Kresmion's Regime Score Has Almost Recovered From Its First Risk-Off Print. The Rebound Is Built on One Backward-Looking Factor.

By Kresmion Research, June 30, 2026

Kresmion's cross-asset regime score has nearly recovered from its first Risk-Off reading, but the rebound rests almost entirely on one backward-looking factor. Strip that factor out and the model is not really healing.

This note describes what Kresmion's regime model currently reads and how its components are moving. It is not investment advice, a price forecast, or a recommendation about any asset.

Key takeaways

What the regime score actually says

Kresmion runs a single cross-asset regime score from minus 3 to plus 3, exponentially smoothed day over day. It blends four factors: liquidity carries the largest weight, then risk appetite, then growth, with volatility the smallest. Above plus 0.5 the model labels the tape Risk-On, below minus 0.5 Risk-Off, and the band between is Neutral. The methodology is described in full on the Kresmion methodology page.

On June 25 the score printed minus 0.56, the first and only Risk-Off reading in the model's history, which begins on April 30. We covered that print at the time in the June 25 Risk-Off note. Over the five sessions since, the score has climbed back to minus 0.22 as of June 30. That is a recovery of about 0.34 points, and it has carried the label from Risk-Off back to Neutral with high conviction at 75 percent.

On the surface, that reads like the all-clear. A closer look at the components says something more specific. Minus 0.22 still sits at roughly the 24th percentile of the model's 62-session history. The tape has recovered, but it has not returned to neutral health. It has recovered to the low end of a range.

The rebound is built on one factor

The growth factor is the engine of the recovery. On June 25 it read minus 0.73. By June 30 it had flipped to plus 0.46, a swing of nearly 1.2 points. The driver is the economic-surprise input, which the model reads at about plus 1.9 standard deviations above its one-year norm. In plain terms, recent hard data has been beating forecasts. That is consistent with recent data, including a firmer labor market and the May ISM manufacturing reading of 54.0, its highest since 2022.

Here is the part that matters for reading the score correctly. Translate each factor into its weighted contribution to the raw score of minus 0.21 and the picture changes. Growth contributes plus 0.09. Liquidity contributes minus 0.08. Volatility contributes minus 0.01. Risk appetite contributes minus 0.21. Growth's positive push almost exactly cancels the drag from liquidity and volatility combined. What is left, and what essentially defines the entire negative reading, is risk appetite.

So the headline recovery is real in the arithmetic, but it is concentrated. One factor, growth, is doing the lifting, and the input behind it is a measure of how recent data came in versus expectations.

What is quietly deteriorating

While growth recovered, risk appetite went the other way. On June 25 the risk-appetite factor read minus 0.55. By June 30 it had fallen to minus 0.69, the worst single reading of the entire June 25 to June 30 episode. Two inputs drove it lower. High-yield credit spreads widened, which the model reads at about plus 0.94 standard deviations, and equity-trend momentum stayed negative at about minus 0.80 standard deviations.

One honest caveat on the decomposition. The high-yield spread input feeds both the liquidity factor and the risk-appetite factor in the model, so the negative contributions from those two are not fully independent of each other. Equity-trend momentum, the second driver of the risk-appetite drop, is a separate input that is not shared with the liquidity factor.

That equity-momentum reading deserves a note, because it sits oddly against a stock market near record highs. The S&P 500 is up in the high single digits for the year, and the Dow set a record above 52,000 this week on the day Alphabet joined the index. The reconciliation is the time frame. Kresmion's momentum input is a trailing trend measure, and June itself was soft, with the Nasdaq down about 6 percent and the S&P down about 2 percent on the month even as the year-to-date number stayed positive. The model is registering the recent trend, not the year-to-date level.

The distinction between the two recovering and deteriorating sides is the distinction between realized data and market-priced risk. The economic-surprise index measures how recent data releases compared to forecasts, a read on the recent past. High-yield spreads and equity-trend momentum are prices set by markets that are continuously trying to discount what comes next. One set is improving. The other is not.

Why this is happening

The policy setting ties the two together. The Federal Reserve held its target range at 3.50 to 3.75 percent on June 17 and signaled no urgency to ease. The 10-year Treasury yield sits around 4.37 percent. In a setting where policy is restrictive and cash earns a real yield, two things can be true at once. Recent data can run hot, because the economy still has momentum from earlier conditions, and forward-looking risk pricing can grow more cautious, because a restrictive stance held for longer raises the odds that something downstream slows.

That is the textbook shape of a late-cycle divergence. Lagging data flatters the headline while leading market-based risk gauges turn down underneath it. Kresmion's regime model does not editorialize about that. It simply weights the inputs and shows where the score is coming from, and right now the answer is that the calm reading is being held up by the backward-looking side.

The timing sharpens the point. June 30 is the close of the first half of 2026, the day desks mark their books and write their reviews. The surface story of the half is a calm one, with the volatility index near 17.5 and indices near records. Kresmion's volatility factor has fallen back to a near-zero contribution, with its VIX input around plus 0.07 standard deviations, a long way from the spike that defined June 25. That fading volatility stress is a large part of why the score recovered at all. The composition underneath that calm is the part a single index level does not show.

The counter-evidence

The honest case against this read is that the recovery is simply genuine, and that is worth stating plainly.

First, the score did rebound by a real 0.34 points, the label is back to Neutral at high conviction, and the volatility factor returning to neutral is a true easing of the acute stress that defined June 25. None of that is an artifact.

Second, the growth improvement is not fake. The economy has been beating expectations on recent hard data, and manufacturing sentiment is at a multi-year high. A model that weights a near plus 1.9 standard deviation economic-surprise reading positively is doing the right thing. The economy beating expectations is a genuine positive, not a statistical quirk.

Third, risk appetite is only one of the four factors, not the whole model, and credit spreads at plus 0.94 standard deviations are elevated but not extreme. A widening this size has happened many times without a downturn following. A reasonable reading is that the model recovered, and risk appetite is one cautious component within an otherwise improving mix.

The argument here is not that a decline is coming. It is narrower and more checkable than that. The recovery is real but lopsided, it is carried by the backward-looking factor, and its pace has already begun to stall. The smoothed daily gain fell from plus 0.092 on June 29 to plus 0.018 on June 30, and the raw score actually slipped from minus 0.200 to minus 0.214 on the day.

What would change the read

This week's hard data is the test, and the calendar is front-loaded because markets are closed Friday July 3 for the Independence Day holiday. The June ISM manufacturing report lands Wednesday July 1, against a prior reading of 54.0, and the June employment report is pulled forward to Thursday July 2.

If the data stays strong and the growth prop holds while risk appetite turns up, meaning credit spreads tighten and equity momentum recovers, the rebound broadens and becomes a real return to health. If instead the data disappoints and the growth prop fades while risk appetite stays pinned at its lows, the score loses the one thing holding it up and would have room to roll back toward the June 25 reading. A recovery carried by a single backward-looking factor is fragile by construction. That fragility, not any forecast, is the observable thing to watch this week. For the data behind the score, see Kresmion research and the latest research notes.

Frequently asked questions

What is the Kresmion regime score?

It is a single cross-asset measure of market conditions running from minus 3 to plus 3, exponentially smoothed day over day. It blends four factors, growth, liquidity, risk appetite, and volatility, and labels the tape Risk-On, Neutral, or Risk-Off. The full method is on the methodology page.

Does a recovering score mean the market is healthy again?

Not necessarily. The score recovered to minus 0.22, but that still sits near the 24th percentile of its history, and the recovery is concentrated in the growth factor. A rising headline number can hide a lopsided composition, which is the entire point of this note.

Why is equity momentum negative when stocks are near record highs?

Because the momentum input measures the recent trend, not the year-to-date level. June was a soft month for equities even though the year-to-date return stayed positive, so a trailing trend measure can read negative while the index level looks strong.

What would confirm or break this pattern?

Watch this week's ISM manufacturing report on July 1 and the employment report on July 2. Strong data plus a turn up in risk appetite would broaden the recovery. Weak data with risk appetite still at its lows would remove the support the score currently leans on.

Sources

• Kresmion macro regime model (macro_regime_history), readings for June 23 to June 30, 2026
• Federal Reserve, FOMC statement, June 17, 2026
• Trading Economics, US 10-year Treasury yield
• ISM Manufacturing PMI at 54.0, May 2026
• Kiplinger, this week's economic calendar
• Yahoo Finance, market close, June 29, 2026
• Kresmion, June 25 Risk-Off note