A weak autumn, not a recession

The Liquidity Map · Macro report

A weak autumn, not a recession

Oil climbed straight into the resistance we were watching, got rejected, and came back down. Right as every market heads into the same weak window. Why that is not a recession, and where we actually expect the real recession.

July 29, 2026 · 8 min read

Key takeaways

High rates are not the danger. We are in a secular high-rate regime, and the indicator to watch is the 3-month yield against the Fed rate, not the rate level itself.
The September 2026 to January 2027 window is a natural weakness, not a real recession. Without inflation falling hard, there is no real recession.
We watch the real recession from September 2028, when the Shmita year begins, the 7-year cycle. The next one falls in 2035.
Gold came down from 5,318 dollars in January to 4,049 now, silver from 115 to 58. A deep correction, but inside a trend that stays constructive.
Oil tested the 87-92 zone, a major resistance, and was rejected there. It is a short zone, but we wait for stabilization before entering, not catch a falling knife.

Oil climbed on July 23 to 92 dollars, right into the zone we were watching as major resistance.

It looked ready to explode higher.

It did not. It was rejected there and came back down toward 83. Exactly the kind of zone where the market runs out of steam, not where it accelerates.

It is a good place to start the wider picture. Every asset class is heading into the same weakness window, from summer into autumn. But weakness does not mean recession. We watch the real recession much further out.

Definition

The 3-month yield

The interest rate the market demands for short-term US debt, 3 months. Unlike the Fed rate, which is a decision, the 3-month yield moves daily, driven by real investor money. As long as it does not fall aggressively below the Fed rate, the bond market is not pricing in a real recession.

Rates

High rates are not the danger

High rates sound like danger, but historically it is the opposite. We are in a secular high-rate regime that likely holds until around 2032-2034. High rates did not cause the big crises, 2008 and 2020. Aggressive cuts did.

That is why the indicator to watch is not the Fed rate level, it is the 3-month yield. It is the market's real voice, updated daily, not a committee decision.

It is not the high rate that scares. Aggressive cuts brought the crises.

The rule for reading the indicator is simple: a yield falling below the Fed rate only carries weight if inflation drops at the same time. Without the inflation piece, a sharp move in yields is market noise, not a panic signal.

Neither sign shows up right now, no deflation, no aggressive drop in yields. That puts the right label on the September-January months ahead: a softer season, not a collapse.

3.96%The US 3-month yield right now, above the effective Fed rate of 3.63%. As long as it does not fall aggressively, recession risk stays low.

The Fed rate, the past 38 years. Grey bands are past recessions, each after aggressive cuts, not sustained high rates.

The gap between the 3-month yield and the Fed rate, the past 3 years. As long as it stays near zero or positive, like now, the bond market is not pricing recession risk.

With that clarified, the natural next question: if not now, when does the real recession arrive?

Recession

Weakness now, real recession from September 2028

A real recession is not just two weak quarters on paper. It means inflation falling hard, toward deflation. Without that, you have a market cooling off, not one entering a crisis.

The September 2026 to January 2027 window reads exactly that way: a seasonal weakness, synced across several asset classes, not a real recession. After it, the picture clears up again.

A real recession comes with deflation, not just a market catching its breath.

We watch the real recession from September 2028. That is when the Shmita year begins, the 7-year cycle we have written about at length here, where periods of outperformance end and the market resets.

Then the cycle reopens.

From 2029 the picture stays constructive into 2034, and the next Shmita year falls in 2035.

In practice, there are two distinct windows, not one: a manageable autumn weakness, and a real recession, much later, one we are only starting to prepare for mentally.

Sept 2028When we watch for the real recession, as the Shmita year begins. The next one falls in 2035. The current window is only a seasonal weakness.

US inflation year over year, past recessions marked. It fell hard every time, down to deflation in 2009. Right now it sits at 3.5%, no sign of an aggressive drop.

With the time windows clear, look now at what the assets are saying. Start with metals.

Metals

Gold and silver, a pause, not an end

Gold remains the underlying macro anchor, the reason risk assets get compared to gold rather than the dollar. It stays watched constantly, regardless of the short-term window.

Short term, the correction has already started. Gold peaked at 5,318 dollars an ounce in January and sits at 4,049 now, nearly a quarter lower.

Silver went far harder: from 115 dollars to 58, close to half its value, in the same window.

It is a pause inside a trend that stays constructive underneath, not an exit signal.

A weaker window toward the end of the year, not the end of a cycle.

2027 is less shaped up. The working assumption is a trendless stretch, more sideways than a firm direction. A return toward the highs stays possible once the weakness passes, but without any guarantee of a fresh record. Patience beats conviction here.

For now, the metals stay underlying context, not a level to trade today.

$4,049 · $58Gold and silver now, against 5,318 and 115 dollars at the January peak. A deep correction, inside a trend that stays constructive.

Gold (navy, left axis) and silver (gold, right axis), real prices in dollars per ounce. Both correcting from the January peak, silver far more sharply.

And if metals are the backdrop, oil is the signal moving right now.

Oil

Do not catch the falling knife

Long term, oil stays structurally in an uptrend, part of the commodity cycle heading toward hundreds of dollars a barrel by around 2030-2032. This year, though, has stayed weak, especially in the first half.

Short term, WTI tested the 87-92 dollar zone on July 23, touched 92, and was rejected there. It is the zone where short-term enthusiasm hits an older ceiling, exactly as we were watching it.

You do not catch a falling knife. You wait for a breather, then decide.

The confirming signal came from RSI.

It pushed above 70, into overbought, on the very day price hit 92, then dropped quickly toward 50.

Buyers ran out of fuel. Right at the ceiling.

Without a quick invalidation, the target on the zone naturally climbs toward 95. The position still gets built in steps, after confirmation, not all at once, eyes on the September-October window, right where several assets turn weak together.

$87-92The major resistance zone I am watching on oil. Without a clear invalidation, the short zone can stretch toward 95.

WTI on top, RSI below, same time axis. On July 23 price hit 92, the top of the resistance zone, and RSI crossed above 70, into overbought, on the same day. Since then price came down toward 83 and RSI toward 50.

Oil gave the first signs because it was the asset that ran straight into resistance. The rest of the market runs on the same clock.

The common thread

Every asset class, the same window

Any rebound on oil tends to weaken from summer into September-November. It is not an isolated coincidence, it is part of a wider sync: several asset classes are heading into the same weakness window, at the same time.

Oil gave the first signs, because it was the asset that climbed straight into a clear resistance and got rejected there. The rest of the market moves on the same clock, just with a small lag.

Not every asset weakens at once by accident. They sync up.

Important to separate: this is a weakness window, not the real recession discussed earlier. The kind of period where the market breathes harder, not one where something structural breaks.

A possible energy shock stays on the radar for next year, mentioned calmly, as a scenario to watch, not a certainty or a reason to panic now.

Sep-OctThe window in which several asset classes, including oil, turn weak at the same time. Sync, not coincidence.

Above every window, one thing matters for you: how you position.

How we're positioned

That is the framework. The part that matters for you is how we sit inside this window.

We do not position on headlines or on how scary a chart looks for a few weeks. We position on structure, with a clear invalidation for every idea, whether it is rates, metals, or oil.

On oil, we do not enter on the way down. We wait for a second attempt into the 87-92 zone, with a lower high and an RSI that no longer reaches 70, before sizing in, and even then in steps, not all at once.

On metals, we treat the weaker year-end window as a pause to accumulate patiently, not as a reason to exit the underlying theme. The average matters, not a single day's low.

And on the autumn window as a whole, we do not rush to the recession conclusion. We treat it as a tougher, manageable season, while the real recession stays a topic for much later, toward late 2028.

Conclusions

One thread remains: the autumn window is a synced weakness across several asset classes, not a recession. We watch the real recession much further out, toward late 2028, in the Shmita year.

High rates are not the danger, the 3-month yield is the barometer to watch. Gold and silver enter a pause toward year end, with 2027 still unclear. Oil gives the first signal, in the 87-92 zone, but without being caught on the way down.

Above all, structure decides. Not the headline of the day. You have a hypothesis, you wait for confirmation, then you move in steps, with a clear invalidation. The rest is seasonal noise.

What to watch this week

The 3-month yield: is it falling aggressively below the Fed rate, with inflation dropping? Only then are we talking real risk, not just a market catching its breath.
Oil: on the next attempt, does price make a lower high while RSI fails to reach 70 again? That confirms the rejection. A close above 92 with strong RSI invalidates the thesis.

A Weak Fall, Not a Recession

Keep reading

2027 is THE YEAR

The consensus wants an October bottom. A green August says otherwise

The math for a rate cut