial automation, transportation, aerospace, and, most recently, robotics. The investment thesis is centered on the higher-value sensing products used in robotics driving stronger growth and margins than its legacy industrial businesses. It turns out that decades spent building sensors and measurement solutions for other industries created valuable capabilities for robotic tactile sensing.

Robots are still quite early and haven’t created a massive surge in revenue for VPG just yet, but they appear to have Figure and Tesla as customers, and management has noted a third customer in the works. The two assumed customers have expectations to ramp production past prototyping levels towards the end of the year. Even if the production increase isn’t to full commercial scale this year, any increase is a great bump in revenue for VPG. That expectation for robotic ramping this year had driven their share price to astronomical levels. I started building a position last summer, averaged up a couple times, took some profits, added again, and then took some profits again as it kept climbing. I sold a $210 strike price covered call back in June, but after this dramatic July selloff, I would have been better off taking a bit more of my profits off the table.

I do think VPG will offer better returns for me over the longer term, but I made a bad call in not unloading a bit more of my position as the shares crested just before $150 per share. The momentum looked unstoppable for a minute there, and to be perfectly honest I have no idea just how large the robotics market will really become. The figures industry research groups release are pretty large further in the future, like the late 2030 or early 2040s, but it’s all a best guess. The industry may underwhelm relative to forecasts for years and then suddenly surprise us when a novel commercial application takes hold. Then those S-curves take over and random industries boom. It was easier to jump out of VCX earlier this year at astronomical prices, because I knew the valuation was absurd. For this one, the future could be way bigger than anything I imagine today, and that possibility made it easier to fall in love with the story and be lulled into less-than-ideal portfolio management decisions.

Should this one fall far enough, I may find myself buying more again, though.

Biggest Losers

The simple price return of thirty-six (36) of my positions underperformed the aggregate portfolio return in the period, with the 10 worst price returns below2.

The Fundrise Innovation Fund does not want to leave the top charts. Whether that is as a top performer or the worst performer in any given period. The price as of this writing is around $34 a share now. So it’s been more than cut in half just since the June 30th closing price. Its NAV is still somewhere around the mid-$20s, so I wouldn’t be shocked to see it fall further. The fund issuer did announce recently that they were pushing up the share lock-up date one month so that locked shares can now be sold in August rather than September. That could create additional near-term selling pressure as previously restricted holders gain liquidity, although the broader float expansion should eventually produce more normal trading conditions.

The other big price return loser in this worst 10 list that had a sizable impact on my aggregate portfolio return was Coinbase. Its shares dropped 18% in the period, and continue to be a problem child in my portfolio. Nothing has changed for the investment case I laid out earlier in the year for the company. I have been sporadically dip buying the name all year, and will continue to do so when I feel the time is right for my own personal situation. It is already the largest position by cost across my portfolios, so I am trying to be judicious in my allocation of dip-buying funds to it.

I definitely jumped the gun in my timing of this investment, but the time horizon on this one was always considered long-term.

Portfolio Changes

Additions (Excluding Dividend Reinvestment)

During the Q2 period, I increased the size of eighteen (18) positions:

• Archer Aviation (ACHR), Bridger Aerospace (BAER), Coinbase (COIN), DLocal (DLO), Fluence Energy (FLNC), Joby Aviation (JOBY), Klarna (KLAR), Kodiak (KDK), Lantern Pharmaceuticals (LTRN), Lemonade (LMND), Ondas (ONDS), PayPal (PYPL), Pfizer (PFE), Shift4 Payments (FOUR), Symbotic (SYM), Synopsys (SNPS), Tempus AI (TEM), and UPS (UPS).

I think I can safely say this about many of the companies on the Q2 additions list: their shares went on a roller-coaster ride. I am not going to spend much time on any single name as I tend to size into my positions slowly unless an idea really snatches me. This list being as large as it was this quarter had to do with some volatility and opportunities to better size up positions I was interested in increasing.

A fun one in terms of volatility was Lantern Pharmaceuticals, the same company I wrote about at the end of last year. They bottomed out around $1.12 per share at the end of Q1. As Q2 began they climbed back into the mid-$2 a share area and fell back below $2 again. I increased my position materially during that time. You’ll also see Lantern in the trims below. I reduced my total position towards the end of the quarter by about 0.9% to harvest a small amount of tax losses on older shares.

Trims

During the Q2 period I trimmed nine (9) positions:

• Duolingo (DUOL), Kodiak (KDK), Lantern Pharmaceuticals (LTRN), Ouster (OUST), Redwire (RDW), Red Cat (RCAT), Rocket Lab (RKLB), UiPath (PATH), and Vishay Precision Group (VPG).

I trimmed a large amount of Redwire when they suddenly jumped to $18-$25 a share in anticipation of the SpaceX IPO. I wish I had exited the entire position at that time, though. Space stocks have been crashing back to Earth since SpaceX peaked. Made some great profits with the shares I did sell, but not selling the rest has left me holding the bags on the remaining shares as they creep closer to my average cost again. The same goes for Rocket Lab, Red Cat, Ouster, Vishay Precision, and Kodiak. I took some gains on all of those as they had their moments in the sun. Given my long-term horizon with most of these companies, I did not sell as much as I would have liked to when looking at today’s price for each of them. I did still take profits with several of these names, however, and sold covered calls on a few as well to generate a bit of extra income as I held.

Exits

During the Q2 period I exited four (4) positions:

• Alta Equipment Group (ALTG), SpaceX (SPCX), Stanley Black & Decker (SWK), and Vail Resorts (MTN).

I covered the Alta Equipment Group and Stanley Black & Decker exits in the April recap, so I won’t repeat myself there. SpaceX is a simple story to tell. I grabbed myself a tiny allocation of shares at IPO and turned around and flipped them right away. The profit was minuscule, but I wanted to be a part of that action anyway.

I had entered a position in Vail as a high-yielding dividend play. Their stock had been crashing for the last few years after it rose dramatically through the pandemic, as so many folks picked up skiing and snowboarding. I saw the beaten-down price and high yield as a good position to be held in my IRA. After nearly nine months of holding them, however, I felt that I was finding myself sitting in a value trap. Vail picked up so many resorts in the last decade, has a lot of debt, and has also faced difficult snow conditions recently. I decided that at the moment, the money I was using for that investment could be used to pick up more shares of other companies I preferred.

I will continue monitoring their situation to see whether I can pick up its shares at an even lower price as they digest the debt levels, snow issues, and figure out if/when they will cut the dividend. At that time, it may become a sustainable long-term investment for me again.

New Positions (11)

During the Q2 period I entered into eleven (11) new positions. Although one was a quick entry and exit with SpaceX while another, Palantir, is a position I previously owned.

• Ajinomoto (AJNMY), Amazon (AMZN), Churchill Capital Corp XI - Agility Robotics (CCXI), Lennar (LEN), Navitas (NVTS), Palantir (PLTR), Penguin Solutions (PENG), ServiceNow (NOW), SpaceX (SPCX), Tesla (TSLA), and Trekor Metals (TGB).

Most of the above names were discussed in the April recap. The new names since that write-up were CCXI, LEN, PENG, and TGB. Lennar was a quick roundtrip entry and exit for me in the period. I like the name and put it back on my watchlist, but I would say that position was a victim of circumstances. I love the technology Lennar is putting together to make 3D printed homes, and I think that will be a big money-maker for them in the future. I wanted some general homebuilding exposure, and thought that was a good diversification option as I reallocated funds from profit-taking in positions like Rocket Labs. The market was gliding upwards easily at that moment and reallocation to some other sectors seemed like a solid idea, but then the selloff began again. I never sized up to a full position in Lennar, and I was not going to be able to do so in the current conditions. Money that could have been used to size up that position needed to go to higher-conviction plays that offered dip buying opportunities.

Should that opportunity in Lennar still be there when I can reallocate again, I would like to reopen that position.

Agility Robotics and Trekor Metals are the other two new names. Agility Robotics has agreed to go public through a proposed merger with Churchill Capital Corp XI. I’ve played with SPAC fire several times before and I know this may have bad news written all over it, but I can’t help myself. I think the robotics theme is further along than many realize, and SPACs have a bit of method to their madness. The desirable names tend to pump before the merger happens, and sometimes a bit into the merger and name change (from the SPAC sponsor to the acquired company name) and subsequently fall below the $10 SPAC floor. How far it goes below the $10 floor depends on the market mood and how strong that company really is. If the bullish spirits reenter this market, then I have no doubt that speculators will push up Agility’s stock price to an undeserved nosebleed level. Hopefully I am smart enough to jump ship at that time and buy back in later.

Trekor Metals is a small diversification play I am setting myself up with. They are a copper producer with operating copper assets in the United States and Canada. Copper sits at the confluence of several thematic trades, the AI trade and the resurgence of American production and mining of critical minerals. It is a theme I have invested in heavily over the last six years, and I will continue to take bites of that apple as long as the theme holds. Trekor recently brought Florence Copper into production using an in-situ recovery process that made the deposit economically viable without a conventional open-pit mine. Their stock had shot up dramatically earlier in the year alongside other copper producers, such as Freeport McMoRan, which I exited at that time, and the price hasn’t recovered since. It’s always good to have some commodities for diversification in the portfolio.

Allocation Summary

Every time I have written one of these, I tell myself I am going to reduce the number of different held securities across my portfolios and every next portfolio recap I still haven’t really done so. I end up reducing some positions but then find a new deal too juicy to pass up.

Perhaps one of these days when I tell myself I’ll reduce the number of holdings, I won’t be lying to myself, but Q2 was not that day. To be fair to myself, the concentration of my top five positions has continued to increase all year. I am making the big positions bigger, while continuing to hunt for new potential winners along the way.

The remaining challenge is ensuring the long tail of small positions earns the research and monitoring time it consumes.

Final Thoughts

While Q2 was a strong quarter, Q3 has been off to an absolutely rough start. I’ve said plenty of times before that the market giveth and the market taketh. It feels great when it’s going up nonstop, but when the market starts taking from you it’s a gut punch. During times like those, I try to remain steadfast and patient with my high-conviction ideas. For my own portfolio, I think it is useful to harvest profits along the way or trade around a core position. If the vision for the future and management execution are both there, then long-term compounding works its magic.

It is also important to separate high-conviction ideas from great stories that have not yet earned high conviction—preferably when the market is in its giving mood. Falling in love with those stocks can be painful when the drawdowns come, and I may wish I had parted ways with them while the market was still giving.

Disclosure: This article discusses securities in which I currently hold, or may hold, financial positions. I may buy or sell any security discussed at any time and without notice. Because I own these securities, I may benefit financially if they appreciate in value.

This article reflects my personal opinions and research and is provided solely for educational and informational purposes. It is not investment advice or a recommendation to buy, sell, or hold any security.

Certain statements in this article reflect estimates, projections, expectations, assumptions, or other views concerning future events. These statements are inherently uncertain, are not guarantees of future performance, and may prove incorrect. Actual outcomes may differ materially from those discussed.

Information is believed to be accurate as of the publication date but may be incomplete and is subject to change without notice. Readers should conduct their own independent due diligence and consult an appropriately registered financial professional before making investment decisions.

There was no shortage of interesting economic data this week. Welcome news on the inflation-front came from the June consumer price index (“CPI”) and the producer price index (“PPI”) reports. (See table above for more details on June’s numbers vs expectations.)

Headline CPI and PPI both contracted month-over-month (“MoM”), while core CPI was unchanged and core PPI rose 0.2%. This led to a 2.6% year-over-year (“YoY”) core CPI, and a 4.7% YoY core PPI print. All four major inflation readings were better than forecast, and have tempered any near-term rate hike concerns that had been bubbling to the surface recently. With the on-again-off-again-on-again war in Iran back on again, oil has climbed back to the $80 per barrel level. If the conflict continues, some of the energy inflation that faded in June is likely to return. That is the big what-if to watch out for on the inflationary front.

Beyond the energy-driven improvement, there were other encouraging details in June’s inflation reports. In this week’s Loh-Down, I will explain a couple of those developments and what they mean for Fed rate expectations.

One Report Doesn’t Make a Trend

One phrase you’ll hear repeated often in finance is “one report doesn’t make a trend.” The phrase is uttered from all angles when a surprising economic data release occurs—whether you were rooting for or against that economic data surprise. The contrarians who were hoping for that economic surprise are tempering their expectations when they utter that phrase. Those who saw the data move against their expectations are defending their egos and reminding everyone not to put the cart before the horse.

That phrase is a good reason why there was not a larger positive market reaction to the inflation data this week. There was a sigh of relief that the data was better than expected and rate hike expectations moved accordingly, but there is no shortage of fear that the relief will be short-lived. The fear of geopolitical tensions creating further supply disruptions hangs over the market like the Sword of Damocles. Those fears are compounded by insatiable demand across the AI buildout (power & electrical equipment, cooling equipment, semiconductors, etc.) With that overhang, markets will be looking for at least a couple more inflation reports that do not indicate a reheating of core services inflation.

If energy and gas prices don’t spike significantly again, then I believe we are likely to see inflation cool further in the CPI and PCE reports. One reason is that housing inflation continues to come down as the real estate market has been mostly frozen the last few years. Current mortgage rates remain elevated relative to the rates held by existing homeowners, while buyer demand remains constrained. Shelter is one of the largest components of the inflation statistics and, outside of a few hot regions of the country, prices are flat to barely up and even down in metros where housing supply has increased dramatically.

The way the BLS measures shelter inflation also creates a known lag in what the official components show versus what real-time market rents indicate. Shelter inflation as measured by more real-time data has come down more than either CPI or PCE inflation since the 2022 peak, and that implies a steady drift lower in future reports. As we previously discussed (here and here), the government shutdown created a quirk in CPI shelter metrics that effectively understated YoY inflation through May before temporarily overstating it through October. Despite this overcounting, shelter inflation has still climbed at a slower than expected rate. This should remain a tailwind for the disinflation watchers.

June’s report also saw a second MoM decline (2.0%) in motor vehicle insurance, which is now down 4.1% YoY. With used vehicles down 1.7% YoY, it is not shocking to see motor vehicle insurance finally show weakness. New and used vehicle prices rose dramatically through the pandemic and for a couple of years after, but eventually stalled out at high price levels. They have only started declining modestly over the last year and a half, but motor vehicle insurance continued to climb as they stalled and declined slightly. Motor-vehicle insurance premiums reset with a lag as policies renew, often on six- or twelve-month cycles, and when vehicle values and replacement costs rise on a sustained basis, you can expect insurance prices to eventually follow. Replacement parts for vehicles are still up YoY, but prices for new and used vehicles have been flat or falling for long enough that it appears we may finally see some relief in the insurance category.

Two months also don’t make a trend, but more good news like this will be welcome relief. Now let’s talk about what this might mean for the Federal Reserve.

Warshach Follow-up

These inflation metrics have lowered futures market expectations for a Fed rate hike at the July FOMC meeting to almost nothing. That was already my expectation, but since the June FOMC meeting, the market had grown quite worried that a July rate hike might be on the table. Prior to this week’s favorable inflation reports, July rate hike expectations had come down—shifting the first plausible hike from July to September. After this week’s reports, the market has pushed that possibility back toward December.

As I said in We’ve Got a Task Force for That, I no longer see a path to the 75 bps of rate cuts I predicted at the beginning of the year. If that somehow happens, I don’t think it would be a happy cut like I was forecasting back then. 75 bps in rate cuts delivered at any point this year would be more likely to indicate an economy in free-fall. My belief today is that, unless the economy noticeably breaks in a specific direction, the Fed will stay put until at least the end of the year. There are several good reasons to expect this:

• When data is murky, the Fed does not act. They bide their time as long as they need to obtain clarity. Several Fed officials have expressed hawkish views, but views do not translate into immediate actions. Patience may not be much of a virtue in the world today, but it is a Fed virtue still.

• If Warsh wants to be an effective Fed Chair during his tenure, he needs to build trust with his colleagues. Without clear alarm bells in either direction, Warsh can gain some needed credibility by appearing patient, independent, and in search of data to make the right move.

• Warsh can also avoid some of the political fervor that can come with cutting or hiking rates before an election. People will still complain about withholding action as political too, but as long as the alarm bells aren’t ringing, it’s unlikely to bring too much negative attention.

• Warsh said he expects all task force groups to wrap up by December. Avoiding any major decisions until then is easy to justify, and based on point number one above, it likely also gives Warsh more time to build support within the committee.

Given enough time I could probably expand that list, but inflation hawks could do the same with a list of reasons why the Fed should hike rates immediately. Something I learned early on in the Fed-watching world, though, is that separating what the Fed should do and what the Fed will do is key to survival.

If you believe the Fed should be cutting or hiking rates right at this moment, then you probably have a ton of data to back up that opinion. We discussed recently how perception and reality are constantly converging and diverging, and the fact of the matter today is that reality isn’t bonking any perceptions on the head. There is data showing some positive disinflationary trends and there is data suggesting inflation pressures, or inflation expectations, could remain unanchored without a forceful Fed response. Neither outcome is set in stone, and the data is not strong enough to force action.

We are in that uncomfortable Fed purgatory stage where the ‘shoulds’ make for a lot of dramatic headlines and volatility, but inaction leads the way.

If I am right that the Fed leans on inaction until at least December, then at some point the equity market should rally again. Whether a new equity rally sustains momentum will then depend on what happens after December. Does waiting until then allow the Fed to maintain the status quo longer, or even cut rates again, or does it merely postpone rate hikes until 2027?

I hope you brought a big book to enjoy during purgatory.

Interesting Reads & Listens:

• Worse on Purpose - Your Appliances Got Worse on Purpose

• Paul Graham - How to Do Great Work

• Rational Optimist Society - Unlocking the Ocean Economy

• Invest Like the Best Podcast - Uber CEO on AI, Autonomous Vehicles, and the Future of Transportation

• Union Square Ventures Podcast - Fred Wilson on 40 Years in Venture and Why USV is Automating Itself

Disclaimer: The author is not currently registered as an Investment Adviser or Investment Adviser Representative. The information contained in this publication reflects the author’s personal opinions and is provided solely for educational and informational purposes. Nothing contained herein should be construed as personalized investment, legal, tax, or financial advice, or as a recommendation to buy, sell, or hold any security.

Information obtained from third-party sources is believed to be reliable; however, the author makes no representation or warranty regarding its accuracy, completeness, or timeliness. Opinions expressed are current only as of the date of publication and are subject to change without notice.

Investing involves risk, including the possible loss of principal. The author makes no representation regarding the suitability of any investment for any particular person. Readers should conduct their own due diligence and consult an appropriately registered financial professional before making investment decisions.

To the fullest extent permitted by law, the author and The Loh-Down disclaim liability for investment decisions made based on information contained in this publication.

On Wednesday, the Federal Reserve’s Federal Open Market Committee (“FOMC”) released last month’s meeting minutes from Kevin Warsh’s first committee meeting as Fed Chair. Am I going to be talking to you in detail about those minutes today?

You can bet your bottom dollar I will not be doing that. Why? Two reasons, really:

• First, the Fed minutes only ever tell you what we already know. The folks that comb through the minutes with a fine-tooth comb are simply looking for things to scratch that confirmation bias itch.

• Second, if you have ever been involved in a board, committee, or similar type of meeting where minutes are taken then you know that the minutes are always abridged and only mildly useful. If there are juicy details from some dramatic infighting that went on in the meeting then it is very likely that the note-taker was producing one of their first sets of minutes. Real life isn’t like the movies, and these meetings aren’t a court of law where a stenographer is capturing every single detail. If there were an unedited, automated transcript—well, that could be spicy and worth reading.

So, if you find yourself cruising the internet and you stumble across somebody’s “hot take” on the FOMC minutes, I have one piece of advice for you: run away from their “not so hot” take. Our minds are like a cup constantly overflowing with useless information these days, and I can promise you that you can find better things to fill that cup with.

If your cup happens to crave a little bit of Federal Reserve information, then you’d be better served looking at the Federal Reserve Board press release from Thursday announcing who will be leading each new task force Kevin Warsh announced in June.

I have gone ahead and included that list below:

• Communications: Review how the Federal Reserve conveys policy deliberations and decisions amid uncertainty.

  • Peter R. Fisher, professor of practice, Foster School of Business, University of Washington

  • Arminio Fraga, founder and chairman, Gávea Investimentos; former president, Central Bank of Brazil

  • Mervyn King, former governor, Bank of England

• Balance Sheet Policy: Examine the costs, benefits, and institutional implications of the Federal Reserve’s current balance sheet regime.

  • Karen Dynan, professor of economics, Harvard University

  • Raghuram Rajan, professor of finance, University of Chicago Booth School of Business; former governor, Reserve Bank of India

  • Jeremy Stein, professor of economics, Harvard University; former governor, Federal Reserve Board

• Data: Improve the quality and timeliness of real economic signals that inform the Federal Reserve’s policy judgments.

  • Raj Chetty, professor of economics, Harvard University

  • Doug McMillon, former president and CEO, Walmart Inc.

  • Kevin Murphy, professor of economics, University of Chicago

• Productivity and Jobs: Assess the economic impact of new general-purpose technologies, including artificial intelligence, to inform the Federal Reserve’s policy judgments.

  • Marc Andreessen, cofounder and general partner, Andreessen Horowitz

  • Charles I. Jones, professor of economics, Stanford University, currently on leave at Anthropic

  • Asha Sharma, executive vice president and XBOX CEO, Microsoft Corp.

• Inflation Frameworks: Revisit how the Federal Reserve understands and responds to the drivers of inflation.

  • Greg Mankiw, professor of economics, Harvard University; former chairman, Council of Economic Advisers

  • Thomas Sargent, professor of economics, New York University; Nobel laureate

  • William White, senior fellow, C.D. Howe Institute; former economic adviser, Bank for International Settlements

This is a diverse group spanning business leaders, academics, and former central bankers. It will be interesting to see what sort of recommendations they come up with. If there is anything to glean about the potential direction of Fed monetary policy, it will come from understanding these folks.

Fed Communications

Fraga and King led their countries’ central banks, while Fisher spent years overseeing market operations at the New York Fed. If I had to place them on a spectrum of more communication to less the order would be Fraga, Fisher, and then King.

Fraga was the head of Brazil’s central bank from 1999-2002 following a currency crisis in the country. He is a big fan of transparency as he needed to establish central bank credibility in the eyes of Brazilian citizens so he could get a handle on unwieldy inflation. It was his view that a clear understanding of the actions the central bank was taking and why would help in managing raging inflation and currency issues.

Fisher spent many years at the New York Fed handling market functions, and he is also pro-transparency at the Fed in terms of ensuring proper market functioning and risk management. He is not known for advocating the style of communication the Fed now engages in, but he does not believe in surprising markets with erratic Fed decisions.

King is the most vocal of the three in keeping central bank communication more limited to why they made a decision. He is associated with the idea of “radical uncertainty” where central banks have to acknowledge how limited their knowledge of the future really is. We try to assign probabilities to several future scenarios, but complex economies interacting with many other complex economies ultimately make the future foggier than most would like to believe.

There ought to be productive discussion amongst these folks to limit some forward guidance, while not trying to make the Fed a “black box” of decision-making.

Balance Sheet Policy

Dynan previously worked at the U.S. Treasury Department, Rajan was a governor for the Indian central bank, and Stein was once a Fed governor. Nobody in this group is an advocate for radical shrinking of the Fed balance sheet, so the folks who worried this task force will push for a quick unwind of the balance sheet can breathe easy.

Rajan has spent years advocating for less quantitative easing “QE” by central banks, however. His focus is on how the prolonged use of QE creates dependencies to the excess liquidity in markets and economies. While he will likely argue against using QE as a routine policy tool, he has also acknowledged that you can’t just rip off the Band-Aid after economies and markets grow dependent on it.

Dynan and Stein are likely to agree with Rajan here, and based on their noted areas of expertise it is likely this task force outlines a few key things:

• A deeper understanding of how QE has permeated lending, markets, and the broader economy and how the Fed can continue reducing the balance sheet over time without creating an unnecessary liquidity crisis.

• Changing the composition of the assets the Fed holds. Many in the financial community are okay with the Fed holding larger amounts of T-bills, Treasury securities with original maturities of one year or less, as those securities align with the short-term lending function of the economy that the Fed is designed to manage. The Fed’s large holdings of MBS and longer-duration U.S. government notes and bonds remove some price discovery that private markets would otherwise perform and create the distortions Rajan fears.

• The conditions, timing, and design of future QE interventions from the Fed. There is substantial agreement in the financial community that QE when the Fed acts as the “lender of last resort” in a financial crisis is important, but many critics strongly oppose QE when it appears to accommodate massive congressional deficit spending.

Data

I am excited to see the results of this task force, and really hope they can help the Fed and the government statistical organizations get the attention they deserve to better capture reality. I have defended the quality of the folks working at government statistical agencies before, and I believe they do an excellent job with the resources at their disposal. Much of the misinformation found around their outputs is the result of politicians doing what politicians do—leveraging data and information to suit their political narratives rather than what the data tells us. If this group can provide detailed recommendations on better real-time data that is more timely and granular than many of the aggregate, survey-based releases available today, then perhaps it can lead to Congress giving the statistical agencies the budget they need to carry out the improvements. (I know, I’m an eternal optimist)

Chetty comes with extensive experience from his “Opportunity Insights” team setup during the pandemic. They built a near-real-time tracker using anonymized data from credit-card processors, payroll companies, and other private businesses. It measured spending, employment, business revenues and job postings by geography, industry and income group. The tracker provided a much more granular picture of the economy than the aggregate data available to the Fed and would have been highly useful for decision-making.

McMillon led Walmart for over a decade, and not many other businesses understand consumer behavior as well as Walmart. There are many macro researchers in finance who pay close attention to large retailers like Walmart to better understand where the economy might be heading before any of the government data starts surfacing those trends. I expect him and Chetty to advocate for better central bank consumption of anonymized private data.

Murphy likely serves as the anchor in this group to ensure that the task force doesn’t just rubber-stamp the idea that old data has to go and new alternative data, regardless of its source, should come in. This will be important because not all alternative data is created equal. Too much reliance on private, alternative data could also eventually create a new set of problems. Capitalism has been a powerful force for improving the world, but there are times when profit-chasing on behalf of shareholders blinds businesses to what is best for other stakeholders (see Worse on Purpose article in interesting reads section below.) If the government doesn’t select alternative sources carefully and pushes out its own expertise in the statistical agencies as a result, there will be a new set of repercussions one day.

Productivity and Jobs

This group seems likely to have the least variation in its members’ underlying views. All three members see AI as a transformative technology that is going to radically change how white-collar work has been performed for the last 50 years.

Marc Andreessen is an AI maximalist and will certainly advocate for AI being a true general-purpose technology that will dramatically raise productivity over time. He will push heavily on the idea of AI raising the non-inflationary growth speed limit the Fed models have historically relied on.

Sharma currently heads the Xbox division at Microsoft where a recent, massive layoff occurred. While Microsoft denies that the reorganization is a result of the “AI will take white-collar jobs” narrative, they did acknowledge that AI is changing processes at the firm and they needed to adapt to stay nimble. Her background will be helpful in understanding just how large, established organizations might be impacted by wide-scale AI adoption.

Jones is an advocate for AI as a transformative general-purpose technology too, but he will provide a measure of balance to this group. His academic work has looked at several AI outcomes, not just the idea that it will be transformative across the entire spectrum of the economy. His work has emphasized that the various bottlenecks or weak links, like the US energy grid, can limit the overall positive impact.

While there doesn’t seem to be a curmudgeon to disagree with AI enthusiasm in this group, I imagine that was on purpose. The Fed has historically relied on the Phillips curve, which essentially tells them to pump the brakes when the economy exceeds its productive capacity, particularly when a tight labor market might lead to demand that exceeds said productive capacity. That model was designed with an economy that looked very different from today’s, and probably even more different than the economy of tomorrow. So institutional Fed staff may be the curmudgeon of this task force instead, and Warsh likely stacked the group the way he did to really convince the Fed to consider a new perspective.

Inflation Frameworks

I think the findings from this task force will be the most interesting to read. Mankiw, Sargent, and White all appear to look at inflation through slightly different lenses, but the work each has produced is heavily respected.

Mankiw is more aligned with the broad framework the Fed has used over recent decades: inflation targeting and interest-rate policy generally work, but policymakers need greater sophistication in distinguishing among shocks. The Fed tries to respond differently to a demand shock versus a supply shock today, but the policy response actions could be better.

Sargent’s views do not differ wildly from Mankiw’s views in terms of differentiating policy response to supply, persistent supply, or demand shocks. However, he believes there are times when the Fed response actions do not have the same effect they did in previous regimes. If people believe the rules of the game have changed, then they will not respond to the Fed reaction function the same way they did before. Applying that framework to the post-pandemic period, consumer spending never seemed to weaken as much as the conventional models might have predicted. When the Fed hiked rates to over 5%, inflation did come down dramatically from its peak levels during that time, but Sargent might argue behavior was altered. The Fed tightened policy but fiscal policy remained, and arguably remains, on overdrive. If people believe the Fed and the government will provide a parachute for everybody the second something goes wrong, then why bother bracing for impact? It’s a more complex version of inflation expectations, and likely explains why Paul Volcker had to raise rates as high as he did back in the ‘70s and ‘80s.

White’s views are further outside the conventional framework. He is not a fan of the arbitrary 2% target as it causes the Fed to react in a blanket way rather than take the appropriate action for the conditions at hand. This would go hand in hand with the AI productivity argument. If the Fed automatically pursues a 2% target, and productivity starts leading to disinflation in several categories that pushes inflation below 2%, the Fed’s conventional reaction would be to ease monetary policy. That disinflation would not be a result of a bad economy, but the Fed has to treat it as such. The stimulative monetary policy then juices other parts of the economy through cheap credit and creates inflationary impulses elsewhere, or in the worst cases it creates the leveraged debt imbalances that then tip the economy into liquidity and credit crises.

The economic nerd side of me is incredibly interested to see how this group turns out.

Phew! That Was a Lot

If you followed along this far, you must really love reading about the Fed. I won’t hold that against you.

It will be interesting to see the outcome of each task force. My view is that part of the reason Warsh created these task forces is marketing. I think he wants to add pomp and circumstance to what is normally a standard process of reviewing policy and procedure at the Fed. There are several areas, including communication, where Powell’s Fed had already considered changes but struggled to move the institution quickly. I see the public nature of each task force as a marketing ploy to encourage a slow-moving institution to move quicker in adopting change. Warsh can also use the groups as market tests, observing which ideas generate significant volatility and which can be implemented with relatively little disruption.

That is just my take on things, and we should be careful not to assume what the outcomes of each group will be. The time to make judgments will be when the task forces reach their conclusions and the Fed begins taking action (or not). Even then, our judgments may prove wrong over a long enough time period.

As the Zen Master said, “We’ll see.”

Interesting Reads & Listens:

• Worse On Purpose - Your Power Tools Got Worse on Purpose

• Rational Optimist Society Substack - It’s Hardware’s Turn to Eat the World

• The Bahnsen Group Dividend Cafe - IPO Mania

• RenMac Offscript Podcast - The Tech-vs-HC Gap, Private Credit’s Bermuda Triangle, A Fragile Ceasefire and the Path Back to Cuts

• Made in History Videos - The Byzantine Empire

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

I hope everybody has a wonderful 4th of July weekend and is able to celebrate it with their loved ones.

I said I wouldn’t do a regular write up this week but I said nothing about not getting on my soap box to discuss something near and dear to me… On our 250th birthday as the USA, I would just like to say never let the current state of our party politics steal your patriotism from you. Loving your country while also being willing to criticize it is the ultimate patriotic act. Patriotism doesn’t actually work without both.

There are two groups of misguided folks that will try to steal patriotism from you. Those who wrap themselves in the flag that feel they have a monopoly on patriotism. They belligerently believe any disagreement with them or questioning of their political champion is a sign you are unpatriotic, and they steal your patriotism through their belligerence—making you feel ashamed to say you love your country. They are nothing more than nationalistic fools that don’t have the slightest idea of what our Founders hoped for our future. Then there are those who instead choose to vehemently hate their country—seeing nothing but atrocity in our history. Both groups are cut from the same cloth and should be seen as nothing but noise that don’t understand what it means to be an American. Don’t let their smallness deter you from being proud of being an American.

Our culture is our idealism and an ability to rise up to challenges and change.

Our country has had its share of problems along its 250-year history, but at the end of the day America has represented hope, opportunity, and the belief that we are all entitled to a life of liberty from tyranny. We teach our kids that mistakes and failure can be corrected and overcome—that those mistakes are not what define us. The brave folks that fill our history books over our 250 year arc (and the years prior to its official establishment) prove that truth time and time again.

Our Founders disagreed on who that idealism applied to and how to best meet that idealism enshrined in the Declaration of Independence, Constitution, and Bill of Rights, but they didn’t disagree on the general idea they were chasing. Disagreement along the journey is expected—hating your neighbor who chases the same goal but disagrees on how to get there is unpatriotic.

Be suspicious of those who tell you otherwise. They are the power-hungry tyrants we have been rebelling against since our nation’s birth.

Chase real patriotism.

Connect with your neighbors, help somebody in your community, or just enjoy some hot dogs and fireworks without raining on everybody’s parade.

Happy 4th of July—cheers to the next 250!

Interesting Reads & Listens:

• GZero Media - America, 250 Years Under Construction

• Canary Media - Pioneering Grid Battery Nudges California Closer to 24/7 Clean Energy

• Jordi Visser Substack - Brick by Brick: Crypto, AI, and the Discipline to See the Future Through the Noise

• Dirtcheapstocks Substack - Calling the Top

• A16z Crypto Substack - A Reading List for the Deeply Curious

• RenMac Offscript Podcast - SpaceX, Semis, and Small Caps

• David Senra (of Founders Podcast) - Rick Rubin on Finding Your Life’s Work

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

The Loh-Down will only publish economic data and an interesting reads & listens section next week

Last week I wrote about the real-time Warshach Test the markets were experiencing after Kevin Warsh delivered his first post-FOMC press conference as the Federal Reserve Chair.

I noted that, in my view, the Fed is unlikely to raise or cut rates until later in the year, if it moves at all.

That view is based on years of Fed-watching, where I have seen the Fed only act in moments of absolute necessity or where it is easy to say an action won’t set off a domino effect of destruction. Nothing in Kevin Warsh’s press conference answers indicated something different to me in that regard—he didn’t say the Fed would become gunslingers shooting from the hip. They may take direction from different data sources or models or lean on certain theories more than others, but that is about the direction of action, not the timing.

There are parts of the economy that are flashing hot today, and the stock market has been a source of loosening financial conditions as the bullish spirits run rampant with the AI trade. The job market also seems to be shedding the malaise it has carried for a bit over a year. There are plenty of indications that growth in the second half of the year will pick up, especially if oil doesn’t have a reason to spike again.

Is inflation too hot right now? Absolutely it is. Will the right move for the Fed ultimately be to hike rates instead of cutting them? That is very possible. But does that mean the Fed will hike rates at the next few FOMC meetings? No, it doesn’t mean that. You can’t rule it out anymore because Warsh has scrapped forward guidance, but that just means every meeting could be “live” going forward.

No More Training Wheels

When people say an FOMC meeting is live, they simply mean that there could be policy action taken at that meeting. Obviously with forward guidance, it was much easier to infer which meetings were live or likely to be live. Now there are no training wheels and the markets are left to infer what policy action might be taken based on the data. The thing is, data is interpreted through the lens of the person looking at it. Collectively, those different opinions coalesce around the right idea eventually, but not without some volatility in the process.

Price discovery is essentially the convergence of our perceptions of what is happening with the reality of what is happening.

That is what we have seen in equity markets in recent weeks. The AI trade had run up dramatically as investors chased momentum, but those same investors were suddenly faced with the reality that rate cuts were no longer a certainty. The frothiest trades are seeing selloffs as some profits are taken out of fear of what comes next.

The key word is ‘some,’ because dip buying is still happening throughout the selloff. This means we do not have a shared reality of the economic picture at the moment. We cannot say with certainty that the Fed will take a specific action, because the data and reaction function to that data are in the perception phase. The hawks and doves are playing one big game of chicken, and eventually the data will reach a point where reality can no longer be ignored in favor of perception.

Hawks prevail if inflation is not really just an oil-supply shock story, but instead a sign that the economy is heating up too much and demand is running too far ahead of supply. Doves prevail if the oil shock is the biggest aspect of the inflation story, and the reduction in oil/gasoline prices lets enough steam out of the inflation picture to let the Fed maintain the status quo. Important to the dovish take is that the Fed doesn’t just maintain the status quo, but that productivity gains from business fixed investment eventually help inflation come down enough to squeeze out a few more rate cuts.

If the hawks prevail, the AI trade is likely to see a massive downdraft. If the doves prevail, the AI trade probably has more innings to go before the bubble bursts.

Last Train Out

I think it’s pretty unclear which way the data is likely to break at this point, and those telling you with certainty should be seen with skepticism. Instead, it is probably worth taking both paths seriously and asking whether your portfolio is too dependent on only one outcome. Many investors may not have enough cash, downside protection, non-AI equities, bonds, or some combination of those things if the hawkish outcome wins.

There are also the folks who are much too hawkish today and thumb their noses at the AI trade—proudly saying their portfolios are basically all cash. These are the exact folks I was picturing when I described the psychological aspect of bubbles in Your Friends & Neighbors (Are Speculating).

If the data breaks in a way that allows the Fed to avoid rate hikes altogether, or even deliver a couple more cuts eventually, then they will see the bubble rise to epic proportions from the sideline. They held out on chasing “some fad” until they couldn’t stand to see the last train leave the station without them. They become the ultimate FOMO candidates, chasing the bubble late and absorbing the worst pain when it eventually bursts.

Pride and gluttony are natural parts of the human condition. I know I’m not immune. Are you?

Interesting Reads & Listens:

• A16z Substack - Flock and the Future of Safety in American Cities

• Reading Ambitiously - The New Knowledge Work

• Bahnsen Group Dividend Cafe - Data Center Drama

• Founder’s Podcast - Kelly Johnson: Skunk Works

• Hidden Forces Podcast - AI and the Collapse of State Power - Miles Taylor

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

The dawn of the Kevin Warsh era has begun for the Federal Reserve.

While the FOMC did not hike or cut interest rates today, the first FOMC meeting chaired by Kevin Warsh was chock full of information to chew on. Whether you believe Warsh and the Fed are now Hawkish or Dovish likely depends on your political or investing perception.

This week’s Loh-Down will discuss The Warshach Test we are all experiencing and dive into what I believe were the two key elements of Warsh’s limited answers during the press conference.

• The simplification of the FOMC statement and removal of forward guidance.

• The creation of five task force groups to evaluate and effect change at the Fed covering:

  • Fed communications

  • Fed balance sheet

  • Use and reliance on data sources

  • Productivity and jobs in an era of transformation

  • Fed inflation framework

AI Kevin Warsh Will Displace Fed Whisperers

Within the finance industry, particularly the institutional space, there are many brilliant people who have spent years really listening to what the Fed says and writes, then translating those words and thoughts into what they actually mean. These folks are colloquially called ‘Fed Whisperers,’ and investment professionals listen to them when the chasm between what the Fed says and what they mean is widest.

That idea can sound a bit bad when you write it out like that, and that may cause some folks to clutch their pearls in disgust that what they say doesn’t always translate to what they mean. I simply pose one question to you in response:

Have your words ever been received differently than what you intended them to mean?

I know the answer is yes. Now think about those experiences and amplify it by 100x as the Fed is under constant public, media, and political scrutiny. This was the case even before the Fed started twisting their words into pretzels to try and give statements that illustrated they’re trying to stay ahead of the curve while not committing to a specific action before it’s needed. When supporters and detractors alike hang on every word you say, you find there is no winning in expressing yourself with extreme clarity. You are trying to answer honestly, but don’t want the world to misconstrue your answer so you have to caveat everything the way a lawyer does in legal disclaimers.

So, a Fed staring down an economic environment that does not need rate hikes in terms of lending conditions, but does have a feverishly bullish stock market, has to figure out how to contain the speculative spirits without accidentally freezing everything else. Instead of saying “We are not acting on inflation today through monetary actions, but we do see issues that, if continued, would force our hand in acting. We will not pre-commit to x data = y outcome, but we will not let the inflation aspect of our dual mandate slip.” Instead, you send out mostly non-voting FOMC officials and let them talk about how they are worried about loose financial conditions and a need for interest rate hikes. That usually talks down the most speculative aspects of a market that might have been loosening financial conditions. Meanwhile, a good Fed Whisperer reminds their institutional clients that while you can’t ignore any member of the Fed, you can discount non-voter remarks when they diverge from the actions taken.

These folks will scour the statements for a removed word that would indicate the next meeting is ‘live’ for a specific cutting or hiking action should certain conditions be met.

Today, Kevin Warsh simplified that statement significantly, and he indicated that will be the case going forward. He also informed us that forward guidance will no longer be included in the FOMC statements. He plainly stated that he sees a problem with the relationship the Fed and markets have developed, where instead of the financial markets reacting to the data, they are reacting to how they believe the Fed will react to the data.

I find the idea admirable in theory and look forward to seeing it in action. As someone who has spent years in the fixed income industry doing my best Fed Whisperer impression, I can tell you it won’t be as simple as plain statements and avoiding questions you don’t believe warrant an answer in that moment. Warsh will be given some grace as the new Fed Chair to try new things and dodge questions the way he did today, but as I said above, the Fed is relentlessly scrutinized from all angles.

The Fed can speak less, but silence can be interpreted too, and you can bet the markets won’t stop trying to perceive what will happen next. With less guidance from the Fed your ability to understand their reaction function will be even more important to not being offsides on any given FOMC day.

Perhaps Fed Whisperers won’t be displaced by Kevin Warsh’s changes, just like radiologists and engineers haven’t been replaced by AI?

You Get a Task Force, You Get a Task Force, Everybody Gets a Task Force

Warsh spent a lot of his press conference answering questions about Fed forward guidance with “We’ve got a task force for that.” Reporters would try to tweak their questions as the presser went on to get him to express his own opinion on current policy, but even those were met with answers they found unsatisfactory. The furthest Warsh went in expressing an opinion on monetary policy today was that it appears “uneven.” By that, Warsh means policy is clearly restrictive in certain economic sectors such as housing, but far from restrictive in areas such as AI or the stock market.

His choice to sidestep reporter questions clearly infuriated a lot of financial folks used to Powell giving comprehensive, if sometimes empty, answers. As somebody who really respects Powell and the job he did while serving as Fed Chair, despite mistakes made under his tenure, I think everybody needs to cool their jets. It was his first meeting and anybody in the industry that isn’t a political hack or finfluencer shilling outrage knows not to expect too much out of a first meeting with a Fed he is actively trying to reshape. It also seems fairly prudent to not express your views today if you are open to changing those views based on the findings of these task force groups he is creating. Instead of complaining about a lack of information let’s briefly discuss what each of these task force groups have been assigned with.

• Fed communications

  • Warsh has been a constant critic of what he views as too many vocal opinions expressed by FOMC members and too many post-FOMC press conferences when one might not be needed. He does not seem to be opposed to them explaining how the Fed looks at data or comes to policy decisions, but he sees a stream of post-FOMC interviews and speeches expressing differing views on what should happen next creates mixed messaging and dilutes the actions taken. He also does not believe the dot plot is useful, and will look for his task force to recommend changes around that. Most institutional Fed watchers don’t like the dot plot either, but now that it might go away it has quite a few defenders.

• Fed balance sheet

  • Warsh’s private citizen opinions on the balance sheet are very clear. He does not like expanded use of it and would prefer to shrink it aggressively. Now that Warsh is in the Fed Chair position he will be forced to accept reality—reducing the balance sheet to zero in an aggressive timeframe would collapse financial markets. His task force is likely to examine and recommend ways they can shed assets slowly over time to limit use of the balance sheet to what is needed to keep the system running smoothly. They will also look at the composition of the balance sheet and see how they might be able to get the Fed out of its MBS holdings faster than the current method of letting MBS mature and then reinvesting those proceeds in T-Bills.

• Use and reliance on data sources

  • Warsh noted that CEOs of major corporations are making decisions with real-time data, but the Fed relies on data sources from the past that are stale by the time they use them to make decisions. He made a clear distinction that many of those economic data sources come from government departments that are not under the Fed, so he is not suggesting the Fed dictate what they do. He is, however, advocating for a full task force assessment of what would make the data better for the Fed’s decision-making process. The task force will also be expected to make recommendations on usage of different data sources that aren’t produced by the government agencies. The financial industry already relies on many private data sources, and the Fed looks at some of them too, but it usually discounts them when making FOMC decisions. The task force is likely to examine if some of those private sources that provide real-time data could be more heavily weighted in Fed decisions.

• Productivity and jobs in an era of transformation

  • This task force is for the Greenspan Fed fans. Greenspan saw the internet as a drastic change to the economy, and kept the Fed from hiking rates dramatically when their frameworks for a different world indicated growth and jobs were strong enough to induce inflation. That proved to be right as inflation didn’t spike despite the booming economy of the ‘90s. If you’re of the opinion that Warsh will be dovish, this task force and the one below are the ones you’ll pay attention to. They are the most likely to provide cover for not hiking rates as we see how AI transforms productivity and economic growth without inducing inflation.

• Fed inflation framework

  • Warsh has been a critic of the lens through which the Fed views and understands inflation. This task force will assess the “first-principles” drivers of inflation and weigh the full range of ideas for delivering price stability in a dynamic economy. Given Warsh’s views on AI transforming productivity and growth without inducing inflation, it is likely this task force provides the Fed recommendations to abandon the idea that strong job and economic growth automatically lead to an inflationary environment.

I have infused a bit of my own opinion into the sought-after outcome of these task force groups, and could be completely wrong about how the assessments turn out. Based on everything I’ve read about Warsh from both supporters and detractors I think the outcomes I have outlined are a reasonable judgment on what he is hoping to see out of each task force, but we just have to wait and see.

Everybody is Suddenly a Fed Expert

Following the FOMC meeting Wednesday, I decided to scroll FinTwit to see reactions to the event. It never ceases to amaze me how many people choose to opine on a subject they spend no effort trying to understand first. The number of posts stating “9 of 18 FOMC officials penciled in at least one rate hike this year, so a 2026 rate hike is basically a 50/50 shot with Warsh casting the deciding vote” was staggering. If you’re a regular reader you know exactly why that framing is so incorrect, but for those that do not read regularly, only 12 member vote at each FOMC meeting. There are several permanent voting members and a group of rotating voting members each year. We can infer who some of the Hawkish or Dovish dots belong to, but, for all we know, most of the 9 Hawkish dots could belong to non-voting members in 2026. If that were the case, then all the hawk talk online would be way off base. The same goes for any dovish talk inferred from the dot plot.

It is okay to have opinions on what you think is the right move for the Fed to take for monetary policy. I certainly express my views. My prediction at the beginning of the year was for 75 bps of rate cuts based on what we knew about the economy at the time and what I know of how the Fed responds to data. That prediction now looks likely to be wrong given where the economic data stands today. If, however, the dot plot is your basis for a particular Fed action and you carry that idea with high conviction, you are nothing more than an armchair expert with a limited understanding of the institution. They write their dot plot figures prior to any discussion among themselves during the two-day meeting. The figures can be stale by the time the Fed Chair hits the podium.

If you find somebody online who has high conviction about what the Fed should do or is going to do, and the next meeting isn’t imminent, RUN. They are your friend who thinks they could outperform an NFL quarterback. Their opinions will not serve you well.

Final Thought

There seems to be only one thing you can say with real conviction today about the Fed. Warsh does not want to be the overseer of rampant inflation, but he also does not want to impede a potential economic boom driven by productivity gains. The Fed will not take action until they absolutely have to—perhaps waiting until the task force results are all in, which Warsh said he hoped to be completed by year-end.

I’ll leave you with one prediction about the new Fed Chair. Four years from now, people who came in liking Warsh probably won’t like him and people who hated him probably will like him.

Be prepared for your quadrennial Warshach Test.

Interesting Reads & Listens:

• A16z substack - Charts of The Week: Memory to the Moon

• Campbell Ramble Substack - How to Short a Bubble

• Joe K. Raymond - 18% IRR for 57 Years

• Founders Podcast - Phil Knight: Founder of Nike

• Invest Like The Best Podcast - Legendary Trader Paul Tudor Jones on AI Risk, Bubbles, and Buffett

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

In Apple’s hit TV show ‘Your Friends & Neighbors’ actor Jon Hamm plays the character Andrew “Coop” Cooper. Coop is a successful hedge fund manager, just below the top brass running the firm. After a brutal divorce and an ill-advised workplace affair, one of the firm’s partners sees an opportunity to push Coop out and take what he built.

Coop suddenly finds himself on the outside looking in to his old life, questioning whether it was all worth it. While he questions the worth of the wealthy life he has lived, he tries to maintain his lifestyle and to regain what he lost. This leads to a steady sequence of bad decisions as he becomes a cat burglar to all the rich neighbors he has known and spent time with for so long.

The show is largely about the material excess surrounding the ultra-wealthy, and how hollow that life can become. While I highly recommend the show and think the lessons it aims to illuminate are topical in modern society, that is not the topic of this week’s write-up.

No, this week’s write-up is about a point a friend and my wife made while we were discussing how much we liked the show. They couldn’t help but notice how calm and collected Coop remains under the most intense situations. Each bad decision led him to deeper forms of chaos that would make a person sweat and break down emotionally or physically, yet Coop’s external demeanor was always that of stoicism.

My wife and friend were shocked at his ease of ignoring the chaos unfolding around him, and wondered how the character was able to do that. My response was an immediate: “Well, the character has been running a hedge fund for 1-2 decades. The volatility and pressure he had to endure to get to that level and maintain it would have been enormous. If he loses emotional control during a bout of extreme volatility, the whole operation can start to wobble. He has had to learn to stay calm under the pressure even if internally he is on the verge of losing it.”

If you read the above and found yourself angry about my willingness to compliment a flawed fictional character, I encourage you to seek help from a mental health professional. I am not trying to mythologize hedge fund managers into some zen masters they most certainly are not. Nor am I trying to condone the fictional character’s behavior.

No. I am simply trying to articulate that any job where your day-to-day could present you with nausea-inducing chaos on a whim is likely to steel a person’s nerves. Whether you make a good decision under those circumstances will vary person-to-person, but the ability to appear calm in the face of great pressure becomes natural. I remember watching the Nasdaq index hit circuit breaker limits daily during the beginning of the pandemic as the index fell 7%+ per day, or watching my company equity dry up like a raisin as Silicon Valley Bank’s failures cascaded into every regional bank that ever mouthed the words “venture banking.” My wife could attest that on the surface, I don’t look much different on these days compared to any other one—quieter, and perhaps easier to annoy, but not panicky.

Panic is guaranteed to produce bad decisions.

Bubblicious, Definition Make Them Markets Go Loco

Where the hell are you going with all of this, Cole? You’re rambling more than normal.

My point is that we are entering the phase of the bubble where a whole lotta people are about to do a whole lotta panicking. Some might read that and think I believe the bubble is about to burst any moment, but let me tell you another tale of panic that occurs before the pop.

Isaac Newton, the man who enlightened us on the mechanics of gravity and other scientific advancements, had the same character flaws as all of us. He was drawn in by the glitz and glamour of speculation. The South Sea Company was granted lucrative trade privileges in the South Seas, but the real financial magic came from its role in restructuring British government debt. Why would the British government back them in such a way? Because the company was being used as a vehicle to absorb and refinance British government debt.

Years of pointless wars left the British government with huge amounts of debt to repay, and at a certain point servicing that debt becomes quite the societal burden when handled in the standard way. Bondholders will tell you to pay them a higher rate of interest if you want to keep increasing your deficit, because they don’t believe you’ll repay them after a certain point. If you don’t want to pay that rate of interest, you can issue less debt or find other investors.

Essentially—pay up or kick rocks.

The British government found other investors by making government debt sexy. The South Sea Company had what every good startup has: a real shot at changing the world through whatever endeavor they are pursuing, and delusions of grandeur that go way beyond what could ever be accomplished in a lifetime. Real changes to society drive the initial hype. Hopes and dreams drive the speculative demand to a fever pitch.

You don’t want that gross British bond? How about shares of a company that is bringing back untold (maybe infinite?) riches from the New World? Never mind that the debt of the government has all been transferred to this company and you’re really just buying a proxy for British government debt now. Didn’t you hear from the lad selling newspapers? The upside is UNCAPPED.

Anyway, Isaac Newton, being the smart man he was, found himself invested in the South Sea bubble early and made at least a 100% profit on his speculation. Newton exited with a profit. He was incredibly wealthy, left his mark on scientific discovery, and could have ridden off into the sunset. Instead, he watched as people much less intelligent than he was make even more money on their rampant speculation as every Jack and Jill bought into the South Sea Company shares.

This is when the panic set in — not panic because something bad was happening, but the FOMO kind of panic. Newton bought back into the South Sea bubble and ultimately suffered large losses. Newton was in a cushy position at that point in his life and did not lose everything in the collapse of the South Sea bubble, but his story is representative of the times. Many folks were not as fortunate as him and did lose everything. The ruin of the overnight wealthy was felt far and wide.

That Won’t Happen to Me

Isaac Newton helped us understand gravity, but he was still not smart enough to recognize that gravity would also eventually bring down any speculative bubble too. Maybe you are smarter than him and won’t be fooled. Maybe you are more pious than the common folk and will not succumb to bubble greed. Or maybe you are just telling yourself sweet little lies.

The madness of crowds is baked into all of us, and it takes ignorance of everything going on around you or impeccable lucidity to walk away unscathed.

Anthropic, OpenAI, SpaceX, and the other marquee names of this cycle will all get their gravity-check moment. Their valuations will collapse dramatically at some point. That does not mean the businesses will collapse or that the technology is not real. The South Sea Company did not disappear after peak mania. It survived for more than a century and continued sending income back to shareholders long after the bubble buyers were wrecked. The winners of this cycle will be truly transformational too. Sometimes, shareholders who hold long enough are even rewarded, but that does not save those who bought at absurd prices.

I am young and would not dream of avoiding this speculative cycle, but I am continually looking for the offramp to do a better job than Newton did of preserving gains. What matters now is whether I maintain nerves of steel — and whether I follow the rules I set before the panic arrives.

WWCD (What would Coop Do)?

Yeah, I think I’ll follow this guy over Newton. Not because he is morally admirable, but because staying calm when everyone else is losing their minds is the whole game.

Interesting Reads & Listens:

• Canary Media - Can the US harness old oil and gas wells to produce geothermal energy?

• Hussman Market Funds Commentary - Extreme Earnings Forecasts Mask Stock Market Risk

• Experimental History Substack - Shame, them, shun them, ban them, beat them!

• China Talk Podcast - WarTalk: Still Very Much Out of Ammo!

• Hidden Forces Podcast - How China is Winning the Iran War w/Jon Alterman

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

NONFARM PAYROLLS!

You’re forgiven for assuming I was going to say Superman, but don’t make a habit of assuming things. Clark Kent will be forced to report salacious things about you in the news.

The last few months have seen a dramatic about-face in the establishment survey of the nonfarm payrolls report in terms of the number of jobs added. March was revised higher again to 214k jobs added, April’s number was revised to 179k, up from the 115k initially reported, and the May survey released today reported 172k.

Considering the layoffs reported in the technology sector, the oil disruptions from the Middle East, and general malaise one finds across the interwebs, one may be surprised when realizing we have three months in a row of strong numbers.

In this week’s Loh-Down we are going to break down this recent shift in labor market trends and discuss the implications for the economy, markets, and the Fed.

Isn’t that super? (I’m sorry, I have a Dad-joke problem.)

Caveat Emptor

This Latin phrase translates to ‘buyer beware’ and encourages you to really understand what you’re getting yourself into before you’re diving in. Applied here, it means balancing three months of increasingly strong payroll reports with enough skepticism to capture the full picture.

That is, job cuts remain elevated—3.4% higher than cuts at this time last year.

The bulk of May’s 97k job cuts came from the technology sector, which reported roughly 38k layoffs. To my regular readers this should come as no surprise as we discussed the continued increase in technology layoffs in ‘Bottlenecks or Bubbles?’ a month ago.

While this is not a new trend, it is important enough to slap the buyer-beware sticker on, because tech workers tend to be on the higher end of the income range. Higher-income workers have been instrumental to continued consumer spending strength over the last 18-24 months. Any headwind in this group of consumers raises my antennae.

Despite tech layoffs picking up steam over the last six to eight months, I don’t believe this buyer beware signal deserves much more than a simple acknowledgement that there are still plenty of risks lurking in the economy.

A few months back I might have felt different but, to me, the pick up in jobs, construction spending, factory orders, and the ISM surveys indicates an economy picking up steam.

Tech Pain, Broader Gain

My point is not to minimize the importance of high-income earners to consumer spending and the broader economy. It is that, so far, the layoff damage has not shown up as broad unemployment in the affected sectors.

Obviously technology layoffs could worsen and a net increase in unemployment in the relevant sectors could continue. If you think AI is coming for every tech worker’s job, then you would likely believe this is just the first cracks in the dam, and you’ll think I am Looney Tunes for believing otherwise. That’s fine by me because market alpha is generated by seeing through the noise, and I think AI doomerism is noise at the aggregate statistical level. People will be displaced, and the pain of losing work — something many people attach identity and purpose to — should not be minimized. If somebody says their career has been hurt by AI trends, then you should not argue that they haven’t. The impact individually is real, but it is logical to expect new companies, industries, and general economic demand that didn’t exist before to ultimately improve things in the aggregate.

Compassion for individuals is necessary; skepticism toward doomers who would block progress is necessary too.

Stepping back off my soapbox… Tech layoffs are likely to continue, but so will the hiring. Meanwhile, the additions in sectors that had been struggling are gaining real momentum.

Mining, construction, and manufacturing jobs were taken to the woodshed last year as tariffs and constantly changing trade and industrial policy took a bite out of American manufacturing competitiveness. As I said recently, America finds itself in a better manufacturing position relative to peers because of our status as one of the largest oil and gas producers in the world. Between that advantage and the Supreme Court ruling against the slapdash way Trump’s tariffs were implemented—manufacturing in America is finding some new life.

It would be unfair to attribute the resurgence to those things alone. There is a clear pivot within America to reindustrialize for the modern era, as can be seen by the startup world moving away from funding dozens of different companies that can deliver Pedialyte to hungover tech workers in 20 minutes or less, to funding physical AI and defense tech companies.

National security within the changing global order dictates an ability to extract, refine, and produce critical areas of the economy domestically or at least with dependable allies. This trend lends itself to additional mining, construction, and manufacturing roles, and it broadens the potential for economic expansion at home.

Construction spending in any given month is still volatile, but the industry is booming from the data center buildout with another 0.4% MoM gain from April. We need homebuilding to pick up to see a sustainable jump in that category month-to-month, but data center construction is keeping the sector buoyant in the interim.

Factory orders are exhibiting similar strength with orders up 4.8% YoY, which is an accelerated rate from last month’s 1.8% increase. In ‘Go For Broke’ we discussed how AI-related capex accounts for the bulk of factory spending just like in construction spending, but based on the improved labor conditions above, it would seem the impact is broadening.

If that economic data wasn’t quite enough to impress you, maybe ISM Manufacturing surveys will. The headline index is above 50, which indicates expansion, and both employment and new orders categories on the surveys continue to improve from prior months. ISM data is considered soft data since it is a survey of manufacturing businesses, rather than the so-called hard data that measures a specific output, but it can be a great leading indicator for hard data to eventually back up.

If this data all seems to support the expansion of the American economy, then why did the market react so negatively to a strong jobs report today?

Too Much of a Good Thing?

“It’s a bird… It’s a plane… It’s Superman!” is not just a phrase from the famous Superman comics, but one of Broadway’s biggest commercial flops. People love Superman. People love musicals. People do not love musicals about Superman.

Sometimes, combining things people love results in something they hate. Today, the markets saw strong jobs in conjunction with other strong economic data this week and the market did not love it. Following the jobs report there was an immediate repricing (again) in Fed Funds futures predicting rate hikes by the end of the year.

It’s fair to ask—was the reaction really about too much of a good thing, or was it because markets rallied an incredible amount the last six weeks? That rally was accompanied by high energy prices that were understood to elevate inflation the rest of the year, but also restrict consumer budgets so that supply disruptions would not be accompanied by demand-induced inflation. Strong economic data coming in with still-elevated energy prices cut directly into that narrative. I would not bet the farm on inflation spiraling out of control based on the data we have in hand, but today’s stock-market selloff certainly reflects that investor fear.

Does this mean we should expect rate hikes from the Fed later this year? I don’t believe you can rule it out under current conditions, however, the selloff and repositioning in bond markets should be contextualized within that recent rally. The market was not pricing in the possibility of an inflationary spiral and the need for hawkish Fed policy. It was a violent reaction, especially for those that only recently chased the hot stocks out there, but this selloff seems unlikely to be the bubble popping.

Just because it isn’t a popping of the bubble selloff does not mean it can’t cause a lot of pain to those not ready for it. The crazier this bubble gets the more likely we see volatile and violent swings like today, but those dramatic selloffs and rip-your-face-off rallies are exactly what lulls investors to fall asleep at the wheel. They get used to huge rallies after huge selloffs and eventually struggle to distinguish a short-lived drawdown from the real one.

Interesting Reads & Listens:

• Reading Ambitiously Substack - The next killer app is AI computer use

• Open Insights Substack - Iran’s Running Out the Clock

• Dwarkesh Patel Podcast - Why China’s manufacturing economy is dominating - Arthur Kroeber

• Sequoia Capital Interview - Andrej Karpathy: From Vibe Coding to Agentic Engineering

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

This week’s Loh-Down is all about the project I have been working on in Codex, with only a smidge of market talk toward the end. I have mentioned the project a few times in recent write-ups, and it is time to stop being a tease about it.

Why discuss this over my normal weekly macro, Fed, or economic topics? Two reasons mostly:

1. I’ve had a lot of fun building this and want to show it off. Hopefully it inspires you to build something of your own.

2. The capabilities available to anyone with a paid OpenAI or Anthropic subscription are incredible and further my belief that AI growth seems to have room to run before supply outpaces demand. So, it actually links back to markets and macro—I just prefer to take the scenic anecdotal route over the shortcut.

Without further ado, I present “The Loh-Down Command Center”

That demo was fully recorded and voiced by AI. Much like the rest of the platform, I provided the instructions and vision, and had final say on what was ‘good enough,’ but the work was carried out by my trusty AI sidekick(s).

So what is the Loh-Down Command Center exactly?

Don’t call it your personal Bloomberg, Don’t call it your personal Bloomberg, Don’t call it your…

It is without a doubt not a personal Bloomberg Terminal, but losing access to a Bloomberg Terminal was the impetus for me embarking on this Codex journey. The Loh-Down Command Center is a way to make my investing process less scattered. I had research notes, models, portfolio data, watchlists, earnings thoughts, options work, and publishing ideas living across too many places. The goal became simple: build a private investing operating system that could organize the way I actually think.

What exists now is much more than a dashboard. It is a locally hosted command center for portfolio monitoring, research workflow, options strategy, market context, earnings evidence, and writing memory. Positions are not just holdings on a screen; they are connected to thesis status, model coverage, technical indicators, research notes, and next actions. The watchlist is not just a list of tickers; it is an idea pipeline. The options module is not just contract tracking; it is becoming a decision desk for income, hedging, pricing context, and assignment risk.

The broader ambition is to make the platform feel like an investment firm’s internal operating system: market data, social sentiment, news, SEC filings, models, quality checks, and publishing workflows all feeding one environment. I want the system to help me understand not only what I own, but what deserves attention, what evidence has changed, what needs to be updated, and where my thinking may be stale.

One of my pet projects is a self-documentation function. The Documentarian layer tracks how the platform evolves: screenshots, demos, capability growth, API connections, build-cost estimates, release notes, and the story of what changed over time. That matters because this is not just a finished product; it is a living build. The history of the platform is part of the platform. That idea is how I came to build out the AI-led demo video above.

The current iteration is already useful, helping inform a couple of buy and sell decisions, but it still feels like the early version of something grander. Next comes deeper automation, cleaner data provenance, better intelligence routing, stronger research ingestion, and a more polished interface. The platform is still raw by professional standards, but the direction is clear: The Loh-Down Command Center is becoming my personal, self-documenting investment platform built around my process, my research, and my decision-making style.

Below I am going to provide some notes about my process of building this all out: why I used Codex, what plan tier(s) I used, how I structured the project for organizational purposes, and some general lessons and thoughts on working through this.

If that’s not your cup of tea, skip that section and I’ll wrap back around to a little bit of AI and markets talk at the end.

Building Notes

I started building the platform out in Codex on the basic $20 per month ChatGPT plan, but as the project picked up steam and complexity, I would hit my rolling 5-hour daily limits and the weekly limits all too quickly. I quickly realized getting the project to where I wanted it to be would take a long time on my basic tier. At the time of writing, I upgraded to the $100 per month plan with the assumption that I would try this out while I do the heavy lifting and then maybe switch back to the cheaper plan later when the infrastructure was in place. I imagine the cheaper plan will be enough to feed updates but not build out an entire project. Famous last words.

I am not team OpenAI or Anthropic, I have just been using ChatGPT since the early days of its public app launch. I have paid for their basic plan for nearly as long, and it was just easier to transition into using Codex rather than moving over to Claude Code. From what I have read, there is not a big enough difference between the two for me to care deeply which one I use. If one pulls much further ahead permanently in the future then I would use that service. Whoever gives me the best agentic sidekicks will have my business.

On organizational structure, I didn’t build Loh-Down OS through one generic development thread within Codex. I organized it in separate workflows with clear mandates, all feeding back into the same command center.

• The Loh-Down Upgrades and Mods workflow became the main product lane, where new modules, data surfaces, workflows, and dashboard capabilities get built. Here, Codex maintains a product roadmap while we chip away at features and improvements.

• The institutional Excel model workflow became the research engine, where deeper company-specific modeling, valuation logic, and IC-style thinking get developed before flowing back into Company Lab and the broader OS. This is basically the knowledge layer for me.

• The Documentarian workflow is the platform’s memory layer: it tracks what was built, why it mattered, what evidence exists, and how the system has evolved over time.

• The self-audit and hardening workflow is the reliability layer. It pressure-tests design pipes, data flows, source-of-truth logic, and fragility points so the platform gets sturdier, not just bigger. I built this after reading how easily AI-generated projects can become messy if you let agents add features relentlessly.

That structure matters because Loh-Down Command Center is not just “a dashboard.” It is an ecosystem: product surface, research engine, historical record, and internal QA process all compounding together. Each workflow has its own job, but the real value comes from the way they reinforce one another into a private investing operating system built around my process.

I also thought this structure was useful because it allowed me to have Codex work on several different things at one time without confusing workflows. If there was ever work that I thought would overlap and cause problems I would hold off on doing more work in one workflow until it could run within the bounds of the new actions taken in another workflow.

A Few Lessons From Building With Agents

• Just do it and have fun. A lot of people have used a chat bot for simpler tasks that you might use a search engine for, but not enough have tried the coding tools. If you don’t have a coding background that can be daunting, but don’t let it be. The stakes are low. Play around and see if you can build something useful for yourself.

• Control API usage. If your project is going to be connected to any third-party data there is a chance you will need to use APIs. One thing I hammered into my agents early was that “free was my favorite flavor.” I am happy to create several different free accounts to use only free API connections wherever possible. If I can’t live without a paid API down the road because certain data is particularly useful, then I am open to paying for the connection, but early on I recommend you ensure your agents adopt a bargain hunter mindset.

  • It is also crucial to ensure that any secret API keys stay local and never end up somewhere public where someone could find them and use your credits.

• Challenge and upgrade agents regularly. Many times when I was going back and forth with an agent, we would come to a point where some underlying principles or repeatable process would be unearthed. I tried to always point them out to the agent and ask it to adopt the principles within the appropriate skills file they were saving. Skills, or repeatable instruction files, are what you use to help Codex and Claude Code handle recurring tasks consistently so outputs maintain a consistent quality bar. For example, my Excel model workflow has a “model builder” skill that we iterated on for weeks that now makes building new or updating existing models much easier. In that iteration process, I used Google’s Gemini to critique my models. For each company I own and several watchlist names, I generated a model, instructed Gemini to be particularly critical, then fed the critique back into Codex until Gemini had little left to complain about. The models are not perfect — no model is — but they are more complex, dynamic, and scalable than anything I could have built on my own.

Bottlenecks, Bubbles, and Barriers to Entry

I am not going to try to convince you there isn’t a bubble forming. There is clearly bubbly behavior all over social media with speculators chasing the hottest AI supply chain bottleneck trends, or breathlessly waiting for Leopold Aschenbrenner’s 13-F from his ‘Situational Awareness’ fund to drop to know which company to chase next. Not everybody in the market is behaving this way yet, but these actions are textbook speculative mania. They’re also a required ingredient to truly make a bubble. Markets become overvalued all the time, but bubbles form when madness swells in the masses—when people believe everybody is getting rich but them. It creates a scarcity mindset.

When you first start hearing about all these random folks on FinTwit pointing out random companies set to benefit greatly from the AI trade you ignore it. Then they appear to get wildly rich and you start to wonder if you should do it too. Especially as more folks are suddenly broadcasting to the world how good at investing (gambling) they are. Maybe you or somebody you know takes advantage of one of the picks and gets a decent win and brags about it. People decide enough is enough and they open up to the mania. I mean, you’re not going to let your neighbor Carl have all the fun are you?

As more folks are drawn into the fervor, snake oil salesmen also come out of the woodworks and immerse themselves among the accounts that were actually pointing out some interesting companies or ideas. They’ll showcase the good ideas, maybe just repeats of someone else’s work, but really pump the awful ideas to anybody who will follow. Their inclusion of a good stock pick earns them the tiny bit of trust needed at the point of mania to engage in speculative purchases.

For the bubble to really take off, though, the technology or trend has to be a legitimate gamechanger to the way the world works. Without that, you just end up with crazy valuations in niche corners of the market that wipe out the speculators that chose to chase the meme. There was a time when 3D printer stocks took off to astronomical levels and then came crashing down for years. One could argue that part of the market was experiencing its own bubble, but investors in aggregate didn’t chase those stocks. It was bubbly but not a bubble. 3D printers have in fact changed the way many things are produced, but the technology’s ability to truly disrupt massive chunks of the global economy was not there.

I’ve spent enough time researching AI and using the tools to understand that the technology is a gamechanger, but building this project was one of those moments when something you’ve been told to be true actually clicks and feels like the truth. Like after a loved one tells you a bit of wisdom dozens of times, but then an acquaintance or stranger gives you the same advice and the light bulb goes off.

I have no real coding experience to speak of. I do have experience with SQL, but that was long ago, and I started a python course once. Otherwise, my credentials here were mostly my industry experience as an investment professional. I was lucky to spend a lot of time with developers at a financial reporting software company at the start of my career, so that certainly helped me with thinking about how the back end of systems need to function too. All you really need to do is have a conversation with the agent and use your experience in other fields to help direct it. When I started this project I simply made a copy of my full investment and writing folder, gave the AI access, and told it some of the things I currently do with my research, writing, and investing. I asked it to ask me questions to understand things better and after some back and forth, it was off to the races on building out the product.

AI tools keep getting more advanced, and that will lower the barrier to entry for all kinds of folks that previously felt limited by not having the technical skills needed for their entrepreneurial ambitions. At many firms, the failure to build a particular upgrade often has less to do with a lack of ideas or ambition and more to do with a lack of staffing and time.

AI is breaking barriers down. Software companies aren’t going to collapse, and engineers still matter. Some companies will end up on the losing side of AI as the technology continues to be adopted, but I have no doubt that some incredible ideas will proliferate as well.

The more I build, the more convinced I am that the next wave of AI adoption will come from people turning domain expertise into software.

Interesting Reads & Listens:

• a16z Crypto Substack - Entering the era of the headless merchant

• Reading Ambitiously Substack - What I am teaching my kids about AI

• Colossus - Beyond the Sky Hyperliquid & Jeff Yan

• On Purpose Podcast - Jan-Emmanuel De Neve: The Simply Daily Habit Linked to Happiness

• RenMac Off-Script Podcast - Burden is on the Bears

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

This week Kevin Warsh’s nomination to become Fed Chair cleared the final hurdles in the Senate. Congrats, Kevin, the tightrope is now yours to walk.

As much as Trump would like Warsh to deliver some rate cuts immediately, that seems very unlikely to happen. Even the most dovish members of the FOMC, such as Chris Waller or Michelle Bowman, think the appropriate move at the moment is to keep rates on hold until they get a clear signal to take action.

At the end of the day, Warsh is one person on the committee. Yes, he is the Fed Chair now and that makes him the most important FOMC member, but he still has to build consensus among the voting members. Warsh is likely to spend the next few months working behind the scenes to earn the trust of colleagues who may view him skeptically, while also changing the communication strategies at the Fed. Warsh has previously criticized the Fed’s forward guidance through the SEP, along with the constant interviews and speeches from FOMC members. His view is not that the Fed should be opaque, but that it should avoid over-communicating, muddying the picture, or locking itself into a policy path when conditions can change quickly. I guess like most things with the Fed, it’s a delicate tightrope to walk. I hope Warsh has been keeping up with his exercises to improve balance.

Consumers Laugh in the Face of Inflation (for now)

The Consumer Price Index (“CPI”) and Producer Price Index (“PPI”) both remained elevated year-over-year (“YoY”), at 3.8% and 6.0%, respectively. The good news for the Fed is that within the PPI, airfare services rose less than they did in CPI, while financial services fell in PPI despite rising sharply in CPI. These two components feed into the Personal Consumption Expenditures (“PCE”) report, which is the Fed’s preferred inflation gauge. As I noted above, PCE will still be too strong for the Fed to cut rates now, but any sign of easing inflation pressure, technical or not, is welcome if economic growth is being impeded by the oil disruption in the Middle East.

Given the pace at which consumers are still spending, some think the Fed may not even need to cut rates. It seems everybody is angry about inflation and higher gas prices, as the University of Michigan sentiment readings show consumer malaise, but it sure isn’t preventing those consumers from swiping their cards. Control-group retail sales, the measure that feeds into GDP calculations, came in at 0.5% month-over-month (“MoM”). That is lower than the March reading, but still strong considering how much of consumer budgets are being eaten up by gasoline right now. Sporting goods and electronics both increased 1.4% MoM and e-commerce increased 1.1% MoM.

Furniture, clothing, and vehicle sales all fell, so there are cutbacks in certain categories. The billion-dollar question is whether those cutbacks spread further or stay contained?

2022 Redux?

If we look back to 2022 when Russia invaded Ukraine, we saw a significant gas price spike. The first month of retail sales after the spike was strong, as consumers continued spending normally while also paying more for gas. The second month was much weaker, but still showed strength. By the third month, spending turned negative before flatlining for most of the year. If history is repeating, that would not bode well for the rest of the year. But is a true repeat of 2022 really in store?

The Fed went on one of the most aggressive rate hiking campaigns in history at the same time the gas prices surged—playing a large role in the reduced consumer spending that year. While some at the Fed seem open to rate hikes, I would be shocked to see that carried out. Could we end up with no rate cuts this year? Yes, that is absolutely in the realm of possibility today given how conditions have changed since the beginning of the year. A rate hike would require a sizable jump in inflation from already elevated levels. Not impossible, but it should not be anyone’s base case. That is especially true given the recent rise in long-term Treasury yields. The 20- and 30-year Treasury yields are comfortably above 5% today, while the 10-year yield has climbed to 4.6%. Markets are doing the job of tightening financial conditions for the Fed.

If the Fed lets the bond market do the work and simply maintains rates where they are for the rest of the year, then what happens?

The gas price surge should continue eating into parts of consumer spending, but it may not be as bad as 2022, especially when you factor in the tailwinds.

The economic data table shows that capacity utilization and industrial production both came in stronger than forecasts and above their prior-month readings, at 76.1% and 0.7%, respectively. These are both positive signs for labor and GDP growth. We also know AI data-center capex remains off the charts. Then, while high gas prices squeeze consumer budgets, the U.S. energy industry will make substantial profits selling their oil and gas to the world at higher prices. This reduces some of the consumer drag.

The U.S. is in a significantly better position to weather this disruption than Europe or Asia. Gasoline prices in the U.S. are much higher today, but electricity prices have not shot up in the same way. There were higher electricity bills across the country generally before the war in Iran, but since the disruption has begun electricity prices have not increased as dramatically as gasoline prices have. The U.S. electricity mix is fairly diverse with strong renewable and nuclear production, but domestically produced gas leads the way. The U.S. produces more natural gas than it consumes, so domestic natural gas prices have not surged the way they have in much of the rest of the world. Lower relative energy costs make American production more competitive in the short term. This is a bad reason to be able to compete on price better, but it may be the reality of the situation for a short-to-medium timeframe—depending on how long it takes to get a resolution in Iran.

From where I’m sitting, it is reasonable to expect some consumer weakness from here, but I would not bet on an outright weak consumer or weak economy just yet.

Warsh has just begun his walk on the tightrope, but investors have been walking one for years now.

Interesting Reads & Listens:

• Exponential View Substack - The Case For Radical Solar Optimism

• Campbell Ramble Substack - Don’t Take The Bait (War, Oil, AI, and Why Everything Is the Same Trade)

• Campbell Ramble Substack - Don’t Invade The Heartland (2,500 Years of Data Explain How To Win From the Periphery)

• Campbell Ramble Substack - Waking The Hegemon (From Quiet Hegemon to Reluctant Arkhe)

• Blind Squirrel Macro Podcast - Stock Market on Shaky Foundations

• Simply Put Podcast - Fabio Natalucci on the Rise and Risks of Private Credit

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

Back-to-back big weeks for economic data and earnings releases. Macro and equity enthusiasts find themselves buried in updates to review and memos to write. Don’t feel bad for them, though—this is their Super Bowl.

Let’s break down some of this week’s highlights.

We Are So Back

Factory orders came in at 1.5% MoM—well above the 0.5% estimate—and on top of an upward revision to the prior month reading to 0.3% from 0.0%. Strong factory orders are good for jobs and good for GDP. A big beat and prior upward revision on this gets two thumbs up.

Construction spending also came in above estimates, printing at 0.6% MoM vs estimates of 0.3%. It was not as impressive as the growth in factory orders as prior month construction spending was revised significantly lower to -1.9% in March. Either way, positive growth is encouraging here, but contextually the 0.6% was not enough to climb out of the hole from the prior month. Good, not great.

ADP’s payroll report and the government’s non-farm payroll (“NFP”) report both came in with solid prints. ADP’s payroll report showed 109k jobs added in April after adding 61k in March. The 109k print was below estimates of 118k, but it was still two months of consecutive job growth. The NFP print of 115k jobs added in April was below the upwardly revised print of 185k jobs added in March, but still a strong showing considering the weak reports earlier this year.

So far, 89% of S&P 500 companies have reported earnings—these are the quick hit stats from that cohort:

• 84% reported a positive earnings per share (“EPS”) surprise

• 80% reported a positive revenue surprise.

• The blended year-over-year (“YoY”) growth rate sits at 27.7%, which is the highest since Q4 2021’s 32%. (Hmm, right before the market meltdown in 2022)

• Analysts already bumped up S&P 500 Q2 earnings estimates 3.2% from their March 31st estimates.

Earnings remain strong and forward guidance showed no real signs of cracking yet. These results keep markets at cruising altitude.

Not So Fast

We would be doing ourselves a disservice if we only highlighted the positive things from this past week, because there are a few things that can give an investor pause.

Building permits fell 11.4% MoM—this follows the -11% report from March. That large housing starts number we saw recently was a good sign for future GDP growth, but without strengthening building permit numbers following close behind, momentum may fail to build.

The Challenger, Gray, & Christmas layoff report showed a 38% jump in layoffs from March to April with 83k workers laid off. For the second month in a row, AI was the highest-cited reason for layoffs. If we expand the scope to 2026 year-to-date (“YTD”) the top five cited reasons for layoffs are as follows:

• Market and economic conditions - 53,058 cuts

• Closings - 52,187 cuts

• Artificial Intelligence - 49,135 cuts

• Restructuring - 42,307 cuts

• Contract Loss - 34,484 cuts

It makes sense for AI to be a top five cited reason for layoffs given the power of the latest models, but I also would discount some of the amount attributed to it so far. It’s not uncommon for companies to lean on the industry topic du jour as an excuse for mass layoffs, when it is more a result of poor management decisions and underperformance from expectations.

While some of the folks that are laid off find new jobs, as can be seen by the recent pickup in the payroll reports, there is a negative trend quietly happening since December. At that time, the labor force participation rate was 62.4%—as of April it has fallen to 61.8%. That’s a sizable decrease to the workforce in five months. This drop in the labor force participation rate helps explain part of the puzzle as to why the unemployment rate hasn’t increased further with the increased layoffs this year and in 2025.

So where does this mix of good and bad leave us?

Smells Like Teen Volatility

When Kurt Cobain wrote ‘Smells Like Teen Spirit’ he was inspired by graffiti left on his wall by his girlfriend’s friend and bandmate that stated “Kurt smells like teen spirit.” Cobain thought it was provocative and tied to prior talks of being revolutionary—leading to the now infamous track. Teen spirit in fact was a brand of spray deodorant and the bandmate was only making a joke that Cobain smelled like a teenage girl’s deodorant spray.

This really has nothing to do with where this economic discussion is headed. I just felt the story was a necessary price to pay to keep Cobain from rolling over in his grave the way I used his song title.

I do think we should expect continued volatility in the markets, though. Over the past few weeks I’ve given several geopolitics and economic reasons for the markets to be fearful, but it has been able to climb those ‘Walls of Worry’ so far. This is mostly attributable to AI-driven economic expansion. There is a lot of doubt about how far that can keep carrying the economy. I think it’s fair to be skeptical about that, but that skepticism, to me, is one of the things that underpins the momentum the trend still carries. Bubbles don’t pop when so many people are skeptical. They pop when disbelief is suspended and folks feel an existential pull to be a part of the financial speculation for fear of being left behind.

The way random company stock prices within the semiconductor supply chains have popped because “they are potential bottlenecks that stand to benefit from years of demand growth” is definitely bubble behavior. I finally feel that we are in the midst of the real bubble, but that doesn’t mean the show is over just yet. I don’t know what metric will be the one to suspend disbelief in the AI-growth trajectory, but I am sure it’s coming. I feel Codex-pilled after building out this entire portfolio management, research, and writing platform I have been working on. I could spend endless amounts of time working on new improvements or features to make my work easier, better, and scalable.

There are plenty of risks to the economy and growth is likely to remain uneven, but I think the AI trade has legs still. A pullback at some point soon seems enitrely reasonable, but otherwise it sure feels like bubble season.

Interesting Reads & Listens:

• The Great Progression Substack - AI Could Trigger the Biggest Productivity Boom Ever - Bringing Many Surprising Benefits

• The Rational Investor Substack - Big Tech: The Perfect Business Model Thrown Away

• Bahnsen Group’s Dividend Cafe - Private Credit Contagion Risk And All The Lies

• Invest Like the Best Podcast - Interview with Energy Trader John Arnold

• DRM News - Full Speech: Finland President Declares End of Western-Led World Order at Raisina Dialogue

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

For April, my aggregate money-weighted portfolio return is ~7.99% vs. the Nasdaq’s 15.31%. This represented an underperformance versus the index of 7.32% for the month, although I am still outperforming the index by 10% YTD. The Nasdaq delivered a classic ‘rip your face off’ rally as negative sentiment over the situation unfolding in Iran was reaching a crescendo. The situation remains unresolved, but the markets are forward-looking in nature. Through the fog of war, if they squeeze their eyes really hard from the crow’s nest, they are totally sure they can kinda/sorta see some land.

I am unsure whether this rally stays strong here, or finds a moment to scare investors with a quick and steep selloff. As I discussed in Climbing the Wall of Worry, the market is primed to climb as long as the situation doesn’t get worse.

Since I wrote that piece, things have neither improved nor worsened, but the tension is clearly simmering. The longer this remains at an impasse the more likely the physical oil market constraints will collide with the paper market prices in a way that squeezes consumer pocketbooks and restricts some economic activity. It doesn’t have to be this way, but the longer the Strait of Hormuz remains closed the more likely these realities collide.

One positive out of left field was a temporary ceasefire between Russia and Ukraine in honor of their holidays commemorating World War II. This may prove to be nothing more than a temporary holiday ceasefire, but if this can be built into something lasting, it would be a big help for turning down the geopolitical thermostat.

Best in Show

Eighteen positions held through the reference period saw simple price returns above my 7.99% aggregate money-weighted portfolio return. Below are the top ten.1

Lantern Pharma went from my biggest loser in March to a top three gainer in April. Nothing screams fun like massive 50%+ swings in both directions. Substantial volatility does give you the opportunity to dollar-cost average and I took full advantage of that during the month. I am hoping the company is able to disclose some commercial revenue contracts for their withZeta AI platform. The company demo of the product was impressive and they provided a few breadcrumbs about some potential deals in the works or perhaps already underway, but nothing investors could sink their teeth into.

The more interesting takeaway from the top four gainers in April is that three of them are related to robotics and autonomy.

• Vishay Precision Group is believed to be providing sensors for Tesla Optimus robots, and Tesla is claiming the Fremont, California plant is being retooled to increase production of those.

• Aurora Innovation is one of several key players in the autonomous trucking space. Their stock has been on a tear lately after announcing an expanded partnership with Hirschbach Motor Lines to scale an autonomous fleet up to 500 vehicles starting in 2027.

• Ouster has closed their acquisition of StereoLabs and just announced their new digital LiDAR sensors, which feature the world’s first native color LiDAR. There are indications that Google and Volvo may adopt this new sensor for their vehicles. We should hear more during the upcoming earnings announcement.

Biggest Losers

Twenty-five positions saw simple price returns of less than the 7.99% aggregate money-weighted portfolio return. Below are the top ten losers.2

Well, the biggest loser by far was the Fundrise Innovation Fund (VCX) that I provided so much detail about in the Q1 portfolio recap.

The fund has actually climbed back a bit since hitting $94 last week, but as I said in the Q1 recap, I expect the price to trade closer to NAV over time as the obsession over private companies dies down a bit and when all the unrestricted shares are available to trade. Either way, I exited all my unrestricted shares and now only hold the restricted position.

The real big losers for me in April were Fluence Energy and UiPath. Fluence continues to drop dramatically from the highs it achieved earlier in the year, and I have added to my position on its way down. I do not believe it will drop all the way to its Liberation Day lows between $3-$5, because the revenue growth and backlog seem to be too strong for that. There could, however, be more room to fall in the short term until a catalyst emerges. I don’t mind rebuilding my position after selling shares at higher prices earlier this year and last year.

UiPath is a tough one. I believe they will eventually demonstrate progress on being the Orchestration layer for enterprise AI automation, but it might be a slow battle that will be reflected in the stock price. I bought a lot of shares in them over the last couple months as they sold back off to the low teens—a larger position size than I originally intended. I will likely reduce the overall position size down at some point by taking some tax losses on the higher average cost shares.

Portfolio Changes

Additions (excl. dividend reinvestments)

I increased my positions in several companies in April:

• Core Positions: Fluence (FLNC), Lantern (LTRN), Tempus (TEM), Uber (UBER), and UiPath (PATH)

• Dividend Payers: Pfizer (PFE), UPS (UPS), DLocal (DLO)

• Speculative bets: Archer Aviation (ACHR), Bridger Aerospace (BAER), Joby Aviation (JOBY), Lemonade (LMND), Ondas (ONDS), and Reddit (RDDT)

For readers who would like more detail on why I am purchasing more of a security on any given day, you can upgrade to my paid Substack. Paid subscribers receive more detailed notes on how I think about markets, individual companies or positions, risk, valuation, and portfolio construction. I also share my positions list, watchlist updates, research notes, and periodic commentary on my portfolio changes.

Trims

In April, I trimmed the following positions:

• Redwire (RDW), Red Cat (RCAT), SentinelOne (S), DuoLingo (DUOL), and Fundrise Innovation Fund (VCX)

DuoLingo, Redwire, and SentinelOne are all examples of me doing some tax-loss harvesting. I had done some dollar-cost averaging in all three positions earlier this year. I increased them to larger portfolio concentrations than I wanted long-term—with the intentions of selling off some higher average cost shares to offset other locked-in portfolio gains this year.

Red Cat was me locking in some more gains on the company. They are a speculative drone company that seems to be carving out an important role in the industry, but they are prone to deep drawdowns. Sometimes you need to take some gains when a speculative name has moved meaningfully in your favor.

Exits

I exited two positions in April Stanley Black & Decker (SWK) and Alta Equipment Group (ALTG).

• SWK was a dividend payer I added to my IRA in Q3 2025. Around that time I opened several dividend payers that had sold off in recent years, and SWK seemed like a reasonable income and maybe price recovery play. After a couple of dividend payouts the stock had increased from my ~$65 entry price to $91, but I didn’t sell there of course. I waited until April and took a much smaller 10% return plus the dividend payouts. I still kind of like SWK around current levels, but I thought the yield and potential price opportunities were better in UPS and Pfizer, which is where I allocated a material amount of the proceeds from SWK.

• ALTG was a position I opened at the beginning of the year. They were never a high-conviction play, but one that sounded good on paper as a speculative bet on a capex construction boom. For this exit, it’s not so much about the thesis changing. I bought into the stock after its initial rally and there was a lot of volatility while I owned it. I have high conviction in American reindustrialization, AI capex, the electro-tech stack, and the supporting energy and power needs that come alongside those themes. But ALTG was only ever an additional, speculative shot on goal. In hindsight, my exit was poorly timed, locking in an ~4.5% loss. If I had waited a bit longer in the month, I could have had a 20 or 30% gain on my exit. I definitely regret the exit on this one and may regret not giving this more time to flourish, but sometimes it’s better to cut positions that make you constantly question whether you chased a theme too sloppily. I can sleep at night knowing I have other shots on goal.

New Positions (5)*

I initiated five new positions in April:

• Ajinomoto (AJNMY): A Japanese food company mostly known for their Umami and MSG seasonings, or their frozen foods in Trader Joe’s aisles. Why would I purchase a company like that? They also happen to dominate Ajinomoto Build-up Film (“ABF”), a specialized insulating material used in advanced semiconductor substrates. ABF is critical in high-performance chip packaging, and Ajinomoto is widely viewed as the near-monopoly supplier, with estimated market share around 95% or higher. I thought I would grab one of the semiconductor bottleneck themes emerging in the market. We’ll see if my gamble pays off.

• Amazon (AMZN): Amazon doesn’t need any introduction on what they do. The crux of the investment here is that they haven’t seen much price appreciation over the last five years the way the other hyperscaler/MAG7 group experienced. They’ve been diligently working to expand LLM usage in AWS, their trainium chips are seeing more production and demand, and their robotics/logistics moves are starting to show returns on investment. I am hoping it may prove to be a solid entry point for me on a longer-term hold.

• Navitas (NVTS): As I watched several supply-chain bottleneck stocks skyrocket this year while I did not get in on the action, I have decided to throw my hat in the ring on a power-semiconductor company that could provide a meaningful improvement in energy efficiency to AI data centers as the buildout continues. See detailed explainers here and here on why GAN and SiC chips are likely to be a big cog in the AI data center of the future due to their better power delivery capabilities to the analog power chips of today.

• ServiceNow (NOW): ServiceNow is a cloud-based AI platform designed to automate and manage enterprise workflows. Historically you might have seen this platform when working with your IT Helpdesk at work. Historically they did not really sit in the RPA space like UiPath, but they are both now looking to sit atop the agentic AI layer helping businesses manage all of enterprise automation. Their stock price has seen a significant drawdown over the last year leaving it near five-year lows. They are leaning heavily into AI and LLM integration across their platform, while trying to position their product as an agnostic orchestrator of autonomy like UiPath is trying to do. It is going to be a heck of a fight going forward and I will likely have to be agile in my positioning in both of these companies as things unfold.

• Tesla (TSLA): Like Amazon, Tesla needs no introduction in terms of what they do. The why now is also similar to Amazon. Tesla’s stock price had struggled over the past year compared to some of the other Mag 7 companies, but all signs were pointing to a re-ramp in growth. The battery pack business has grown significantly for Tesla, and given the solar + battery trends globally, I would expect that growth to continue. The robotaxi business will obviously roll out commercially slower than anything Musk ever says, but the safety-driver testing is increasing and when it is good to go the economics scale quickly. Tesla semi-production has finally begun at their factories, and I would guess Optimus enthusiasm picks up by the end of the year—regardless of whether it is actually go-time in that arena.

Allocation Summary

The allocation summary section of these portfolio recaps…The section where I swear next month I will reduce the total number of positions in my portfolio.

Once again, that month of position reduction is set for a future date. When excluding Cash, MMF, and consolidating the Bridger Aerospace investment into one parent level “position” I am still holding 48 different companies across my portfolios.

Over time, I do intend to whittle this down—it’s a marathon and not a sprint, though.

Final Thoughts

While I think a bit more concentration in my equity portfolios will make it much easier to manage the investment process, I have also been leveraging Codex to build myself a comprehensive portfolio management platform. It’s my own personal Bloomberg, albeit nowhere near as good. It is very personalized, though, and provides me with a tool that can help me analyze and manage positions, monitors various market momentum gauges, builds complex financial models for owned and watchlisted positions, screens covered call and other option ideas for me, monitors earnings, and includes a publishing dashboard to help me aggregate and track ideas in my writing over time. I will continue sinking my teeth into this project as many things I’ve built still feel rather unpolished in my eyes. One tool to help in this endeavor is a self-documenting process I had Codex build into the platform to monitor the improvements made to the system over time.

This will allow me to do a full write-up with screenshots and a demo video of the platform soon. I will showcase what I’ve built to manage my investments more like I did at the institutional level.

Today, I can include a teaser screenshot of the dashboard page.

The goal of this platform is to ultimately scale my research, writing, and investment process. It has already done that, and really demonstrated just how powerful Codex and Claude Code really are.

One final note: I will be moving my Portfolio Recap write-ups to a quarterly cadence. It will be a little easier that way and they can be a bit more comprehensive.

Go For Broke is the 1965 board game where players race to see who can lose a $1 million fortune first. Players try to squander their fortunes through gambling at the casino, stock market, racetrack, and buying unvisited properties. Nobody would fault you for thinking this sounds a lot like scrolling through FinFluencer pages on social media.

I think one day this may also become an apt description for some of the hyperscaler companies—whichever ones crash and burn from the buildout. They are all spending increasingly large sums of money every year to try and ‘win’ the eventual AI fortune. The big name executives acknowledge the spend is eye-watering, but the choice is easy to make. If they don’t spend, they will lose. If they spend, they may still lose, but there’s a chance they win. So, they have to spend. Let’s use this week to dive in to what ‘Going For Broke’ looks like in the economy.

Capex, Capex, Capex

There was no lack of top-tier economic data releases this week: Durable Goods Orders, Housing Starts, PCE inflation, GDP growth, Employment Cost Index (wage growth), Personal income and spending, and ISM Manufacturing survey data for April.

Strip away the inflation data and you have some pretty solid numbers. Non-Defense durable goods orders, excluding the volatile aircraft sector, came in at 3.3% month-over-month (MoM). This was on the back of an upward revision of February’s number from 0.6% to 1.6% MoM. This surge is mainly related to AI data center spend, and shows no signs of letting up. Growth in capital goods purchases is almost always a sign of economic engines firing, as those goods are used to produce future output.

This large surge in durable goods orders is easy to pinpoint in the 2.0% quarter-over-quarter (QoQ) Q1 GDP growth figures. This is evident in the ‘investment’ category.

The investment category is the belle of the ball right now, although the real star of that show is the ‘business fixed investment’ sub-category—surging 6.2% in the quarter. The AI data center continues to buoy the economy while residential investment lags and consumer spending muddles through tariffs and energy supply shocks. A 2% real GDP growth rate for Q1 is nothing special, but in the context of the energy supply shock, it’s acceptable. I’ve discussed several times that the OBBA provides some massive capital expenditure (capex) incentives for companies this year. Spend more on large capex projects domestically and you can depreciate the full expense immediately. This incentive was made permanent last summer and the race for AI supremacy is hot, so business fixed investment slowing down this year seems unlikely. Get in while the gettin’ is good.

Hungry Hungry Hippos

To this day, I still believe my childhood friend manipulated his Hungry Hungry Hippos board game so that he could always consume more marbles than the rest of us and win the game. The case has gone cold, and the evidence is lacking, but the suspicion will always burn bright.

Today, if you removed the growth from the AI data center buildout it appears the economy would be operating at stall-out speed. Yes, the hungriest of all the hippos is hands down the AI data center buildout. One could even make the argument that AI and data center spend are hoovering up all the available capital—to the detriment of other hippos industries—and that capital could be allocated elsewhere. There may be some merit to that argument, and perhaps if that capital was allocated across a wider array of industries we’d see a stronger economy as wider participation induces stronger consumer demand. We want this because strong U.S. economic growth rate is typically accompanied by strong consumer demand.

Business fixed investment is growing significantly, but the breadth of that growth is narrow. Consumption is growing at a slower than usual pace, well below what the AI data center buildout is contributing—accounting for ~1.5% of the 2% real GDP growth in Q1. The ideal outcome from here is that business fixed investment continues increasing and widens to other industries, while consumer demand picks up as business fixed investment benefits accrue.

Based on Q1 earnings so far, that outcome certainly looks possible. That positive scenario largely hinges on the outcome in Iran. A resolution sooner rather than later gives the US an economic speedbump, but then the path to growth should be clear from there.

Otherwise, if the other hippos don’t start getting fed, we may have to switch board games.

Housing Starts, Powell’s Final Presser, & What’s on Deck

Housing starts in March were a positive surprise this week, increasing 10% MoM after February’s numbers were also revised upward from -3% to +1.8%. We will want to focus on construction spending in the months ahead to see how these new housing starts filter into GDP growth, but another month or two of strong housing starts would be a clear indicator for GDP expansion as the year progresses. We will be sure to circle back to this topic as more data comes in, and we will want to really scrutinize the second half of the year as the 45L homebuilder credits expire June 30, 2026.

Wednesday was Powell’s final press conference as the official Fed Chair. The Fed left rates unchanged as expected, but they did have the most dissents at a FOMC meeting since the early ‘90s. Stephen Miran dissented in favor of a rate cut. Meanwhile, Beth Hammack, Lorie Logan, and Neel Kashkari dissented over the “inclusion of an easing bias in the statement at this time.” Those weren’t the only fireworks at the meeting. Powell indicated that he would stay on as a Fed Governor until he believes the investigations into him and his colleagues are completely done and not going to be reopened.

Powell also made clear that he will support Kevin Warsh as the new Fed Chair and will not be running some sort of Shadow Fed Chair. He knows how tough the job is and will only look to support Warsh in his endeavors. Powell seemed to indicate that his staying on as a Governor is a principled stand that policy disagreements between the Fed and the White House must not be settled with manipulation of the justice system. The Fed, under Powell, has both made mistakes and acted in ways that most certainly prevented a financial crisis. Everybody loves to be an armchair QB, but it’s a tough job up there. Given the way the Administration kicked the investigation back to the Inspector General to finalize, the case should be closed fully again. It’s hard to imagine Powell staying on the Board for many more meetings. The man making this joke at his final press conference as Fed Chair is not trying to avoid retirement for long.

The era of the Powell Put is now over. We will be on Warsh Watch going forward.

Speaking of things to watch for—next week is another busy economic data week: ADP employment, non-farm payrolls, JOLTS survey, Challenger job cuts, weekly unemployment claims, factory orders, building permits, and the ISM services index.

Unless one of these reports is significantly above or below expectations, it is unlikely they will create much of a splash in markets. Not while the market is locked-in to the unfolding geopolitical situation, and earnings season. It will still matter to read the economic tea leaves a bit, but focus is elsewhere.

At present, markets are only accepting ‘applications for attention’ from problem children.

Share

Interesting Reads & Listens:

• Sophie’s Notes Substack - The Man in the Arena Isn’t a Hall Pass

• Pinebrook Capital Substack - U.S. Economic Growth Update

• a16z Substack - Institutional AI vs. Individual AI

• One Useful Thing Substack - The Shape of the Thing

• The Ideas Letter - The Next Great Transformation

• Global Macro Update Podcast - The End of the Middle East Energy Era

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

I don’t know how many more times I can talk about a “will they, won’t they” ceasefire agreement in the Middle East and the potential implications of that, so I simply won’t talk about that this week. In fact, I will keep this week’s write-up short as I have spent most of my time this week working with Codex to build out a platform for my research, writing, and investments. Yes, Codex ate my homework this week. Reaching my token limits every day was much more enticing than coming up with clever ways to write up about geopolitics or economic drama. I have no intentions of making the platform a ‘for public use’ tool anytime, but I will eventually show off what I am building for feedback. I hope to share more in a future write-up.

With that in mind, let’s do a brief dive into the bit of economic data that mattered this week, and the recent move by the administration that should clear Warsh to take over as the new Fed Chair.

Survey Says…

This week’s economic data was light. The most important print of the week was the retail sales report. The report came in above consensus at 1.7% month-over-month (“MoM”), but much of the increase appears to be from higher gasoline prices—rather than an actual expansion of spending activity by consumers. The University of Michigan survey reflects sentiment, not actual spending behavior, so it is considered “soft” data. That being said, consumer expectations and current conditions had been falling continuously for some time. Last December, sentiment reached a decade low. The numbers are still in the “too low” territory for the survey historically, but seeing consumer sentiment stabilize and begin to tick higher over the last few months is an interesting turn. This feels notable given ongoing energy disruptions stemming from the conflict with Iran. Certainly worth watching as the year progresses, even if there is no real story there today.

While the data itself didn’t move markets much this week, the policy backdrop may be about to shift.

Next week’s FOMC meeting is expected to be Powell’s final meeting as the official Fed Chair, as his term concludes in May. While he could remain in a temporary capacity depending on the timing of Warsh’s confirmation, the transition in leadership is now approaching. Senator Thom Thillis said he would only hold up Warsh’s nomination until Powell was safe from prosecution over the construction costs for the Eccles building that were not within Powell’s control. Based on the DOJ officially dropping their investigation into Powell, I would say this clears that path.

When a Fed Chair finishes their term, they can still stay on as a Governor until their term limit runs out, but most former Fed Chairs tend to leave the organization to let the new leader have their time in the sun. Powell has kept his cards close to his chest regarding his likelihood to stay or leave, but this had everything to do with the questionable investigation brought against him. If Powell believes the investigation is done for good, then he is more likely to leave—letting Warsh preside over the Fed without his input.

Warsh will have his work cut out for him as he tries to build consensus for further rate cuts. He is unlikely to build that consensus for at least a meeting or two, possibly longer, depending on how the data evolves. With his hands tied by the data for now, Warsh is likely to focus his efforts somewhere else. He has been critical about the problems he sees with the way the Fed handles forward guidance forecasts. He believes they are too rigid in how they are presented and pigeon-hole the Fed into staying on a certain course longer than they should to create an air of credibility.

Powell shared that same complaint of the Fed, and had been pushing for a change to how they present their summary of economic projections at quarterly Fed meetings. He was unsuccessful in implementing the changes, but he has laid the groundwork with many Fed Governors and Regional Presidents to get on board with the idea. I believe we’ll see Warsh run with that idea quickly, and perhaps manage how other Fed communications are handled. This would be his “low-hanging fruit” to grab while he waits out inflation data that prevents the Fed from cutting rates.

With the case against Powell dropped, Kevin Warsh’s nomination can now move forward. Less confusion and drama at the Fed is a positive for investors. It does not change the picture on what will give us further rate cuts and when, but it is one less question mark on the market’s mind.

That is always welcome news.

Interesting Reads & Listens:

• Visual Capitalist - Charted: The Growth of $100 by Asset Class (1965 - 2025)

• Canary Media - Balcony Solar is Taking State Legislatures by Storm

• Sophie’s Notes Substack - Science Fiction Isn’t Predictive, but It’s Important Anyway

• Experimental History Substack - The One Science Reform We Can All Agree On, but We’re too Cowardly To Do

• RenMac Off-script Podcast - Is The Bull Case &@#%!?

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

Less-bad is Good

The most notable economic data this week was the March Producer Price Index (“PPI”) print. The PPI essentially measures wholesale prices—the price domestic producers receive for their output. As we know from dissecting other economic data measurements, the devil is in the details, and we should use reported numbers as a guide—not gospel.

For March, the PPI is still elevated at 4% YoY and 3.8% YoY for core PPI, but the estimates for the print were notably higher at 4.6% and 4.2%, respectively. Less-bad data than expected is welcome news, especially alongside a dramatic rise in energy prices. This data also matters because the Fed’s preferred measure of inflation is the Personal Consumption Expenditures (“PCE”) index. In sectors like healthcare, financial services, and air travel, producer prices—not consumer prices—feed PCE inflation data. In the March PPI, all of these categories except for air travel showed smaller increases than the March CPI.

This tells investors that the March PCE number should come in lower than feared prior to the PPI release. It may even come in below March CPI. Now, we’re unlikely to see the Fed deliver rate cuts over the next few meetings, unless the economy deteriorates quickly. Markets try to price moves by the Federal Reserve before they actually happen. So, while a better-than-feared print for March PCE won’t bring us fresh rate cuts, it helps prevent the markets from pricing in rate hikes this year.

This matters because when markets price in future rate hikes, financial conditions tighten and liquidity contracts. Tighter credit and liquidity can create a negative feedback loop that slows down the economy and makes some of the doomsday scenarios discussed in financial media more likely. In How I Met Your Redemptions, we discussed how private credit issues are more likely to stay contained rather than become systemic.

A negative feedback loop that dramatically tightens economic conditions right as energy supply chains are disrupted could risk that containment. I am not saying that is what will happen, but if private credit issues were to boil over into something systemic, those would be two clear catalyst candidates to watch.

Howard Marks’ latest memo on private credit is a good read if you want to dive deeper.

Vertical Limit

I watched 2000’s heart-pumping, mountain-climbing movie Vertical Limit recently. At the time, I took away two lessons from this movie—the first was to not trust Bill Paxton on a mountain. Ever.

The second was that summiting massive peaks is a high-risk/high-reward endeavor for climbers, analogous to trading in markets during periods of chaos. To seek the reward at the top, one must climb high, but the higher you climb, the farther there is to fall. You can’t think about falling while you’re climbing higher, of course. You must not lose your focus, but the reality lurks in the back of your head the whole time. In moments of chaos, problems pop up at every turn—threatening the markets from climbing higher. If traders want to reach the reward at the top, they must don their climbing gear and attempt to scale these obstacles.

This is known as “climbing the wall of worry” in markets.

Today, we find ourselves at a juncture where every data print that is less bad than feared lowers the likelihood of something blowing up, and that is a balance we should hope to maintain until there are less unknowns swirling around. Each time we avoid the worst-case, markets will burst higher—like they have over the past two weeks. For those who bought the dip, it makes for a euphoric feeling. A euphoric feeling that can be snatched away quickly if any of the obstacles prove too tough to climb.

As long as the Middle East situation doesn’t ratchet up in intensity from here, then the U.S. equity markets appear positioned to run higher. The U.S. is shielded from the oil disruption relative to peer countries and most oil is denominated in USD, which helps explain the run-up in the dollar since the conflict began. A stronger dollar, relative outperformance economically, and the AI trade bring money into the U.S. markets in the near term. It wouldn’t be surprising to see a small pullback soon after moving up this quickly. Perhaps after a long-term ceasefire is finally announced—a ‘buy the rumor sell the news’ type of event.

Where markets could actually stumble as the year progresses is if the second and third-order effects from the energy disruption cause more problems for a wider swath of companies than is currently expected. The breadth of that impact is key. If financials and AI remain strong, then the economy and markets can stay buoyant while the worst of the disruption is worked through. An economy doesn’t go off the rails when the financial engine and its leading sector are still chugging along.

Final Thoughts

I know when I see my portfolio marching higher day after day—I can’t help but feel anxious about a fall. Last year, I felt it constantly when the markets climbed relentlessly after the DeepSeek and Liberation Day selloffs—that there were so many ways things could go wrong and the market was ready to drop at any second. That may be the case one day, but it is difficult to predict with real certainty. That’s especially true if you’re not someone who lives and breathes markets. I do my best to ignore the chaos—stay invested, take some profits along the way, and occasionally buy index put options for insurance.

Scared money don’t make money, but prepared money can.

Interesting Reads & Listens:

• Rational Optimist Society Substack - What I’m Telling My Kids About AI

• Mauldin Economics: Thoughts from the Frontline - Conflicting Data, Conflicting Results

• Musings of a Wandering Mind Substack - The Imagination Curriculum (book list)

• Campbell Ramble Substack - The Death of the New Deal

• Artificial Ignorance Substack - The Emerging “Harness Engineering” Playbook

• Founders Podcast - How Roger Federer Works (Fantastic listen)

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.

I’m back from a week of wedding shenanigans in Mexico, and I might actually be less rested than when I left. Scratch that—I am less rested. I’m here, though, and ready to talk about the economy.

Jobs Are Here

The biggest economic news last week was the nonfarm payrolls (“NFP”) report and its massive beat over expectations. Consensus was for 65k jobs in March, and the report came in at 178k. January numbers were revised up by 34k to +160k, while February numbers were revised down by 41k to -133k. Q1 average monthly job growth now stands at 68k. Better than last year, but still low for the size of our economy.

The unemployment rate dropped from 4.4% to 4.3%, but this drop stems from a decrease in the labor force—those actively looking for jobs. That is not the reason we want to see the unemployment rate drop.

The clearest sign in the report that the higher job additions and lower unemployment rate are not as good as they initially appear is the continued decline in average hourly earnings.

Year-over-year earnings growth has slowed from 4.1% to 3.5%. We don’t have a high unemployment rate, but the labor market is not exactly signaling strength. If it were, wage growth wouldn’t continue decreasing. When companies are desperate for labor and the labor market is as tight as it appears, wage growth would climb as companies pay up to get roles filled.

This labor market elicits a shoulder shrug and a “phew, glad it’s not worse.” Meanwhile, inflation still holds center stage as the problem child.

Inflation Angst—It’s Not a Phase, Mom

The Consumer Price Index (“CPI”) for March printed at 3.3% year-over-year (“YoY”) and Core CPI, excluding food & energy, printed at 2.6% YoY. Skeptical folks will tell you that Core CPI, or other slicing and dicing of data, exists to serve politicians. That is not the case. Statisticians just want to look at the data from all kinds of angles—politicians being able to pick and choose what they highlight as relevant is just a byproduct. Core CPI exists because food and energy prices are heavily influenced by volatile factors that monetary policy does not directly control. The Fed can influence broader demand, but it cannot do much about a sudden oil shock or a weather-driven jump in food prices.

When you look at Core CPI, you are attempting to see what everything else is doing in the economy. If oil is causing headline inflation to spike, are other categories spiking in conjunction, or are they falling? If they’re falling, then there is slack in the economy and consumers and businesses can’t keep up with price increases elsewhere. If prices are increasing alongside it, then the economy is running hot enough that price increases are met with more, or at least the same demand.

For March, we are seeing other categories tick up slightly but not at the pace of the energy categories. April through June will be important inflation series as they will paint the picture of whether price increases downstream from energy are coming, or if there is no room for that in consumer or business budgets. If it is the latter, affected businesses would have to take squeezed profit margins or preserve margins by cutting labor. Any relief in energy prices alleviates some of that problem as well.

A happy(ish?) medium would be some relief in the energy category, and the other categories continuing at a moderate pace. This would signal businesses can pass along some costs, avoid making the labor market any worse, and the economy can muddle along until the energy disruption is in the rearview mirror.

What does this mean for the Fed and the markets?

Dueling Mandates

The Fed has a dual mandate: price stability and maximum employment. These two mandates have been in tension over the last few years—fighting for attention. Given the current inflation situation, the NFP report puts the Fed policy Hawks in the driver’s seat. Talk to anybody that is unemployed and they will tell you this job market is difficult. The time between jobs can be incredibly long, but the low unemployment rate gives the Fed cover to hold off on further rate cuts. They know this job market is challenging, but inflation is too high and rising again. They feel they must focus on the inflation side of the dual mandate.

Chair Powell has made attempts, however, to quell market fear that this energy disruption will cause the Fed to hike rates this year. Whatever inflation arises from the energy spike is likely to take a bite out of economic growth rather than perpetuate an inflationary cycle. The Fed is willing to look through the price spikes and not hike rates to get in front of inflation, but they won’t cut rates until they have good reason to move.

As I’ve said in recent pieces, we are in a waiting game.

One that I still think ends up with several rate cuts by year-end. Either the energy disruption clears up faster than expected and the Fed can deliver soft-landing-style cuts later this year, or energy remains elevated long enough to roll the economy over and force cuts for a much worse reason.

There is also a risk that Congress learned nothing from the extra consumer stimulus provided after we were already coming out of the pandemic, and some sort of bill passes that materially stimulates consumer demand. That could turn a potential transitory energy and supply chain disruption into another inflationary cycle. That outcome is likely to put the fear of rate hikes back on the table.

There are so many other ways all of this could turn out, because the world is a complex place. There are always countless variables at play, but today’s world seems especially uncertain. Tectonic level shifts beneath our feet in how the world and geopolitics operate around us. Change breeds instability and instability leads to volatility in the markets.

The markets are starting to price in some stability out of the Middle East going forward. I wouldn’t be surprised to see starts and stops along the way, but the direction appears to be towards some form of “normal” again. That’ll be good for the markets overall, but if markets rise too far too fast we could find ourselves facing another pullback at some point. Q1 earnings should come in strong for most businesses, but the rug pull on the bulls would be management teams providing weaker Q2 or full-year guidance out of caution. The flip side of that is U.S. energy players will have stellar earnings and forward guidance for the year, so S&P 500 earnings in aggregate could continue to climb or at least maintain expectations.

If energy earnings are very strong while tech earnings remain strong, then it might be the bears who see the rug pull on them. Bears expect the downstream effects of high energy prices to crush overall earnings—leading to a deep selloff. They would be positioned short and things turning out better than expected could lead to a total melt-up in the market that torches their short positions.

I don’t know which way things will break, and I don’t think you can position for certainty these days. For my book personally, I am trying to maintain my equity exposure so I capture the seemingly inevitable upside over the longer term, hold some cash to buy dips, and buy index put contracts for downside protection when I can’t make heads or tails of things. Everybody has their own coping strategy.

Mine is a little cautious optimism…and a Peroni.

Interesting Reads & Listens:

• High Agency Substack - How to Solve Any Problem

• Questionable Intelligence Substack - Intention is All You Need

• Canary Media - Heron Power Raises $140M to Modernize Electrical Transformers

• One Useful Thing Substack - A Guide to Which AI to Use in the Agentic AI Era

• Noahpinion Substack - China is Killing the Fish

• RenMac Off-Script Podcast - Is The Bull Case &@#%!?

• Geopolitical Cousins Podcast - The New Rules

Disclaimer: The author is not currently licensed to provide personal investment advice and is not professionally managing funds. All information provided in this publication is the opinion of the author for educational purposes only and should not be construed as investment advice. The third-party information contained herein has been obtained from sources the author considers reliable, but the author does not represent that the third-party information is accurate or complete. The author is not making any recommendations regarding the suitability of particular investments for any particular person. Before making any investment decision, it is important to do your own research, or contact a certified professional. The Loh-Down assumes no liability for any investment decisions made based on the information provided in this newsletter.