https://stochvoltrader.substack.com/p/distribution-breakdown-logopt-trade

Subscribe now

Let’s look at today’s positioning…

Read more

https://stochvoltrader.substack.com/p/distribution-breakdown-logopt-trade

Subscribe now

Let’s look at today’s positioning…

Read more

ATTENTION:

As I mentioned last week, I won’t be around this week.
You’ll still receive the daily script inputs every day, though, along with the probability distributions and the heatmap. I was able to automate those.
I might write a short note on NFP and PMI that day.
The next weekly post will come out on September 7, since the market is closed.

This week I put out several posts analyzing Warsh, gold and yields, looking at market positioning through year-end and the probabilities that follow from it, as well as oil and geopolitical developments.
Please read them.

These are free posts, so you can share them too. People could stand to read this.

Subscribe now

Now I’ll look at this week’s customer positioning and walk you through the probability distribution the market is pricing in.

Read more

The market is pricing today’s Jackson Hole as a normal day (0DTE sold, dealers +$3.7B/1% long gamma, no event premium in either 0DTE or the front end). The 0DTE straddle is pricing in ±0.6%.

Although the market is pricing in by year-end that Warsh is hawkish, it expects this effect to be absorbed. There is an $88B synthetic long funded at SOFR+84, the spread compresses from ~105bps (Sep to Oct) to ~60bps (Oct to Dec), so positioning sees the Q4 hike priced in and pushes leverage into the Sep16 FOMC/quarter-end window. My rhogamma model, which uncovers bets on the SPX/yields beta, shows that this is not a hedging risk in the book. Any unexpected rate shock (whether hawkish or dovish) would cause a maximum swing of ±$40-50 million. Compared to multi-billion dollar options moves, this is a mere rounding error.
The situation is different regarding vanna positioning for Q4 (post-midterms). The dealer book’s break-even spot/vol beta is -0.58 pts/1%; the Sept smile (-0.71) is below it, the Dec smile (-0.42) is above it, meaning the term structure is pricing a transition from a stabilizing regime to a trend-following one for Q4. This coincides with the ATM rising from 11.2% to 14.0% and the VIX curve’s contango from 16.85 to 19.5.

Customers sold skew correlation on the SPX (long vanna +$668M/point, every vol point in a 5% drop costs them $160M), and bought crash convexity in the VIX. The bet-measure overweights the crash bucket by 3–4x (10.6% vs a 3.6% q-probability for Sep30), but as a VIX event, not an SPX skew event. Dealer flow aligns with this: from -2% to -5% it’s ‘vol up’ (−$0.4B...−$0.9B), but at −10%, vol absorption is already priced in (dealers would have to buy like 6.2B against the momentum). To the upside, there is a big call wall above 8016.
So the probability-weighted expected flow for Sep30 is +$1.3B net buying. Therefore, the hedge is for a -3% to -8% correction and a VIX spike, while the deep SPX tail is sold.

The market will start feeling the changes from the end of Q3, or at least that’s what it is signaling.

Subscribe now

The probability distribution looks like this:
Up to Sept 18, the market shows a tight range (1-sigma ≈ ±2.6%, 68% band 7535–7940), the crash bucket is <= -8% and only 1.9%, while the tail above 8000 is 8%. Along the horizon, the right tail grows faster than the left, while the forward rises from 7741 to 7808, and the median for December is 7898. This is the combined effect of the market-priced drift and the rising vol for Q4.

So, even though market pricing sees Warsh priced into the forward and absorbed, and times the risk for Q4—when the left tail price rises to 13% and vol to 14%—CBOE positioning data shows that hedgers disagree with this timing. They are buying the near-term tail because it’s cheap, viewing Sep30 as the day the Warsh reaction will materialize. The combos, however, overweight the December upside, funded at SOFR+84.
So, the distribution does not show a consensus, but rather a gross mispricing. Prices reflect a grinding higher and Q4 tail risk; however, customer positioning data prices in near-term tail risk, and leverage prices in a December melt-up.

...However, there is a risk that the market is too complacent about, and this is exactly why Bessent’s program happened.
Since Warsh wants to keep long-end yields high for a synthetic QT, this would result in a slow, mechanical SPX bleed [or grind lower] where the VIX doesn’t spike. As a result, the downside vol-absorbing positioning dynamic is not triggered, leaving the trapdoor open roughly below the June lows.
This would currently be the only true bear scenario in the market.

Today at 10:00 ET, it’s not the word ‘hike’ that matters, but the reaction of the 30Y. If Bessent’s long-end program is credible, we get a relief grind; anything else triggers a tantrum. The market believes he will be able to hold the front-end, but not the long-end. And if the long-end breaks in a disorderly fashion, a VIX event is coming around the midterms. However, 0DTE pricing shows that the market doesn’t expect this to be decided today, but rather at the Q3/Q4 transition.

Update

Warsh’s speech is out.
Warsh laid out a standard, not a decision, but the shape of the standard is asymmetrical: inaction requires certainty ("confident... clearly and at sufficient speed"), while action is the default ("otherwise, we have work to do"). To this, he added the antecedent—the trend has not improved, the good summer prints do not count, breadth is 54%—and he dismantled the two constraints: financial conditions are not restrictive even with the 30Y at 5.3%, the labor market is at full employment, and the dual mandate is "not working at cross-purposes." He also named the tool: the short rate is primary, unconventional policy "sparingly, if at all." By modus ponens, tightening follows from this; the text only withholds the timing, and he even justifies why with the Romer–Romer footnote: guidance slows the reaction, the absence of guidance accelerates it. The Kohn/Bernanke wordplay says the same thing playfully: there are two kinds of hikes, the excluded middle applies to the pace, not the direction.

By my calculations, a system without guidance is only consistent if it is not contrarian. The Fed communicates the axioms, and the market must derive the theorem; otherwise, the decision is a surprise, and a surprise is a breach of credibility.
The speech, therefore, turned the hike from a risk into the baseline, the hold from the baseline into a risk, and the long end from an instrument into information. Warsh stated what I have been writing for a while now: inflation is not falling fast enough (also noting that PCE is 3.7% annualized, 4.1% annualized over six months; and that the good summer data doesn't mean the trend has improved), and he also stated that the tool is the short-end.
More can be said after the September CPI. Tightening logically follows from these three—he just didn't state the timing. The market has priced this in: the probability of a September hike jumped to 66%, and they are already pricing in 1.4 hikes by December. So from now on, a hike is the expected outcome.
However, if a hike is the expected outcome, then a non-hike is the surprise. And according to Warsh's own rule, a non-hike can only be justified by one thing: certainty about the downtrend. A good CPI print on September 11 is not yet a trend. If he doesn't hike on September 16, and there is no trend improvement in the meantime, the market will be forced to conclude that he is not bound by his stated rule. This is a severe credibility issue, and credibility is priced by the long end of the bond market. If the 30-year yield were to rise (pricing in the missing tightening instead), equities would drop to roughly around 7500.
So, before the speech, this drop was associated with a hawkish surprise; after the speech, it is associated with a hold.

Furthermore, he ruled out supporting the long end, meaning long-end yields will remain high, exactly as I forecasted in my previous Warsh analysis as a synthetic-QT strategy. But this will become a trend later; for now, the front-end (by pricing in the September hike) has credibly done the tightening, so the long end didn't have to.
The market also calmed down because it received the hawkish content in a credible, surprise-free format. (Today moved ±0.4%, they bet on ±0.6%. As I signaled pre-market, put selling was the winning strategy.)

There are three outcomes:

1. Warsh hikes in September (or justifiably holds because the data actually improved), long-end yields remain calm, and equities grind higher to the 7850-8016 range. The priced probability is 63%;

2. A September hold without data improvement. Here the long end becomes unhinged, going up +50bps to 5.65%+, which would mean a -7% SPX drop, retesting the June lows. The key here is that bond and equity vol spike simultaneously. Like in 2022, when equities fell not on a growth shock, but on a duration shock, and vol did not cushion it. Importantly, Warsh did not promise a Fed put if the long end breaks. On the other hand, the risk is an oil/Hormuz shock that moves inflation expectations. It's important to note, however, that the market is not trading this with deep OTM SPX puts (they are shorting those), but with VIX calls. Why? Because the Sep 20–25 calls for zero premium (BE 16.8 = the futures) are +$395M at 30, +$882M at 45; the Nov 60 calls (159k, for the midterm+FOMC cluster) for $26M are +$662M at VIX 80. In this scenario, the customer VIX book pays +$441M, the SPX options book +$2.6B, the combos −$6.1B: the aggregate is −$3.1B, meaning the hedge covers one-third of the leverage losses. Below 7300, SPX spot gets a zomma cushion; the VIX does not. Bear trap. The probability of this is roughly 7%;

3. This is the only true bearish scenario. Either he doesn't hike in September without data improvement (this is a credibility issue leading to long end up/SPX down), or he hikes but signals an aggressive pace (50bps, or "this is the first of several," per the Kohn reference in the text). This is a yield-driven bleed where downside support is not triggered. The probability of this is roughly 30%.

Commodities

I looked at gold, copper and long-end yields, because they are sending important signals.
Gold had been behaving as a rates-reactor: when bond yields rose, gold fell. In August, though, the market started to worry that the Fed was losing control, and gold shot higher as a stagflation hedge. Today gold dropped 3.4% while long yields barely moved. That means the gold market priced in that the Fed still has the situation in hand. The 60-day correlation is −0.13, the beta −0.063%/bp — today’s −3.4% against a +2bp 30Y is far above that beta, so this is not a yield-driven drop. It is a credibility-driven drop.

I re-ran copper/gold versus the 10Y too…
The ratio is now 0.0971, the midpoint of the two-year range (49th percentile). That is +10% over one year, but not because gold fell: copper rose +41% while gold rose +27%. The two-year regression is weak at only +0.19 (distorted by the 2024 yield-decline window); the one-year is strong at +0.79, slope 20% per unit. The 10Y implied by the ratio is 4.48% against an actual 4.72% — so the 10Y is 24bp “rich” to what the growth proxy justifies (+42bp on the two-year fit). This residual turned persistently positive in February 2026 — at the same time as the gold crash — and has sat between 0 and +30bp since. That is the credibility / term-premium surplus on the curve.
Today the ratio jumped +2.6% (gold −3.4%, copper −0.8%), which per the regression implies +5bp on the 10Y — and the actual 10Y moved exactly +5bp. So today’s yield rise is “fundamental” according to the ratio, not term premium: the residual did not widen. That is the imprint of the 63% scenario I laid out in an earlier post.

Copper is expensive not because of demand but because of a supply shock. The world’s largest mines — Grasberg, Kamoa-Kakula, Cobre Panama, Codelco’s El Teniente — have seen accidents, shutdowns and capacity cuts. Sulfuric acid, essential to copper processing, is in short supply where it matters most, after China halted exports. Smelters are already processing concentrate for free (treatment charges are at zero), which is the cleanest signal of shortage. Inventories have piled up in the US on tariff bets, and are tight in the rest of the world; that US stock will mechanically press copper lower once it is released onto the market.
Right now the actual 10-year (4.72%) is about 24bp higher than what copper/gold would justify (4.48%).
Also note, that the ratio uses COMEX copper, which carries a US tariff premium — so the 10Y is, if anything, richer than the 24bp suggests, and a ratio drop driven by that premium unwinding is a tariff artifact, not a growth signal.

• If gold falls (credibility), the ratio and the implied yield rise, the actual yield stays with them, and the +24bp surplus closes or stays stable — that is the 63% priced branch: bear-flattening with an anchored long end, 10Y held in 4.5–4.7.

• If gold rallies (hedge branch, credibility break), the ratio and the implied yield fall while the actual yield rises on term premium, and the residual widens above +40–50bp. That is the imprint of the more bearish scenario: 30Y at 5.4–5.5.

Today showed the first path: residual unchanged, gold −3.4%, 10Y +5.

At the same time, remember that the copper supply shock is Warsh’s own input (“the recent rise in overall commodity prices bears watching”), so copper strength is hawkish through the reaction function. A rise in the ratio — whether via copper or via gold — lifts the floor under long yields (~4.5% on the one-year fit) and leaves the back end of the curve no dovish room before 2027.
This is consistent with everything I wrote today and earlier. The commodity leg of the stagflation hedge (gold) is in the Fed’s crosshairs; the copper leg is holding up for supply reasons.

So the market is hedging against his loss of credibility. In all other cases, the market is supported. Geopolitical risk remains underpriced.

Share

Subscribe now

Let’s look at today’s positioning…

Read more

Subscribe now

Let’s look at today’s positioning…

Read more

PCE is out.
The print is last week’s “melt-up under stagflationary cover” framework in a single number.
Real consumption is stagnating while core inflation stands at 3.3% — except that of this 3.3, roughly 0.4–0.5pp is pure measurement feedback, the imprint of the equity rally and the AI investment frenzy in the price index. Of the 3.7% headline, another 0.8pp is carried by the energy carry from the Hormuz shock.
The trimmed mean, which measures the broad base of price increases after stripping out the swinging components, meanwhile sits at just 2.2%, only a little above target.
The Fed, however, is not reacting to this but to the headline number. At the July meeting three dissenters — Hammack, Kashkari and Logan — already voted for a rate hike. The market is pricing roughly a 40% probability of a hike for September, and Friday morning Kevin Warsh, as Chair, steps onto the Jackson Hole podium for the first time.
Yet it is precisely Warsh who is leading the review of the Fed’s inflation gauges and who has openly criticized PCE. If he remains true to his own logic, the breakdown above — the 2.2% trimmed mean and the asset-price artifact sitting inside core inflation — is nothing other than the door left open for a dovish turn.

Yet core CPI at 2.5% y/y against core PCE at 3.3% is a +0.8pp PCE excess, while the historical norm is about −0.2/−0.3 (CPI usually runs above PCE because of the shelter weight). Asset-price components (+0.4–0.5), healthcare weight/scope difference (+0.2–0.3), imputations/NPISH (+0.1–0.2), minus the shelter disinflation that is pulling CPI (shelter +0.1 m/m). That is a ~1.1pp swing.
So household out-of-pocket inflation is 2.5%, the national-accounts construct is 3.3% — and the gap is not demand, but a bubble artifact plus administered prices.

Compared with this, the options market’s verdict is merely a 0.97% SD — a modest event premium for a day that carries PCE + hike-repricing risk. The 2% left tail is the price of the “if core gets rounded to 0.3%, then September hike repricing” scenario, and the mode sits above the forward.
The market has therefore treated PCE as a formality now — as confirmation of the disinflation path — and has deliberately allocated the real risk budget to NVDA, which it pushed behind the settlement fence: $91B ±2% guide, $92.07B consensus, $2.09 EPS on a $5.1T name. The 7790–7815 short-call cluster is, in effect, the sale of the pre-NVDA euphoria cap. (Plus Jackson Hole over the weekend…)

Subscribe now

Let’s look at today’s positioning…

Read more

This video of an American Boeing C-17A Globemaster III landing in Moscow has been circulating around the internet today.
People are going back and forth about it. Let me explain what’s actually happening, after carefully looking into it…

This is CIA Director John Ratcliffe and his delegation.
They arrived from Washington to Riga on a comfortable VIP aircraft (C-40B), but there they transferred onto a C-17 heavy cargo plane. The C-17 is equipped with built-in missile approach warning and countermeasure systems and was capable of transporting the delegation’s armored convoy and protected communications systems directly to Moscow.
(The United States requested in advance that Ukraine suspend deep drone strikes against the Russian capital for the duration of the multi-day Moscow stay.)
Since the Russian strategy is to play for time, the process itself is the product.

Every diplomatic round yields him three things for free:
1) a sanctions deterrent, because while the “negotiations are alive,” Congress holds back the secondary sanctions package and Trump does not open the interceptor spigot;
2) and this is the most important mechanism☝️ Washington asked Kyiv to suspend strikes until the delegation leaves Russia. So every diplomatic event becomes a free mini-ceasefire for Russian refineries and logistics. Ukraine pays the price; Washington hands it over. The only thing that actually hurts Putin right now is that the deep campaign running from the Baltics to Western Siberia is switched off for the duration of the diplomacy. Putin can harvest these windows one after another;
3) It’s a status trophy. Hosting a CIA director costs zero commitment and yields a threefold return. Of course the plane was allowed to land…

Geo context

Washington faces a binary policy fork this autumn: 1) escalate (sanctions package, interceptor sales, ISR intensification) or 2) pressure Kyiv into concessions. And the decision hinges solely on whether Putin’s runway is real or a bluff. (spoiler: bluff) How long can he hold out fiscally, socially, and in terms of munitions? (spoiler: a long time, because it is an autocracy and interests are aligned with China). This is not a diplomatic question; it is an intelligence-assessment question. My June analysis already said the same thing when I wrote about a “time-limited runway,” a “tapering Iranian windfall,” “social-crisis warnings,” etc. But the variance of the estimate is enormous, and everything turns on it.
Ratcliffe went in order to obtain the direct read before the decision, and also to deliver the Burns mirror — that is, the price list and the warning (“we know what you are preparing for this winter and this is what it will cost you”). You do not negotiate with a time-buyer; you price the time for him. And that has to be delivered personally to be credible. Meanwhile Putin can gauge the quality of the intelligence and the readiness of the diplomacy.
By having the CIA director go in person, Putin gains a clear advantage, because the intelligence channel (as opposed to political envoys) gives maximum flexibility. There is no territorial map on the table, no official framework — this is the ideal terrain for the Russian time-buying strategy.

If everything hinges on the U.S. assessment of Russian sustainability, then the most valuable target for Putin is the assessment machinery itself. A Naryshkin–Ratcliffe (or higher) meeting is the highest-value injection point for perception management of the messages “we can hold indefinitely; your crisis will come first — look at your own bond market.” That is precisely the narrative he wants to feed directly into Langley, the narrative on which his entire strategy rests. Ratcliffe went to observe; Putin wants to write the script. The meeting is a mutual audit of two decay curves under maximum incentive to lie — which is why it can only happen on the intel channel, because that is the audit channel.

And something important to keep in mind: against a post-truth regime, a political promise has no binding force, so the political channel is structurally empty. What remains is the channel that has sensors, not words.
Another point: if they had wanted to keep it secret, it would not have leaked already today. So why is it public? Because the whole of life is theater.

On the one hand this allows Trump to show that diplomacy is in motion, to lull Congress’s sanctions legislation, and to stage a show of force toward the EU (see the Canadian case right now… still the same record I explained last year. Who exactly is the king of Canada?).
For Putin it is a dominance role, because the mere landing of the CIA director in Moscow is itself a prestige trophy that can be communicated at home as the restoration of superpower status (“the Americans come to them”). The Kremlin — through Peskov — functionally did not confirm the preparations, thereby locking the Americans into the role of “supplicant.” See TASS shortly… In essence, therefore, the Americans paid the cost of publicity; Putin collects the yield of publicity.

The process-theater will be spun at maximum until the midterms — more rounds, more leaks, more “near-breakthrough” headlines — because the process is a product for both principals and a cost only for Kyiv. Putin wants nothing from Trump. His calculation is that the Western political (electoral/budgetary) decay curve will break earlier than the exhaustion of Russian economic reserves, and the approaching oil shock plus the social crisis give good odds of that. And I note: according to Western calculations as well…
So even if an air moratorium is reached in the autumn, it will be Putin’s instrument, not a step toward peace — air-defense replenishment, refinery protection, an escalatory option re-activated for winter, exactly as Zelenskyy has also warned.

The Russians are only testing Western cohesion, diplomacy, intelligence, etc. Every air moratorium is rearmament time. It is the threshold test required for the transition from hybrid warfare into conventional warfare; the precondition of that threshold is reconstitution, and the precondition of reconstitution is silence. The ceasefire is not the end of the war; it is the first phase of the Western campaign. The Ukrainians understand this precisely — which is why Kyiv will be the hardest signatory of any moratorium module.
The Eastern powers are waiting for time to run out, for the typical post-midterm Western “shitshow” to arrive, and for the crisis that the West has been accumulating since 2018 to hit. That is when they will move. China on the first island chain; Putin somewhere in the Balkans or around Kaliningrad.
Never forget that time passes completely differently for an autocratic regime than for a Western showcase democracy in which only four-year cycles of public fooling take place…

Share

What to watch?

Watch the diplomatic signals:

1. Watch how long the drone strikes against the Russian rear remain suspended after the CIA visit. If the Ukrainians restart deep strikes immediately after the delegation departs, it means the pause served purely logistical and VIP-protection purposes — i.e., no deal was struck. Otherwise Washington has successfully forced a deal on Ukraine in the background. This remains true even if Ukraine stops bombing Russian infrastructure and hits only military targets, because energy stability is now of elevated importance to America until the midterms. The deep-strike campaign is currently Kyiv’s only genuinely painful offensive instrument against Russia. If they are willing to switch it off, it means they are vulnerable to the extreme. They will only go along if Washington has promised something vital in return (e.g., immediate Patriot shipments for winter survival) or if Washington has blackmailed them (e.g., by threatening to withdraw the American ISR — reconnaissance and satellite — support that is indispensable for Ukrainian drone targeting).

2. If the Kremlin, citing the need for “thorough preparation,” only agrees to another Trump–Putin summit after November, that confirms the wait-and-see strategy.
   If the legislature suspends Russian sanctions for the duration of the “diplomatic process,” the Russian time-buying is working very well indeed.

3. If an energy moratorium is established (on the 2025 model), watch the asymmetry: if Russian refineries (Russian interest) are exempted from strikes while the Ukrainian energy grid continues to be attacked (Ukrainian pain), that signals the balance of power and the American pressure on Kyiv.

4. If Washington sells Ukraine the vital air-defense missiles, it signals that it has recognized the Kremlin’s time-buying tactic. Withholding them means forcing Kyiv toward surrender.

5. If Russian strategic strikes thin out even without a moratorium, that signals stockpile replenishment for a devastating winter offensive (this is a “bearish,” warning signal). This is especially important!

6. If Kash Patel’s visit materializes, the normalization process is alive. If it does not, the channel has frozen.

7. And if after 3 November Moscow’s tone immediately hardens and the negotiation process slows, the thesis will have been fully confirmed: the theater lasted until the onset of winter.

On the market, any “peace-deal” headline is merely sentiment-control for the algorithms that makes the left tail cheaper. The crisis will start from nowhere.
Because of the winter re-escalation and the possible oil-shock scenario, it is worth accumulating the Q1 2027 tail options at discounted prices under the euphoria caused by the moratorium.

Subscribe now

Dessert

And just as a curiosity at the end, a little Eastern propaganda.

This is how it works. They know where society is divided, and what the Western person is missing that the Eastern one receives. That’s what they play on: division, the breaking of cohesion, discontent and unrest. Don’t forget.
Read and share my geo posts, because there is great darkness in people’s heads about this.

Share

Let’s look at today’s positioning…

Read more

Let’s look at today’s positioning…

Read more

Unfortunately I most likely won’t be around next week.
On the 30th you’ll get the weekly post, then the next one on Labor Day, but between the two I’ll most likely not be able to be here.
But I will still update this.

First of all, in this new article of mine I explained the current Bessent measure, Warsh’s plan, its expected effects on the yen, the oil situation, and I examined what the SPX and the VIX are pricing in through year-end, while also explaining the current geopolitical landscape.

Make sure you read it — don’t even continue until you have.
This is a free post available to everyone.

Subscribe now

Now I’m gonna scrutinize the weekly positioning, PCE expectations and implied distribution…

Read more

First, let’s start with the broader geopolitical picture…
The Kremlin’s goal is not a settlement, it is the removal of Ukraine as an independent strategic fact — the punishment apparatus was built deliberately dual-readable, satisfiable either kinetically or by replacing the “office” in Kyiv — and its method is compellence: raise the pain until accepting Moscow’s terms becomes the only door left open. There is no built-in de-escalation duty anywhere in the system, only maximal exit optionality. Russia can stop whenever it wants by relabeling any state of the world as victory, while everyone else has to buy every outcome from Moscow at Moscow’s price. Negotiation, in this architecture, is not an alternative to the war; it is a weapons system of the war. That is the non-naive premise, and everything currently observable is downstream of it.
Ukraine is where the premise met resistance. Russia is in net territorial loss for 2026 — the first negative year of the war — while Ukrainian deep strikes grind down its refineries, fuel and rail. But NATO drew the correct, uncomfortable conclusion by seeing that a compellence machine that fails on its primary target does not stop, it retargets. The revised US assessment says precisely this — battlefield frustration makes the Kremlin more risk-acceptant, and the window for a limited test of NATO itself opens this fall and runs to 2029, with cyber, sabotage, deniable armed groups or a small land grab in the Baltics or Poland as the menu.
Kaliningrad is the hinge of that scenario, which is why NATO massed a fifty-aircraft electronic-warfare and strike package around it last week and framed it openly as deterrence, and why the satellite layer matters: infrastructure for a six-fold larger northern force grouping, an Oreshnik site in Belarus, a GRU sabotage unit parked next to Baltiysk — a steep capability curve, though with no acute mobilization signature yet.
The probing has already crossed into kinetics from both directions. A Russian-intelligence drone with military explosive at Leipzig airport, NATO fighters shooting down intruding drones over Latvia and Romania. The remaining battle is over attribution, which is exactly where the bot networks, the false-flag warnings and the fake “America quits NATO” posts operate — reflexive control aimed at one message, that helping Ukraine means war, and one prize, the perception of Western exit, which is cheaper than any real exit.

The Iran war and Ukraine together have drained the Patriot, THAAD, SM-3 and ATACMS stocks that a China contingency would also need, the oil shock refilled Russia’s war budget at the precise moment its fiscal clock was breaking, and Beijing — building the ability to win a Taiwan war by end-2027 while holding a blockade option convertible within hours — harvests the distraction as European market share.
None of this requires a master plan; it requires only that each member of the axis keep one Western front hot while the others bleed the pool, and wait not for economic collapse but for the cheaper thing: the moment Western political will breaks before Eastern capacity does.
That test has a date. It is scheduled against our own calendar — the US midterms, the Israeli election, the post-Duma Russian mobilization — and it runs through Q4 2026. NATO’s task is not to guess Moscow’s red lines, which the system can drop silently anyway, but to make the test fail: visibly, early, and at a price Moscow cannot relabel as victory.

Subscribe now

The new “Bessent-put”… really?

I call it a Bessent put, since this is primarily aimed at the headlines, and only a temporary solution, and mainly manages the sentiment. But let’s take a closer look.
On Wednesday the Treasury said it would at least double its long-end buybacks, four billion or more per operation across the 10s-to-30s sector, running September 9 through November 4. The announcement came the same day the national debt passed forty trillion, one day after the 30-year printed its highest yield since 2007.
The bond market needed 72 hours to deliver a verdict. 10 basis points lower on the headline, all of it given back by Thursday, and Friday’s close slightly above where the week began. Yields rejected the flow, and the arithmetic says they were right to: a few tens of billions over eight weeks against a deficit that ran past four hundred billion in July alone. Buybacks retire old bonds funded by new bills. It is a duration twist, not money creation, and a twist that size cannot set a level.
What the announcement actually did was reveal a reaction function: losing the long end is not acceptable. The market repriced that immediately, just not in bonds. Gold rose about 8% on the week. The dollar sold. And inside the index option complex, the implied financing rate on equity exposure moved 20 basis points richer.
That is how you pay for a put like this. Not in yield. But in the denominator.

The yen shows the same machine from the outside.
Japan and US intervened jointly at the end of July — the first American operation to support the yen in decades — and the design leaned on the Fed's repo facility for foreign central banks, built so Japan can raise dollars without selling Treasuries outright. Three weeks on, the currency has given back about half of those gains and sits near 159 per dollar; it jumped almost 1% on the buyback headline and returned more than half of that within a day.
The yen now trades as a derivative of the American long end. And Japan's own curve is removing the marginal buyer: the 10-year JGB touched 2.95%, its highest since 1996, the 30-year sits above 4%, inflation is accelerating and a September hike is priced — the money that used to fund the American long end finally gets paid at home.
The buyback, the intervention and the repo facility are one trade on three screens: keep the 30-year bid.
The BOJ sits at 1.0% with a September hike priced and its own minutes flagging upside inflation risk; the path points higher and faster. The 10-year differential has already narrowed to about 1.85 pts, the tightest in years, and a pure rate model would put the pair near 135 to 142 against a spot of 159 — the yen trades roughly 12% to 15% cheap, with the gap split between the oil import bill, a fiscal premium on Japan's own long end, and carry positioning.
That last piece is the fuel. The forward pays 4 to 5 yen a year to stay short the currency; the 2024 unwind took back 12 in five weeks.
So the base case is a 155 to 162 range with intervention defending the top and drift toward 152 to 156 into year-end if the September hike lands. A failed repression regime tests 163 to 168 and forces either a bigger joint operation or a faster BOJ. And a carry unwind snaps the pair to 145 to 150 in weeks — the branch that lands directly on the levered equity book.

I already written about Warsh. His doctrine, consistent from the confirmation hearing onward, is the opposite trade: remove guidance, lean on the single mandate, and let the term premium do the tightening.
A high long end is not his problem. It is his instrument. So the Treasury is now capping the very lever the Fed chair wants free. As long as the operation stays labeled liquidity support, the two frameworks can coexist on paper. The moment it looks like level targeting, it is fiscal dominance in the open. That is why Friday at Jackson Hole matters more than the usual August speech: PCE on Wednesday, Warsh at the podium after. If he validates back-end discipline, the steepener is back on. If he gestures at coordination, repression gets a central bank behind it. Positioning says the market expects him to say nothing structural, with a hedge on the first branch.
And note the calendar. The program ends November 4, the day after the midterms. The market has priced the window as the calm and the expiry as the event.

Subscribe now

Oil

The reason the long end will not stay down is sitting in the product markets. Diesel cracks settled above a hundred dollars for the first time on record. The economy does not run on crude, it runs on refined product, and by most counts around five million barrels a day of refining is offline between the strikes on Russian plants and the damage in the Gulf. There is no strategic reserve for diesel.
The one for crude is below 299 million barrels, the lowest since 1983, with a legal floor at 252 and a practical floor, according to the engineers who study the caverns, roughly where it stands today. A quarter of what remains may not be deliverable at all. The coordinated IEA release is three quarters done and ends within a couple of months. In Hormuz, the June memorandum expired on August 17 with no deal. Transits are running around ten ships a day against 130 before the war, crude moves again only under naval escort, Iran’s new security chief is threatening the workaround routes, and the same Treasury that is defending the long end is expected to roll out a fresh sanctions package on Monday. The oil lever, the FX lever, the dollar backstops: used. Buybacks were the last rung before the Fed’s balance sheet. With crude and the 30-year running a correlation of 0.85, every product-market headline lands directly on the bond Bessent is buying.

Over the past month the October heating oil contract rose 7% while October crude rose essentially zero. The entire move is a refining story, not a crude story. Crude itself round-tripped: down 15% into the August 5 low as the market digested the glut the memorandum had let out, then 17% back once it expired.
Volume confirms the breakout rather than betraying a squeeze — turnover in the October crude contract roughly tripled in the sessions after the crack cleared a hundred. The curve still is not pricing an event, but a regime, where the October-dated diesel crack sits near 97 against the front-month record of 102, so the market concedes about 5% of relief through the entire window in which the strategic reserve runs dry.
Crude implied vol has fallen roughly 20 pts in a month, from the high 60s to just under 50, while the underlying rose and the crack set records. The managed price is calm; the unmanaged price is screaming. That is the administered-scarcity signature and it means the crude tail is being offered cheaply precisely because the visible benchmark is the one the releases suppress. In the December crude book the largest open line among the key strikes is the 100 call, about 30,000 contracts, roughly double the biggest put strikes, with calls outnumbering puts nearly 2-to-1 in the wings. The fear is positioned above the market, in supply-shock insurance, and its center of mass sits exactly where the post-release clearing arithmetic has pointed since July.

One more piece of arithmetic that sharpens the calendar…
The reserve draws are running near six million barrels a week, and at that pace the system reaches its practical floor in mid-October — 2-to-3 weeks before the vote, not on it.
The sequence matters: the oil prop expires in mid-October, the bond prop on November 4. The final weeks of the Treasury put will be spent absorbing the very repricing the reserve’s exhaustion sets off, with that 0.85 correlation as the transmission belt. So crude implied vol touched 44 intraday last week, the neighborhood where this scarcity read would begin to break, and bounced.
Watch that level, the Wednesday draw, and the physical spread.
The schedule, not the news flow, sets the calendar.

Everything I coded is happening:

Subscribe now

SPX & VIX

Customers did not de-risk into any of this. Net customer call buying more than doubled, put selling grew by more than half, and the levered core held: roughly 80 billion dollars of synthetic long forwards laddered across September, October and December strikes, financing locked through box structures.
The December upside cap was bought back and the overwrite rolled down and in. A new quarter-end structure appeared for September 30 that only pays in dislocation. And about 2 bln dollars of crash convexity moved from dealer books to customer books in two days, which leaves dealers, for the first time in this dataset, net short the left tail after delta hedging. Customers now pay roughly 8-to-9 nine million dollars a day in theta to hold the package. Conviction has a carry cost, and they are paying it willingly.

The rates read is the part I find most interesting.
That 20 basis point rise in implied equity funding is not flat across the curve: it is highest for the terms that span October into early November, where buyback expiry, the refunding announcement and the midterms stack. And the cross-greek that works as the SPX-to-yields beta bet, the rhogamma of the complex, roughly tripled, customer-long. Positioning is paying for a world where equities and yields drift up together, the nominal co-movement regime, while realized correlation is still negative on every window from one month to six.
The book is betting the trend correlation. The tape is still pricing the shock correlation. That gap is the regime call, and it is the Bessent-credibility trade written in derivatives.

So can we say that customers are not afraid of inflation? — Not at all.
In the textbook, a positive stock-yield beta means growth is driving yields, inflation is anchored, and the residual fear is recession, so you hold bonds as the hedge. Run the cross-asset consistency check and that reading dies on contact. Gold up 8% in a week, while diesel cracks at records. A vol floor sold and close to half a million contracts of crash wings bought. And the positioning itself is short duration, meaning the levered long is a borrower, financing locked near 4.5%, holding a rhogamma tilt that pays when equities and yields rise together. Nobody hedging a recession builds that book.
What is being priced is repression, not disinflation. The positive beta is manufactured, not organic: the state caps the yield expression of the inflation shock, so the fear does not disappear, it migrates — off the curve and into gold, commodities, the vol surface and the equity funding spread. The beta bet tripled in size in two days while realized correlation still runs -0.3 to -0.5, so the book is paying to be early on a flip the tape has not delivered. It lives in the smile ride, works in the melt-up zone, and mechanically switches itself off below roughly 7350.
And the hedge choice is the confession basically. Duration protects only the deflationary break and loses in the stagflationary one, so the customers skipped bonds entirely and bought VIX convexity, which pays in both.
Nobody buys quadrant-agnostic crash insurance in a world where inflation stopped being the risk. You buy it in a world where you believe the state will suppress the symptom until it cannot.

Spot vol sits in the mid-15s and looks asleep.
The structure underneath is a staircase. Forward one-month vol runs about 12 into September, 14 into October, 16 by year-end, VIX futures plateau near 20 through the fourth quarter, and the premium of VIX futures over forward ATM vol widens with tenor. The market is paying up specifically for convexity, not level, the further out it looks.
The floor is sold: around 165,000 VIX puts at strikes from 15 to 17, roughly a coin flip to finish in the money, financing a tower of upside wings whose November leg only pays above a 30 VIX.
It’s not a bet on a vol crush, but maybe a regime break. That is the whole trade in one sentence.

Jackson Hole week is priced for movement, not regime: 1-sigma near 2%, about a one-in-fifty chance of finishing below 7300.
The September FOMC into OPEX window carries a 1-sigma of about 4%, with both tails fatter once dealer feedback is applied, and the upside compressing toward 8000, just below the fresh 8300 overwrites.
The midterm window prices election night itself as an ordinary macro day, worth an extra 1% or so of implied move on the day, but the distribution through November 20 has its median at the old high and a 16-to-18% chance of sitting below 7300. This is because the risk is not the vote, but what expires the day after.
Into the year-end the median is in the 7830 to 7870 area, above the August high, about a two-in-five chance of finishing above 8000, roughly one in nine below 7000.
Price says new highs. Positioning says pay for the corridor in between. A market long the nominal outcome, insured for the transition, and financed at a rate that tells you exactly which 8 weeks it is worried about.

So why do they expect the calm to expire right after the midterms? - Well, because the calm is not a state of the world, it is a program, and the program has dates.
The buyback window closes November 4, the day after the vote. The SPR and the coordinated IEA release are three quarters spent and run out on roughly the same schedule. The administration’s incentive to sit on gasoline, yields and the tape is maximal until election night and undefined after it. And everything that forces disclosure stacks into the weeks that follow: the refunding announcement that reveals the true issuance path, a December FOMC under a hawkish chair, the October and November CPI prints where product-price pass-through should first show up, the year-end funding turn, winter energy.
The market is pricing this, not guessing at it. The crisis odds embedded in VIX options roughly triple between the September and November expiries. Implied equity funding is most expensive for exactly the terms spanning October into early November. The most bearish SPX expiry book on the board is November 20, with its payoff peaking about 6% below spot.
Trump is not on the ballot, but his Congress is, and the markets that trade it are blunt. Democrats are at near 85% to take the House on a majority so thin a handful of seats flips it, the Senate a coin flip for the first time in the cycle, repriced from under 20% a year ago largely on the war, and a full sweep against him close to even money. Call it roughly one chance in seven that unified control survives the night.
So the scenario the option market is insuring is precise: the put’s political sponsor loses at least one chamber the day before the put expires. That is why the crash tower sits in the November 18 VIX expiry, not the October one.
The market is not afraid of the vote. It is afraid of the morning after, when the props are gone and the political cover behind them may be gone with them.

Subscribe now

Share

Here I wrote about expected OpEx-flows:

Bessent put

The Treasury is at least doubling the size of its liquidity-support buyback operations in long-dated nominal coupon securities (the 10–20yr and 20–30yr sectors). The current $2 billion maximum/operation will become at least $4 billion, from 9 Sept 2026 until the end of the refunding quarter (4 Nov), with further size details to be announced at the 4 Nov Quarterly Refunding.
The buybacks are financed by bill issuance: you sell at the front end (where $7.91 trillion of MMF assets sit as buyers) and buy at the back — net duration extraction from the market, yield-curve management without rate hikes. A pure piece of the financial-repression toolkit: the Fed sits at 3.50–3.75 with hawkish dissents, inflation is sticky, and the fiscal side is performing the duration management that monetary policy is not. They are pinning the nominal long-end with administrative tools; the relief valve remains inflation.

In this quarter a doubled official bid is now working against a supply-driven long-yield spike. The medium term, however, remains untouched — in fact, the very need to double the size precisely documents the selling pressure in my thesis; repression does not eliminate the problem, it merely shifts it into the inflation premium and into gold.
This is moderately risk-supportive for the weekly structure — it precisely cuts off that left tail in which a long-end supply scare would land in the Thursday–Friday –$2.2 billion short-gamma window.
The doubled long-end bid goes live on 9 September — one week before the 16 September node. The authorities are preparing for the same date the market is watching.
Not by chance. They see it too…

Geo

Meanwhile, the probability of a military incident in the Baltic region has risen to 15–20 percent over the next year, while the likelihood of the Middle Eastern war re-escalating this autumn is already above 35 percent. In practical terms, a global war is already being fought on one and a half fronts.
From a financial and market perspective, this means that beneath the seemingly calm surface the risk of a sudden collapse — the so-called left-tail risk — has swollen enormously. We are now sitting on two completely independent powder kegs — in the Baltics and in the Strait of Hormuz — either of which could trigger a global shock.

Let’s look at today’s positioning…

Read more

Here I broke down the SOQ, VIX-expo expected flows in details:

Take your time with it.

Bessent put

The Treasury is at least doubling the size of its liquidity-support buyback operations in long-dated nominal coupon securities (the 10–20yr and 20–30yr sectors). The current $2 billion maximum/operation will become at least $4 billion, from 9 Sept 2026 until the end of the refunding quarter (4 Nov), with further size details to be announced at the 4 Nov Quarterly Refunding.
The buybacks are financed by bill issuance: you sell at the front end (where $7.91 trillion of MMF assets sit as buyers) and buy at the back — net duration extraction from the market, yield-curve management without rate hikes. A pure piece of the financial-repression toolkit: the Fed sits at 3.50–3.75 with hawkish dissents, inflation is sticky, and the fiscal side is performing the duration management that monetary policy is not. They are pinning the nominal long-end with administrative tools; the relief valve remains inflation.

In this quarter a doubled official bid is now working against a supply-driven long-yield spike. The medium term, however, remains untouched — in fact, the very need to double the size precisely documents the selling pressure in my thesis; repression does not eliminate the problem, it merely shifts it into the inflation premium and into gold.
This is moderately risk-supportive for the weekly structure — it precisely cuts off that left tail in which a long-end supply scare would land in the Thursday–Friday –$2.2 billion short-gamma window.
The doubled long-end bid goes live on 9 September — one week before the 16 September node. The authorities are preparing for the same date the market is watching.
Not by chance. They see it too…

Geo

Meanwhile, the probability of a military incident in the Baltic region has risen to 15–20 percent over the next year, while the likelihood of the Middle Eastern war re-escalating this autumn is already above 35 percent. In practical terms, a global war is already being fought on one and a half fronts.
From a financial and market perspective, this means that beneath the seemingly calm surface the risk of a sudden collapse — the so-called left-tail risk — has swollen enormously. We are now sitting on two completely independent powder kegs — in the Baltics and in the Strait of Hormuz — either of which could trigger a global shock.

Let’s look at today’s positioning…

Read more

Here I broke down the SOQ, VIX-expo expected flows in details:

Take your time with it.

Bessent put

The Treasury is at least doubling the size of its liquidity-support buyback operations in long-dated nominal coupon securities (the 10–20yr and 20–30yr sectors). The current $2 billion maximum/operation will become at least $4 billion, from 9 Sept 2026 until the end of the refunding quarter (4 Nov), with further size details to be announced at the 4 Nov Quarterly Refunding.
The buybacks are financed by bill issuance: you sell at the front end (where $7.91 trillion of MMF assets sit as buyers) and buy at the back — net duration extraction from the market, yield-curve management without rate hikes. A pure piece of the financial-repression toolkit: the Fed sits at 3.50–3.75 with hawkish dissents, inflation is sticky, and the fiscal side is performing the duration management that monetary policy is not. They are pinning the nominal long-end with administrative tools; the relief valve remains inflation.

In this quarter a doubled official bid is now working against a supply-driven long-yield spike. The medium term, however, remains untouched — in fact, the very need to double the size precisely documents the selling pressure in my thesis; repression does not eliminate the problem, it merely shifts it into the inflation premium and into gold.
This is moderately risk-supportive for the weekly structure — it precisely cuts off that left tail in which a long-end supply scare would land in the Thursday–Friday –$2.2 billion short-gamma window.
The doubled long-end bid goes live on 9 September — one week before the 16 September node. The authorities are preparing for the same date the market is watching.
Not by chance. They see it too…

Geo

Meanwhile, the probability of a military incident in the Baltic region has risen to 15–20 percent over the next year, while the likelihood of the Middle Eastern war re-escalating this autumn is already above 35 percent. In practical terms, a global war is already being fought on one and a half fronts.
From a financial and market perspective, this means that beneath the seemingly calm surface the risk of a sudden collapse — the so-called left-tail risk — has swollen enormously. We are now sitting on two completely independent powder kegs — in the Baltics and in the Strait of Hormuz — either of which could trigger a global shock.

Let’s look at today’s positioning…

Read more

The VIX edged higher by +0.5 points this morning. The very first thing to decide when sitting in front of the screen is why.
If the VIX is rising because investors are actually selling stocks, then ES will open around 7710-7720 today (i.e. lower). This is the more dangerous scenario, because we are close to the edge of the abyss.
If the SPX does not fall (opens around 7740) but the VIX is still higher, it means nobody is panicking — they are simply buying insurance ahead of tomorrow’s Fed minutes. (The first half-hour of trading will give the answer.)
VIX dealers will be buying futures, which in the current correlation has a relatively moderate effect, but creates convexity to the downside.
The trapdoor remains at 7635/50.

Today neither the VIX can spike (it’s pinned between 15.5–16.3) nor can the SPX break down. The two markets are perfectly neutralizing each other. Even if the VIX pushes above 16.3, its impact on the SPX remains limited.

Let’s look at today’s positioning…

Read more

Let’s look at today’s positioning…

Read more

They remain completely relevant.

The equity market has entered a euphoric state, with call option volume exploding to historic highs. Although the market is pricing a relatively quiet week, the downside is underpriced, put skew is cheap due to VIX complacency, and this is typically when the odds of a vol squeeze start to rise. Behind the paywall, I will examine what the market is pricing for this week.
One of the main supportive flows right now — alongside the short squeeze and put decay buyback flow — is that the share buyback window is open, retail investors are buying again, NQ futures positioning has dropped to very low levels, and XLP has switched r/r to the upside.

Inflation is far from gone, and monetary conditions remain far too loose. Although core CPI printed 0.22%, core PPI is still elevated at 4.7%. Goods prices are rising again, software prices (driven by AI demand) have surged, and inflation in many services categories remains stuck above pre-pandemic levels. Supercore inflation is still running at 3.28%, creating a stalemate for the September rate decision.
Warsh, of course, wants to keep the long end elevated, and as I wrote earlier, this will produce a stagflationary and then recessionary environment. Deliberately elevated long-term yields will destroy the average small and medium-sized businesses, mortgage holders and industry much sooner than they can cool the tech sector’s AI investment boom. But that comes later — not now. Inflation has been left untreated and will not simply disappear.
The Fed’s policy rate remains below the Taylor rule. If the Fed stays on hold now, far more aggressive tightening will be required later.
The clash between the Fed and the market will not break inflation — it will break the labor market and the real economy, ultimately making the most drastic intervention inevitable.

Subscribe now

And now let’s analyse the priced-in probability distribution for OpEx week…

Read more

Let’s look at today’s positioning…

Read more

I’ve posted about PPI release…

Let’s look at today’s positioning…

Read more