The U.S. 10-year has fallen back to roughly 4.74–4.75% after reaching 4.818% Wednesday, while the 30-year is around 5.23%. Both have moved back below the levels I’ve been watching for renewed long-duration stress:

10Y warning zone: 4.80–4.85%

30Y warning zone: 5.30–5.35%

That’s important because long-term yields remain one of the biggest variables affecting growth stocks and the AI infrastructure trade. For now, that pressure is easing.

The catalyst today came from the Fed. Governor Christopher Waller indicated he could support leaving rates unchanged in September if disinflation continues. Markets responded by reducing expectations for another hike, helping pull Treasury yields lower.

But Waller also said something much more important for the longer-term thesis: the United States is losing some of the premium historically attached to safe, liquid government debt.

His explanation gets directly to what I’ve been writing about. Massive federal deficits and an enormous AI infrastructure buildout are competing for the same pools of capital. That may keep the neutral rate, and ultimately long-term borrowing costs, structurally higher than investors became accustomed to during the post-GFC era.

So today’s message is somewhat paradoxical:

Near term: bullish.

Structurally: still a warning.

Japan is helping today

Japan is moving in the right direction as well.

The 10-year JGB has fallen back below 3%, while the 30-year JGB declined after a successful long-bond auction. That reduces one of the transmission risks we’ve been monitoring: sharply higher Japanese yields encouraging capital to leave U.S. Treasuries and return home.

USD/JPY has also fallen sharply toward 155.5–156.

That’s important. The further the yen moves away from 160, the lower the immediate probability of another disruptive currency intervention and associated Treasury-market unwind.

The BOJ is still moving toward tighter policy, so Japan remains a structural risk. It simply isn’t making the U.S. Treasury problem worse today.

The dollar confirms the improvement

DXY has fallen to roughly 98.9.

Put the major signals together:

U.S. Treasury yields ↓

JGB yields ↓

Yen ↑

DXY ↓

Equities ↑

That’s almost the opposite of the cross-market combination that would make me increase hedges.

The dangerous configuration remains something closer to:

U.S. long yields ↑ + JGB yields ↑ + USD/JPY toward 160 + DXY strengthening + equities weakening.

We’re not seeing that today.

What the GP Alpha Risk Monitor is telling us

This is exactly why I built the GP Alpha Risk Monitor.

I don’t want readers making portfolio decisions from one headline, one Treasury yield or whether the Nasdaq happens to be green that morning.

The monitor continuously combines the major variables affecting market risk, including Treasury yields, the dollar, Japan/JGB conditions, liquidity, monetary policy, inflation, market internals, technical conditions and other macro inputs…and converts them into a live risk regime.

Today’s live Market Risk reading is approximately:

33 — CONSTRUCTIVE

Remember how to interpret the number:

Lower risk = more constructive.

Higher risk = progressively more defensive.

But the number by itself is not an automatic buy/sell signal.

GP Alpha separates two questions:

Strategic regime: How much underlying market risk exists?

Tactical posture: Are the verified conditions present right now to add, hold, hedge or reduce exposure?

That’s why readers should pay attention to both the Risk Score and the Action Plan.

How I’m using it today

With Treasury pressure easing and the broader cross-market signals improving:

HOLD: Core S&P, growth and high-conviction AI infrastructure positions.

SELECTIVE ADD: Increasingly attractive if the 10Y holds below approximately 4.75%, the 30Y moves toward 5.20%, DXY remains below 100 and Japanese yields stay contained. Individual stock technical signals still matter.

RE-HEDGE / TRIM: If the 10Y returns to 4.80–4.85% or the 30Y returns to 5.30–5.35%, particularly if Japanese yields and the dollar begin rising simultaneously.

MATERIAL DE-RISK: Still reserved for approximately 10Y ≥5.00% or 30Y >5.50%, especially if accompanied by DXY above 100, USD/JPY moving back toward 160 and renewed JGB stress.

One important point for readers

If the Risk Monitor ever says TACTICAL UNRESOLVED, don’t interpret that as bearish.

It means exactly what it says: GP Alpha does not currently have enough verified evidence to authorize a new portfolio action.

In that situation, the instruction is simple:

Hold existing core positions. Pause new adds. Wait for verification.

I’d rather have the system refuse to make a call than manufacture confidence from incomplete data.

Bottom line

Today’s move materially reduces near-term equity risk.

The structural Treasury problem hasn’t disappeared. If anything, Waller’s comments reinforce the longer-term argument that federal borrowing and the AI capital cycle are competing for an extraordinary amount of capital.

But that’s not what the tape is signaling today.

Treasuries are improving. Japan is behaving. The yen is strengthening. Equities are responding.

For now:

Hold cores. Add selectively when both the GP Alpha Risk Monitor and individual technical signals confirm it.

That’s exactly what the monitor was built to do, separate long-term concerns from what the market is actually telling us to do today.

Finding a company I want to own is only part of the job.

The harder part comes afterward.

The stock runs 50%. Or 100%. Rates move. Earnings come out. The macro environment changes. A market moving Trump Tweet on X. The stock gets crushed for no obvious reason.

Do I buy more? Hold? Wait for a better entry? Trim?

That’s what I’ve been building GP Alpha to help answer.

For subscribers, I want to explain how I’m actually using it.

Start with the portfolios

I’m gradually moving the stocks and investment themes I’ve written about into portfolios inside GP Alpha.

These aren’t random screens of stocks.

They’re companies and themes I’ve researched and written about that I want readers to be able to follow over time.

Every stock has an actionable signal:

STRONG ADD — long-term trend is confirmed and the stock is near what the model considers a favorable entry zone.

ADD — conditions remain constructive and I’m comfortable building the position.

HOLD — I want to own it, but I don’t want to chase it here.

TRIM / REDUCE — risk/reward has deteriorated enough that I want less exposure.

Because a great company isn’t automatically a great buy at every price.

These signals aren’t random

I want to make this part clear.

I’m not sitting behind the app manually changing AVGO from HOLD to STRONG ADD because I woke up bullish on Broadcom.

Each stock is evaluated through the GP Alpha proprietary signal framework.

The model looks at several layers simultaneously: the long-term trend, price and entry levels, momentum, technical structure, price extension, portfolio weighting and the broader GP Alpha Market Risk environment.

That last part matters.

The individual stock doesn’t exist in a vacuum. Rates, liquidity, the dollar, Fed policy, credit conditions, Japan and the broader market environment can change how much risk I’m willing to take even when I still love the company.

Those inputs ultimately produce an actionable signal:

STRONG ADD → ADD → HOLD → TRIM / REDUCE

I’m deliberately not publishing the exact formula or weighting behind the model. That’s proprietary.

But readers should understand that there’s a repeatable framework underneath the signal. It’s not a green button generated from somebody’s opinion that morning.

The objective is to remove as much emotion, FOMO and guessing from the process as possible.

Example: AVGO — STRONG ADD

AVGO is a good example.

The current signal is STRONG ADD.

But notice what GP Alpha is not saying.

It isn’t telling you to immediately put 10% of your portfolio into AVGO.

The target weight is 10%, but for someone opening a new position:

Start with 1/3 of your 10% target position.

That’s roughly a 3.3% initial position.

If you already own AVGO:

Add only if you’re below the 10% portfolio target.

That’s how I want people using this.

Build positions.

Don’t chase them.

ADD doesn’t mean go all-in

Example: TSM — ADD

TSM currently shows ADD rather than STRONG ADD.

The trend remains constructive and price-extension risk is limited.

Again, look beyond the green ADD label.

New position: Start with one-third of the target.

Already own it: Add only if you’re below the target.

The signal answers:

Is this an attractive place to add exposure?

Position sizing answers:

How much risk should I actually take?

Those are two different decisions.

Portfolio construction changes the target

Not every GP Alpha portfolio has ten stocks.

A ten-stock equal-weight portfolio has a 10% target per position.

A six-stock equal-weight portfolio has approximately a 16.7% target per position.

GP Alpha adjusts for that automatically.

Example: COHR — ADD

COHR is in a portfolio where its target is 16.7%.

That does not mean someone starting today buys a 16.7% position.

The app tells a new investor to begin with one-third of the target and build from there if the setup continues to qualify.

If you already own it, the system considers how much exposure you have relative to the portfolio target.

That’s the difference between a signal and a portfolio-management system.

Don’t confuse performance with the signal

Example: AAOI — ADD

Look at those returns.

YTD: +204.7%

Since 1/1/25: +188.2%

It’s easy to look at that and think:

I already missed it.

Or worse:

It’s going up. I better buy it before it goes higher.

Neither is an investment process.

GP Alpha isn’t giving AAOI an ADD simply because the stock has gone up.

The system is evaluating the setup today.

Trend.

Extension.

Entry levels.

Momentum.

Risk/reward.

Portfolio exposure.

Broader market risk.

That’s also why you’ll see VIEW LEVELS AND SCORE underneath each company.

If you want to understand what’s underneath the headline signal, go one level deeper.

The individual stock is only half the equation

This is where GP Alpha becomes much more useful to me.

I don’t evaluate AVGO, TSM, COHR or AAOI in isolation.

GP Alpha is simultaneously monitoring the broader environment.

Treasury yields. The dollar. Liquidity. Fed policy. Japan. Credit. Market breadth. AI infrastructure. Crypto. And other variables feeding into the risk framework.

So there are really two questions:

Is this stock attractive here?

And:

Does the overall market environment justify taking more risk right now?

Those answers won’t always agree.

A stock can flash ADD while the overall market says:

HOLD / DON’T CHASE.

That’s useful.

Maybe I initiate a smaller position.

Maybe I wait.

Maybe I add to something I already own but don’t start something new.

The model isn’t supposed to replace judgment.

It’s supposed to make the decision process more disciplined.

How I would use GP Alpha

When you open the App, don’t immediately go hunting through PORTFOLIOS for green ADD signals.

Start with TODAY.

Look at the Market Risk score.

Look at the current action.

Look at what we’re watching.

Then go to PORTFOLIOS.

Find the portfolio or company you’re interested in.

Look at the signal.

Look at the target position.

Look at YTD and cumulative performance.

Then hit VIEW LEVELS AND SCORE if you want to see what’s underneath the recommendation.

I’m trying to answer four questions:

What do I want to own?

At what price do I want to own more?

How large should the position be?

Does the current environment justify taking the risk?

GP Alpha now checks the market four times a day

I’ve also expanded the App to four scheduled checkpoints:

6:00 AM Arizona — PREMARKET

What changed overnight and how I’m approaching the open.

8:00 AM — POST-OPEN

We’ve seen actual trading. Did the market confirm or reject the premarket setup?

10:00 AM — MIDMORNING

Another read after more of the opening volatility has cleared.

1:30 PM — CLOSING

Where things stand heading toward the close and what I’m carrying forward.

Live market information continues updating between those checkpoints.

That’s intentional.

The data should be live. The investment call should have a timestamp.

That lets us see what the actual decision was at a given point rather than rewriting the narrative every five minutes.

The research tells you why. GP Alpha tells you what now.

I’ll continue writing the deeper research explaining why I like a company, industry or investment theme.

GP Alpha serves a different purpose.

The research might explain why I want to own AVGO.

GP Alpha should help answer:

Do I add AVGO today?

How much?

Am I already at my target?

Is the stock extended?

Has the risk environment changed?

Should I be buying anything at all?

More of the companies and portfolios I’ve written about will continue moving into GP Alpha.

I’m not trying to build another stock-picking app.

I’m building a system around the way I actually manage investments after the research is published.

The Substack article explains why I want to own it.

GP Alpha helps answer what to do with it now.

If you’re riding the Fourth Industrial Revolution with me on Substack, download the GP Alpha app and get the most out of the research. Follow the portfolios, track my proprietary ADD/HOLD/TRIM signals, and know when the risk/reward actually favors putting more money to work.

The Photonic CPO portfolio tracks six companies positioned across co-packaged optics, optical networking, lasers, transceivers and the connectivity layer required to scale AI infrastructure:

AVGO, NVDA, MRVL, COHR, LITE and AAOI.

Each holding begins at an equal model weight of approximately 16.7%.

The performance has been exceptional:

• YTD: +96.3%

• Since Janua…

Read more

I call it the Fourth Industrial Revolution portfolio. It is a concentrated, equal-weight portfolio built around the companies I believe can lead the next five years of AI, automation, semiconductors, cybersecurity, cloud infrastructure and precision medicine.

The ten holdings are:

TEM, TSM, TSLA, NVDA, PLTR, AVGO, NET, MRVL, CRWD and SNOW.

Each company beg…

Read more

I built this as a separate risk model because crypto does not trade on the same inputs as equities. The score runs from 0 to 100. Lower is more bullish. Higher means risk is building and you should become more defensive.

The model combines:

• Bitcoin and Ethereum price trends

• Bull-market support bands, the 200-day average, RSI and drawdowns

• Spot ETF inflows a…

Read more

Add it to your phone’s Home Screen for live prices, real-time market risk and clear action levels—so you can quickly see when conditions favor adding risk, staying patient or protecting capital.

The portfolios and companies I write about will be added soon. Each position will include live …

Read more

I’ve written enough about the power side of the AI buildout, so I’m not going to rehash it here.

The question is how I want to invest around it.

Vistra $VSTR has fallen roughly 15% YTD and is trading around $138 despite earnings moving in the other direction. Q2 adjusted EBITDA grew 31% YoY, management is guiding toward $3.9B–$4.7B of FCF before growth th…

Read more

US Treasury already doubled long-duration buybacks from $2 billion to at least $4 billion per operation. Now officials are discussing using the roughly $950 billion Treasury General Account to give themselves considerably more firepower.

Yet long yields still won’t stay down.

Reports say people familiar with Bessent’s thinking believe he’s prepared to “do whatever it takes” to bring yields down.

Then Trump took it another step.

Asked Friday about another form of intervention after yields started rising again, Trump said:

“We have many types of intervention… The ultimate intervention is our military.”

What the hell the military has to do with Treasury yields, I have no idea. But the message from Washington couldn’t be much clearer:

They want long rates lower.

And there’s a reason.

Washington has too much debt, too much interest expense and too much new issuance. Meanwhile, the Fed is boxed in by inflation.

Cutting rates doesn’t necessarily solve it. The Fed controls the short end. The bond market controls the long end. Cut too aggressively and investors can simply demand a higher term premium.

So Treasury is stepping in.

Buy back long bonds. Issue more at the short end. Reduce duration. Use the TGA if necessary.

They can call it liquidity support, or debt management.

But we’re getting closer to YCC.

That’s the progression I’ve been warning about:

Fiscal pressure → intervention → financial repression → eventually monetary policy becomes subordinate to the government’s financing needs.

We’ve spent decades hearing:

Don’t fight the Fed.

The bond market may be about to learn a new one:

Don’t fight the U.S. Treasury.

HYPE and PURR

Hyper-liquid is becoming one of crypto’s most important financial platforms. It dominates decentralized perpetual trading and uses 99% of protocol fees to purchase HYPE from the market.

$HYPE is its native token, and one of the relatively few major crypto tokens trading higher over the past year. $PURR is the Nasdaq-listed company that owns and stakes approximately 27.6 million HYPE, giving stock investors regulated exposure to the ecosystem.

Stanley Druckenmiller’s Duquesne Family Office disclosed 2.94 million $PURR shares. Days later, Trump said the CFTC is working to bring Hyperliquid into the United States legally.

Druckenmiller validates PURR as an institutional investment. Trump validates Hyperliquid’s potential place in the U.S. financial system.

$HYPE is near record highs, but if Hyperliquid receives a workable U.S. pathway, this rerating may be just beginning.

https://desalle.substack.com/p/hype-at-the-center-of-defis-on-chain?r=1d4gvs&utm_medium=ios

If I’m right, the next five to seven years could be a materially more bullish environment for Bitcoin than anything it has experienced in its history.

Look at the chart.

DXY has spent roughly 45 years ca…

Read more

For equities, the bond market is flashing a warning.

The long end is sending a message Washington doesn’t want to hear. The market is demanding substantially more compensation (higher rates) to finance America’s fiscal trajectory.

Kevin Warsh can talk about fighting inflation, but the long end is questioning whether the Fed will ultimately be able to hold the line.

Scott Bessent can continue leaning on T-bills to reduce duration hitting the market, but he can’t hide a structural deficit in the front end forever.

And Trump is pouring gasoline on the problem: enormous fiscal deficits, tariffs that have already pushed prices higher, and the Iran War driving oil back above $90.

This is what fiscal dominance looks like when the bond market starts setting the price.

I am not bearish, but at some point, 5%+ long rates stop being a bond-market story and become an equity-market story, so pay attention, and have some risk management in place.

The problem is the combination of high valuations + rising long-duration yields. The S&P entered this week around 20.2× forward earnings.  At that valuation, a sustained 5%+ Treasury complex starts competing seriously with equities while simultaneously increasing the discount rate applied to long-duration growth stocks.

I would frame the risk roughly this way:

• 30Y 5.2–5.4%: uncomfortable but equities can absorb it if earnings keep ripping.

• 30Y 5.5%+: I’d expect meaningful multiple compression and probably a 5–10% S&P correction.

• 10Y approaches 5% + 30Y approaches 5.5–5.75%: much more dangerous. A 10–15% correction becomes quite plausible.

• Oil stays >$90 while yields continue rising: worst combination because it attacks both inflation expectations and corporate/consumer margins. Brent closed around $90.87 today.

But there’s an important wrinkle for AI infrastructure. Today’s market was almost a perfect demonstration: the broad market fell, yet the semiconductor index rose strongly, with Coherent, Sandisk, Marvell and Applied Materials among the leaders.

So I wouldn’t automatically expect the next selloff to resemble 2022 where all long-duration technology gets indiscriminately crushed. AI infrastructure companies with explosive earnings/cash-flow revisions and genuine scarcity economics can continue outperforming even while the market multiple contracts.

The thing I’d watch most closely now isn’t the S&P.

Watch the 10Y/30Y curve.

If economic data continue weakening while the long end keeps rising, that’s the nasty signal. Reuters reports exactly that divergence today: weak retail sales pushed investors to reduce expectations for near-term Fed tightening, yet the 30-year still surged to 5.31%.

That tells you the bond selloff isn’t simply “the economy is too strong, therefore rates are higher.” The market is increasingly charging the U.S. for duration, fiscal and inflation risk. Goldman is explicitly pointing to rising term premium associated with fiscal/inflation concerns.

That’s the scenario I’d worry about for stocks. The Fed can eventually cut the overnight rate. It can’t simply order the long end down without moving toward some form of balance-sheet intervention/YCC.

And if equities fall 10–15% while long yields remain disorderly, that is where the political pressure for intervention becomes enormous.

For concentrated positions, this is one of the better environments to consider buying insurance, because the insurance is still relatively inexpensive before the equity market fully reacts to what bonds are doing.

The striking divergence today is:

30Y Treasury: 5.31% — highest since 2007

10Y: 4.72%

VIX: only 15

That last number matters. The bond market is flashing substantially more stress than the equity-volatility market. In other words, equity protection hasn’t become expensive yet.

For large individual positions, I would generally prefer defined, modest insurance rather than aggressively reducing the underlying positions if the long-term thesis hasn’t changed. One reasonable structure is 2–4 month puts perhaps 10–15% OTM, sized so you’re insuring against the ugly tail rather than trying to eliminate every 3–5% drawdown. Another is a put spread, buying, say, a ~10% OTM put and selling a ~20–25% OTM put, which dramatically reduces the premium if you’re mainly worried about a conventional 10–20% correction.

I wouldn’t over-hedge your high-conviction AI infrastructure positions. Those stocks can continue appreciating even while the broad market struggles, as today’s semiconductor strength illustrates.  The objective would be to protect against a disorderly repricing where the 30Y goes through 5.5%, the 10Y approaches 5%, and equity vol suddenly catches up.

The asymmetry right now is interesting: VIX ~15 while the 30-year is at a 19-year high. If the bond move finally breaks equity complacency, puts bought while implied volatility is subdued can appreciate from both negative delta and expanding volatility.

That makes this a good time to think about cheap insurance rather than reducing high-conviction positions. Take this $MU position as an example. Let’s say you own 4,000 shares at roughly $1,019, a position worth just over $4 million. I don’t want to hedge normal volatility or sacrifice the upside because the memory/HBM thesis remains intact. I want protection against the tail risk of a disorderly 15–30% correction.

One way to do that would be buying November puts roughly 10–15% out of the money. At a $900 strike, 20–30 contracts would provide partial protection, while 40 contracts would cover the entire 4,000-share position. An even cheaper structure would be a $900/$750 put spread, which sacrifices protection below $750 in exchange for substantially lowering the cost of the hedge.

Forty $900/$750 spreads would provide as much as $600,000 of gross downside protection:

($900 − $750) × 4,000 shares = $600,000

I’m not trying to eliminate every drawdown. I’m willing to eat the first 10–15%. What I want to insure is the scenario where the bond market finally breaks equities and a normal correction turns into something much uglier.

Spend a relatively small amount when volatility is cheap to protect several million dollars of concentrated exposure, while keeping 100% of the upside if the AI infrastructure thesis keeps working.

The Fed capped Treasury yields while inflation, nominal growth and financial repression steadily eroded the real value of the debt. By the late 1970s, debt/GDP had fallen to roughly 30%, giving Volcker the room to crush inflation with double-digit interest rates.

Today we’re starting from the opposite position. Debt/GDP is back near post-WWII levels, deficits remain enormous, and interest expense is consuming an increasing share of federal revenue.

There will be no Volcker 2.0.

The government cannot tolerate Volcker-era real rates on today’s debt load for long. The path of least resistance is financial repression: keep the government’s effective funding cost below nominal GDP growth, tolerate higher inflation, and let time and currency debasement reduce the real burden of the debt.

That changes what you want to own.

Not just gold and Bitcoin. Productive scarce assets.

Power generation. Data centers. Energy. Copper and uranium. Infrastructure. Scarce real estate. And increasingly, the physical infrastructure behind AI: compute, semiconductors, memory, photonics and the electrical systems required to power all of it.

The AI trade isn’t just a technology trade anymore. The winners increasingly control scarce physical capacity that takes enormous amounts of capital, energy and time to reproduce.

Cash and long-duration nominal bonds are ultimately claims on a currency the government has every incentive to debase.

You want to own scarcity, productive assets, and the bottlenecks in an era of fiscal dominance.

Lab-grown diamonds have gone from a niche product to the majority choice for engagement rings in the United States. That shouldn’t be surprising. Consumers can now buy a larger, higher-grade diamond for a fraction of what an equivalent natural stone costs, with no visible difference to the person wearing it.

The price compression has been brutal. A one-carat lab-grown diamond that retailed for roughly $3,400 in 2020 can now be bought for around $600. And there is little reason to believe the decline stops here.

As manufacturing improves and capacity expands, lab-grown diamonds should increasingly be priced like other manufactured goods: production cost, cutting and polishing, certification, distribution and a retail margin.

But the more interesting story isn’t simply that lab-grown diamonds are getting cheaper.

Technology is destroying diamond size as a signal of wealth.

Twenty years ago, an enormous tennis bracelet or a three-carat engagement ring communicated something almost automatically: it was expensive. The size of the diamonds was itself evidence of scarcity.

That relationship is disappearing.

Someone who once had the budget for a one-carat natural diamond can now buy a two-, three- or even larger lab-grown stone. As more consumers do exactly that, large diamonds become commonplace.

And once everyone can wear a huge diamond, size stops communicating exclusivity.

The feedback loop is pretty straightforward:

Lab prices fall → consumers buy larger stones → large diamonds become commonplace → size loses its value as a status signal → consumers become less willing to pay enormous premiums for ordinary natural diamonds.

This is where the greatest damage to the traditional diamond industry will occur.

The natural-diamond market isn’t necessarily going to disappear. Instead, it is likely to split.

At the bottom and middle of the market, lab-grown diamonds have an overwhelming economic advantage. If someone can choose between a one-carat natural diamond and a visually equivalent two- or three-carat lab-grown diamond for the same money, the traditional industry has an increasingly difficult argument to make.

The most vulnerable part of the market is the ordinary commercial natural diamond, the 0.5-to-2-carat stones whose value proposition for decades was essentially that diamonds are scarce, diamonds are expensive, and therefore diamonds signify wealth.

Lab-grown diamonds have broken that equation.

But the very top of the natural-diamond market could behave quite differently.

Truly exceptional natural diamonds, very large stones, rare fancy colors, unusual specimens and stones with important provenance, could become more collectible precisely because everything else has been commoditized.

Mechanical watches provide a useful analogy.

Quartz technology made accurate timekeeping almost free. It didn’t destroy Patek Philippe. It destroyed the economic rationale for an ordinary mechanical watch as a necessary timekeeping device.

What survived moved toward craftsmanship, history, rarity, provenance and collecting.

Diamonds could follow the same path.

The market may eventually divide into three distinct categories: lab-grown diamonds as mass-market jewelry; ordinary natural diamonds as a shrinking middle caught between cheap lab stones and genuinely rare natural stones; and exceptional natural diamonds as collectible luxury objects valued for scarcity and provenance rather than simply carat weight.

That’s also why I don’t think natural diamonds necessarily settle at some fixed multiple, 2×, 5× or 10×, of lab-grown prices.

Eventually they may be treated as fundamentally different goods even though the stones themselves are chemically and visually nearly identical.

A $30 quartz watch and a $100,000 mechanical watch both tell time. Nobody determines the value of the Patek by multiplying the price of the quartz watch.

The same separation could happen with diamonds.

And there is another consequence the diamond industry may be underestimating.

As five-carat lab-grown stones and enormous tennis bracelets become affordable, ostentatious size itself may cease to look luxurious.

Luxury signals work because they are difficult to obtain. Once technology democratizes the signal, wealthy consumers move elsewhere.

Subtlety, provenance, unusual natural stones, craftsmanship and genuine rarity may become more important than sheer carat weight.

So the endgame isn’t necessarily that diamonds disappear or that every natural diamond becomes worthless.

It is something more fundamental:

Technology is separating the diamond from the scarcity.

Lab-grown diamonds give consumers the physical object, the brilliance, hardness and visual impact, at increasingly commodity-like prices.

Natural diamonds will have to sell what technology cannot manufacture:

Rarity. History. Provenance.

There can still be an extraordinary luxury business built around those things.

But there probably isn’t enough demand for that story to support the enormous mass-market natural-diamond industry that was built over the last century.

So What Should You Buy?

None of this means I would never buy a natural diamond for my wife.

It means I would be much more selective about which natural diamonds deserve a scarcity premium.

If you’re buying primarily for appearance, an engagement ring, earrings, a necklace or a tennis bracelet, lab-grown increasingly makes economic sense. You’re getting essentially the same physical object without paying an enormous premium for geological origin.

But if you’re spending serious money on a natural diamond, I would go in exactly the opposite direction.

Don’t buy ordinary. Buy genuinely scarce.

Look for exceptional natural stones: rare fancy colors, particularly strong pinks and blues; unusually large, high-quality stones; important provenance; exceptional vintage pieces; and signed jewelry from major houses such as Cartier, Van Cleef & Arpels or Harry Winston.

The point isn’t simply to buy a better diamond.

Buy something whose value extends beyond the fact that it is a diamond.

A generic one-carat natural diamond has a problem: technology can manufacture an effectively identical-looking substitute by the millions.

A significant vintage Cartier piece cannot be reproduced in the same sense. Neither can an important Van Cleef & Arpels jewel, a historically significant Harry Winston piece or an exceptional natural pink or blue diamond whose geological rarity is itself part of its appeal.

This distinction will become increasingly important as lab-grown prices continue falling.

The diamond market may ultimately follow the same path as mechanical watches. Once technology makes the basic product cheap and abundant, value migrates toward rarity, craftsmanship, history, provenance and collectibility.

So if I were buying diamonds today, my rule would be simple:

Don’t pay a scarcity premium for something that isn’t actually scarce.

Buy lab-grown for the diamond.

Buy natural for the rarity.

And if you’re paying serious money for natural, buy something technology can’t commoditize.

SpaceX is generating an estimated $30–50 million in annualized revenue for every active megawatt of compute capacity.

Pure-play neoclouds like CoreWeave, Nebius and IREN sit around $9.4–10.4 million per MW, while traditional colocation players like Digital Realty and Eq…

Read more

Anthony Fauci became the face of America’s pandemic response, but the real story was never about Anthony Fauci.

It was about the system that gave one unelected bureaucrat, and thousands of others like him…the power to shape the lives of hundreds of millions of Americans.

If every ounce of public anger ends with Fauci, then the system has done exactly what it was designed to do.

Bureaucrats Hold the Power

Americans are taught that Congress writes the laws and presidents run the government.

In reality, much of Washington is run by a permanent bureaucracy that remains in place regardless of who wins an election. Administrations change. Bureaucracies don’t.

Career officials write regulations, interpret laws, distribute funding, and influence policy long after elected leaders have left office.

The Incentive Is More Power

Every institution responds to incentives.

For the federal bureaucracy, success often means larger budgets, broader authority, more employees, and greater influence.

Money isn’t the only motivation. Prestige, influence, and institutional power are rewards in themselves. Once authority is granted, there is rarely an incentive to voluntarily surrender it.

Government programs have a tendency to expand because the people running them benefit from that expansion.

“Science” Becomes a Political Shield

Science is one of humanity’s greatest tools for discovering truth.

But science can also be invoked to give political decisions an aura of inevitability.

During the pandemic, Americans repeatedly heard that policies were simply “following the science.” In reality, public policy always involves trade-offs between competing values: health, liberty, education, economic opportunity, and individual rights.

Those are political judgments, not “science”…

Invoking science can inform policy. It should never end debate.

The System Needs Scapegoats

Large institutions survive by protecting themselves.

When public confidence collapses, people naturally look for someone to blame. A single face is easier to understand than an entire bureaucracy.

Anthony Fauci became that face.

Whether he is remembered as a hero or a villain is almost beside the point.

If all the blame is aimed solely at Fauci and not the system he was operating within, then, regardless of whether he escapes any accountability or spends the rest of his life behind bars, he will have fulfilled his role.

The Machine Remains

Fauci pleaded the Fifth, and is now retired.

The bureaucracy he represented is not.

The agencies remain. And so has the regulations, and incentives. The authority accumulated during the pandemic still exists, waiting for the next crisis.

Holding Fauci and other officials accountable for their decisions is necessary if the country hopes to rebuild public trust.

But accountability alone is not enough.

The larger question is whether Americans are willing to reconsider the laws, regulations, and institutional incentives that allow unelected bureaucracies to accumulate ever greater power.

Otherwise, the names will change, but the system won’t.

1. Demographic Inversion

The global population pyramid is inverting. Every major asset class and financial system was built on expanding populations, growing workforces, and rising consumption. That foundation is now reversing.

2. Capital Concentration

As wealth becomes increasingly concentrated while public debt continues to expand, political pressure for redistribution will intensify. Whether through wealth taxes, higher taxes on capital, inflation, or financial repression, the relationship between capital and the state is likely to be rewritten.

3. Artificial Intelligence

AI is redefining the economics of work by compressing the value of much of labor while increasing the value of owning capital, compute, energy, data, intellectual property, and networks. The intent-based economy will reward ownership more than employment.

These three structural forces are not unfolding independently, they are reinforcing one another. Their convergence is reshaping demographics, economics, politics, and markets simultaneously, creating the conditions for a period of institutional and societal transformation.

A $5–10 billion intervention is small relative to the roughly $7–8 trillion traded daily in the global FX market. By itself, that amount is unlikely to permanently change USD/JPY.

However, the signal matters more than the cash. Coordinated intervention by the U.S. Treasury, the Bank of Japan, and potentially other G7 central banks can influence market expectations, especially if traders believe it signals future policy changes.

A weaker dollar improves U.S. export competitiveness and make Japanese imports more expensive. However, trade volumes typically adjust over months or years, so the near-term GDP impact is likely modest.

A weaker dollar increases import prices, but a single FX intervention of this size would probably have only a very small effect on overall U.S. inflation unless it marked the start of a sustained dollar depreciation.

The pertinent question is whether this is:

1. A tactical intervention to reduce excessive volatility in USD/JPY, or

2. The opening move in a broader policy to weaken the dollar.

If it’s the second, the implications become much larger. A sustained weaker-dollar policy could:

• Improve the U.S. trade balance over time.

• Ease pressure on countries and companies with dollar-denominated debt.

• Support commodity prices (including gold).

• Increase the dollar value of foreign earnings for U.S. multinationals.

• Potentially contribute to somewhat higher U.S. inflation over the medium term.

If the U.S. is signaling a structurally weaker dollar rather than just smoothing volatility in USD/JPY, the second-order effects could be much larger than the intervention itself, several asset classes could benefit simultaneously:

• Gold and silver → higher dollar prices and increased reserve diversification.

• International stocks → gains from both local equity performance and currency appreciation.

• U.S. multinationals → stronger translated overseas earnings and potentially improved export competitiveness.

• Emerging markets → lower debt-servicing burdens, improved liquidity, and stronger capital inflows.

This is one reason to pay close attention to any signs of a shift in U.S. currency policy. The intervention itself may be modest, but if markets conclude it signals a sustained change in the dollar’s trajectory, the ripple effects can extend well beyond the foreign exchange market.

My base case this is an 80% tactical intervention aimed at stabilizing USD/JPY and supporting Japan, and a 20% chance it’s the opening move toward a more managed, gradually weaker dollar.

That 20% isn’t a prediction, it’s recognition that markets price trajectories, not just events. If future policy actions begin matching today’s rhetoric, investors will increasingly price in a structural shift.

The next 3–6 months will be far more revealing than this single intervention. If the Treasury follows with additional FX operations, coordinated G7 messaging, or a consistent policy narrative favoring a weaker dollar, the odds that this represents a genuine regime change would increase meaningfully.

The fund was forced to liquidate its entire public book after severe losses the past couple of weeks.

Situation Awareness was 4x leveraged.

When a fund this large is forced to deleverage and liquidate, the selling becomes non-economic. Both winners and losers get sold. The most liquid high beta names sell off the hardest.

That’s the exact scenario why we’ve seen indiscriminate selling of quality tickers like $NBIS, $MU, AI infrastructure, and semis.

Is Situation Awareness the last large forced sellers or are there other fund blowups out there still deleveraging?

Ironically, forced liquidation events often occurred near important lows.

Markets bottom are marked by margin calls, hedge funds blowing up, indiscriminate selling rather than fundamental selling, capitulation sentiments. All of which checked the box the last couple of weeks.

Incremental supply disappears once the force sellers are gone. This removes one of the biggest downward pressure. The AI demand story is still intact with hyperscalers still aggressively spending on AI infrastructure. Prices will stabilize as natural buyers step back in.

This is more about market structure (degenerate leverage, see KOSPI) than deteriorating fundamentals.

Forced liquidation often creates temporary price disconnection from business performance and fundamentals. The next several trading sessions should tell us if selling has exhausted that supply or if further deleveraging remains.

Personally, my PUTs on my large positions saved my portfolio. And I’ve been deploying cash to add to my position on quality names like $NBIS $MRVL $MU….

Aging is simply the increase in that force over time.

In humans, the force of mortality rises roughly exponentially with age. That’s why even a modest reduction in midlife mortality can translate into a surprisingly large increase in life expectancy.

Some animals, such as turtles and crocodilians, appear to exhibit negligible senescence. Their force of mortality remains relatively flat. They still die,

from accidents, disease, predation, or starvation, but not because age itself makes them dramatically more likely to die each passing year. Their remaining life expectancy stays remarkably constant.

That’s essentially the Lindy principle. If the hazard rate remains constant, surviving another year doesn’t make the following year appreciably more dangerous.

This is where much of the modern longevity movement goes off the rails.

Bryan Johnson, for example, seems to equate slowing biological aging with approaching immortality. But immortality requires something far more demanding than slowing senescence.

It requires reducing the force of mortality to zero from every cause, heart disease, cancer, infections, neurodegeneration, accidents, natural disasters, and simple bad luck, for an infinite amount of time.

That isn’t biology.

It’s mathematics.

As long as any non-zero hazard remains, the probability of surviving forever is effectively zero.

Ironically, swallowing 100 pills a day may lower one source of mortality while increasing another through drug interactions, side effects, unknown long-term consequences, or simple human error.

The objective isn’t to minimize biological aging.

The objective is to minimize total mortality.

Think of longevity as a risk management problem…not a contest to produce the youngest biomarker panel.

The force of mortality is the sum of many competing hazards:

• Cardiovascular disease

• Cancer

• Metabolic disease

• Neurodegeneration

• Infectious disease

• Accidents

• Medical interventions themselves

The strategy that follows is surprisingly unglamorous.

Attack the largest proven risks first.

Maintain healthy blood pressure, ApoB, glucose, body composition, cardiorespiratory fitness, and don’t smoke.

Exercise consistently. Build muscle. Preserve balance. Stay metabolically healthy.

These interventions reduce mortality far more reliably than most expensive longevity protocols.

Detect disease early, but intelligently.

Screen for diseases where early detection clearly improves outcomes.

Avoid turning every scan into a fishing expedition. False positives lead to unnecessary biopsies, procedures, radiation exposure, anxiety, and sometimes treatments for diseases that would never have become clinically relevant.

Don’t die from preventable stupidity.

Wear your seat belt.

Don’t text while driving.

Don’t drive sleep deprived.

Prevent falls.

Keep smoke detectors working.

Respect electricity, firearms, motorcycles, ladders, and alcohol.

At a certain level of health, preventing one fatal accident may add more life than shaving another few percent off an inflammatory biomarker.

Minimize iatrogenic risk.

Every drug, supplement, peptide, or procedure should answer one simple question:

Does it reduce my total force of mortality more than it increases it?

More interventions do not automatically produce better outcomes.

Complexity itself becomes a source of risk.

The fewest interventions with the greatest proven benefit usually outperform elaborate experimental stacks.

Preserve resilience.

Muscle.

Aerobic capacity.

Bone density.

Sleep.

Balance.

Cognitive function.

Social connection.

These aren’t simply markers of health, they’re reserve capacity. They determine whether you survive infections, surgery, trauma, hospitalization, and the inevitable physiological insults that accompany aging.

My operating principle is simple.

1. Large, proven interventions with low downside.

2. Personalized medicine based on measured risk.

3. Experimental therapies only when the expected benefit clearly exceeds the uncertainty.

4. Avoid accumulating interventions simply because they’re available.

Longevity isn’t about achieving the lowest biological age score.

It’s about maximizing the probability of remaining alive, functional, and independent without creating new ways to die in the process.

A decisive breakout on strong volume would likely signal the beginning of the next leg higher…not just for gold, but silver, and across precious metals.

Consolidations like this often serve to reset sentiment before the trend resumes. The next few sessions should tell us whether that’s what’s happening here.