Risk-On vs Risk-Off Regimes: How to Read the Market's Mood From Cross-Asset Data

Every macro newsletter uses the phrase "risk-on" or "risk-off" like it's a settled fact about the day, the way you'd say the sky is cloudy. In practice the regime is a fuzzy, multi-signal read, and the clean textbook version — stocks up, bonds down, dollar weak, VIX low, equals risk-on; flip every sign for risk-off — has broken down hard enough times in the last few years that trading it mechanically will get you run over. The useful version of this framework isn't a binary light switch, it's a basket of cross-asset signals you check against each other, because when they disagree, that disagreement is itself the information.
The Textbook Signature, and Why It's Incomplete
The classic risk-on/risk-off (RoRo) framework groups assets into two buckets. Risk-on: equities (especially high-beta names and small caps), commodities tied to growth (copper, oil), high-yield credit, and carry-trade currencies like AUD and NZD. Risk-off: US Treasuries, the US dollar, the Japanese yen, gold, and the VIX. The idea is that capital rotates between these risk assets and defensive assets based on investor sentiment, and you can read the rotation off a handful of charts. Risk-on conditions tend to show up alongside dovish central bank signals, strong corporate earnings, and easing economic tensions — the backdrop that makes investors comfortable reaching for yield instead of parking in Treasuries.
The VIX itself has rough, widely-used bands: under 15 is a calm regime where mean-reversion strategies work at full size; 15–25 is a normal regime; 25–40 is elevated and most desks cut size by a quarter to a half; above 40 is crisis territory where you're sizing down 75% or standing aside entirely. Those bands are a reasonable starting point, but the VIX level alone tells you about expected volatility, not direction — and a rising VIX during an equity rally is itself a signal worth flagging, since it often means traders are buying protection even while price grinds higher, a classic late-rally tell.

2022 Is the Case Study Everyone Should Have Studied by Now
The biggest crack in the simple framework happened in 2022, and it's not ancient history — it's the most recent full year where the standard playbook failed completely. A 60/40 stock-bond portfolio lost roughly 17% that year, its worst showing since 1937 and one of the four worst results in a 200-year sample. The reason: the stock-bond correlation, which had been reliably negative for the better part of two decades (bonds rally when stocks sell off, cushioning the portfolio), flipped to roughly +0.50, the highest reading in the modern data. Inflation and rate hikes were the dominant macro force, and under that regime stocks and bonds trade as one correlated bet, not as hedges for each other. If you were reading risk-off purely as "bonds should be rallying while equities fall," 2022 handed you a wrong signal for the better part of the year.
The dollar-yen relationship told a similar story. In 2020, during the pandemic shock, USD/JPY carried roughly a +0.35 correlation with the MSCI World index — meaning when equities fell, the yen typically strengthened against the dollar, acting as the safe haven it's supposed to be. By 2022, with USD/JPY trading above 140, that correlation with equities had collapsed to effectively zero. The yen had lost its safe-haven behavior because Japan was absorbing its own negative shock — a collapsing trade surplus from surging energy import costs, combined with the Bank of Japan holding rates near zero while the Fed hiked aggressively. The lesson isn't "the yen is no longer a safe haven asset" as a permanent rule; it's that safe-haven status is conditional on the nature of the shock and the currency's own balance-of-payments position at the time, which most retail explanations skip entirely.
Credit Spreads: The Signal That Leads Instead of Confirms
If you only watch one additional series beyond the VIX, make it the high-yield credit spread — the gap between junk bond yields and Treasury yields. Equity markets can stay buoyant on short-term optimism or speculative positioning well after the fundamentals have turned; credit markets are less forgiving, because bondholders get paid back a fixed amount and care disproportionately about default risk. Widening high-yield spreads have historically preceded significant equity corrections, functioning as an early-warning signal rather than a lagging confirmation. A stable or narrowing spread alongside a wobbly equity tape is actually a point in favor of "this is noise, not a regime change"; a spread that's blowing out while equities are still near highs is the more dangerous combination, because the credit market is telling you something the equity tape hasn't priced in yet.
A Practical Checklist Instead of a Single Indicator
Rather than trying to collapse the regime into one number, run a short checklist and weight the signals by how many of them agree:
VIX level and direction. Low and falling supports risk-on; rising even from a low base is worth noting on its own, independent of the absolute level.
High-yield credit spreads. Widening spreads are a genuine early warning; use them to sanity-check an equity rally that credit isn't confirming.
Stock-bond correlation, not just bond direction. Check whether bonds are actually behaving as a hedge right now, because that relationship is regime-dependent, not a law of nature.
USD/JPY versus equities. Don't assume the yen will strengthen on every equity selloff — check the current correlation, because it moves with Japan's own trade and policy position.
Dollar index (DXY) direction. A rising dollar typically pressures commodities and gold and often coincides with risk-off flows, but treat it as one vote among several, not the deciding one.
When four or five of these line up in the same direction, you've got a real regime and it's worth adjusting size and exposure accordingly. When they're split — say the VIX is calm but credit spreads are quietly widening, or equities are grinding up while the yen is also strengthening — that disagreement is the actual trade signal: something is unresolved, and mechanically sizing up on a "risk-on" label because one chart looks green is how traders got hurt in 2022 and will get hurt again whenever the next regime shift breaks the easy pattern-matching.
The Honest Takeaway
Risk-on/risk-off is a useful shorthand for talking about markets quickly, not a tradeable signal on its own. The individual legs of the framework — the VIX, the yen, credit spreads, the stock-bond correlation — each carry real information, but they can and do decouple from each other for months at a time when the dominant macro driver changes, as 2022's inflation shock and the yen's lost correlation both demonstrated. If your process depends on a single cross-asset chart telling you the regime, you're one decoupling event away from a bad trade. Track the basket, weight by agreement, and treat disagreement between the signals as data in its own right rather than noise to be ignored.