Everything loosened except the cycle

Where we are

The market regime reads YELLOW and has since August 3, when it softened from RED. It is still sitting on the line. Consumer sentiment is 55.2 against a RED/YELLOW threshold of 55 — a margin of 0.2 — and the reading on file is unchanged from a week ago because it is the same print, dated July 31. The consumer-sentiment feed Jawz ingests publishes its final reading at month-end; the collector ran on August 14 and returned no new number, which is the feed behaving normally rather than a gap. So the margin that made last week's colour provisional is unresolved, not narrowed. A colour change from this position remains a data artefact until proven otherwise.

The business cycle is the steadier read and it has not moved. FALL — stagflation — has held in the weekly series since July 20, four weeks, and the engine marks the quadrant confirmed with no provisional reading pending. Confidence in the read rose to high this week, from moderate in each of the two prior weeks. That is the first firming of the cycle signal in this window, and it firmed in the direction the cycle was already pointing.

The quadrant is nonetheless on its own boundary, and for the same structural reason as last week: the growth cycle score reads neutral rather than negative, so the quadrant would turn on a small upward move in growth. Two of its three components are positive — initial claims and consumer sentiment both improving — with the manufacturing proxy neutral. The growth indicator pillar, a different measure, reads green outright. The two disagreeing is information: growth data looks healthy in the level and is not weak enough to be doing the work here. What holds the cycle in FALL is the other axis. Core PCE and core CPI are both still rising, only five-year breakevens are stable, and inflation remains the dominant risk factor.

The holdout eased

Financial conditions read loose for a second week, and this week the composite direction printed easing — the only easing reading in thirteen weekly rows. Every other row in the window reads stable or tightening. That is the single clearest change in this edition, and it is worth being precise about what produced it, because the label moved further than the underlying pillars did.

The pillar that moved is the one flagged in the last edition as the standing obstacle. The 10-year real yield fell 8 basis points to 2.36%, its lowest in four weeks. It is still classified restrictive, exactly as it has been in all thirteen rows, and it is still up 20 basis points over the twelve weeks. One week down does not undo a quarter up, and the classification has not changed. What changed is that the pillar setting the cost of carry stopped moving against every other pillar.

The rest of the picture is calm and getting calmer. Volatility is 14.6, down from 15.2, the lowest reading in the window and down about 2 points over the quarter. The broad dollar index is 119.06, down 0.5% on the week and also a window low, holding the neutral classification it regained on August 10 after four weeks classed restrictive. Credit is priced for nothing at all: high-yield spreads are unchanged at 271 basis points and investment grade widened a single basis point to 79, both calm, and over the full twelve weeks high yield is 3 basis points tighter while investment grade is 5 wider. Neither of those is a signal; the flatness is.

The honest caveat is about what a loose-and-easing composite is not. It describes the price and availability of money, not the economy those conditions are financing. Conditions have now been loose for two weeks while the cycle has been in stagflation for four, and neither read is contradicting the other — they measure different things. An investor who treats the easing print as evidence the cycle is turning is reading a conditions series as a growth series.

Two weeks is not three

Measured on a constant three-bank basis — the Fed, ECB and Bank of Japan at both ends, so coverage changes cannot fake a move — global liquidity is $17.70T, down from $18.06T twelve weeks ago. That is a 2.0% contraction and it remains the standing trend. But the trough was August 3, at $17.42T, and the two weeks since have added $220B and then $60B. Liquidity has now risen two consecutive weeks and stands exactly level with where it was on June 22.

The last edition set the bar for calling this a genuine change at three consecutive weeks of expansion. This is week two, and it is a much smaller week than the first. Over the two weeks together, the Bank of Japan added $180B and the ECB $80B while the Fed added $20B — US net liquidity has been close to flat all quarter, and the entire move is Frankfurt and Tokyo, which is to say the same two banks that did the draining.

That composition carries a specific caveat that should be stated with the number rather than after it. These are dollar-denominated readings of non-dollar balance sheets, and the dollar fell 1.4% over the same two weeks. A weaker dollar mechanically raises the dollar value of the ECB and Bank of Japan holdings, so part of this expansion is revaluation rather than policy — and it is the same dollar move that eased the conditions composite, which means the two readings are not independent evidence of each other. It does not account for all of it: the Bank of Japan's dollar figure rose 4.6% against a 1.4% currency move. But the dollar index used here is a broad basket rather than a yen-specific rate, so that is a bound on the revaluation share, not a decomposition of it.

Two further things belong in the open. Five of the thirteen weekly rows are reconstructed point-in-time rather than recorded live, so the early-window detail is less reliable than the recent end. And the headline four-bank series flipped basis twice inside this window — China dropped out of coverage on June 29 and re-entered on July 6 — which moves the headline by the size of the PBoC with nothing real happening. Every figure above is on the constant three-bank basis for precisely that reason.

China resolved the item that had been on the watch list for two editions. The July PBoC publication is now on file, 46 days old, putting the Chinese component at $7.44T against $7.32T for the June print, and the four-bank aggregate at $25.14T. That is a second consecutive monthly increase, which by the standard set in the August 2 edition makes it a stance rather than an inflection. The same currency caveat applies to the size of it — this is a dollar conversion of a yuan balance sheet — so the direction is the durable claim here and the magnitude is not.

What to watch

• The third week of liquidity expansion. Two weeks is the bar half-cleared, and the second week was a quarter the size of the first. A third week of ECB and BoJ expansion would make the dominant force of this quarter genuinely in question; a flat or negative week would make the last fortnight look like a dollar move wearing a balance-sheet costume.
• Whether the easing direction survives a second print. One easing row in thirteen is an event, not a trend, and it was produced largely by two pillars at window lows — a volatility reading of 14.6 and a dollar index of 119.06 — both of which have more room to revert than to extend.
• Consumer sentiment at month-end. The 0.2 margin has now gone unresolved for two weeks. The next final print decides whether the headline colour reads YELLOW or RED, and either outcome would be a statement about a survey crossing a threshold, not about markets.
• The growth cycle score. It reads neutral with two of three components positive, and it is the thing standing between FALL and SUMMER. The cycle firmed to high confidence this week without that score moving — worth watching whether the next move firms it further or tips the quadrant.
• Data quality is clean: all 27 sources report current, nothing stale, no release alerts outstanding, and the daily collector has run 26 times without a failure. Where a figure above is reconstructed rather than observed live, it is labelled as such.

The frame, not the trade

The specific way to be wrong this week is to let the labels vote together. Conditions turned easing, liquidity rose, the dollar and volatility hit window lows, and the Chinese balance sheet expanded a second month — four separate readings pointing the same way, arriving at once, and at least two of them driven by the same 1.4% move in the dollar. Counting correlated readings as independent confirmations is how a comfortable week becomes an expensive conviction. Meanwhile the cycle, which is set by inflation rather than by conditions, did not move at all, and the regime is still 0.2 of a point from changing colour on a survey revision.

The question that puts to a real book is about what your positions are actually indexed to. If something is sized for tightening conditions, this is the second week of evidence against that premise, and the useful move now is to write down how many more weeks it would take to change your mind — before the weeks arrive and the answer becomes retroactive. If something is sized for the stagflation quadrant, note that it has held four weeks and firmed to high confidence, but that it turns on a growth score currently reading neutral, so the invalidation is worth stating rather than assuming. And if a position rests on global liquidity turning, ask how much of your evidence for that turn is balance sheets and how much is the dollar, because this week those two answers are not the same size. That is the exercise: not what the labels say, but which of your positions would need reconsidering if the one label you are relying on is the one that moves next.