Multi-Asset Regime Models

Multi-Asset Regime Models

An equity-only regime model is reading half the page. The other half is being written in rates, FX, and credit, often days before the equity regime catches up. The discipline is to layer regime states across asset classes and read concordance and divergence as a unified signal.

5.4.1Why Equity Regime Alone Is Incomplete

An equity-only regime model labels the world based on equity returns, equity volatility, and equity internals. It captures what the stock market is doing but is structurally blind to what the stock market is being told. Stocks are downstream of capital flows that move first through the bond market, the dollar, and the credit complex. A regime that looks Bull from the equity tape can already be Crisis in credit, and the equity regime has tended to catch up, usually within two to four weeks — though that timing is probabilistic, not mechanical.

The intuition is simple: equity is the residual claim on the cash flows of corporations. When the cost of capital changes (rates), the demand for the currency the cash flows are denominated in changes (FX), or the perceived risk of those cash flows changes (credit), the equity tape is the last to know because it is responding to the slowest derivative. A multi-asset regime framework gives equity the context that equity alone cannot provide.

The current SomerQuant cross-asset state is equity Bull (0.89) + credit risk-on (HYG/LQD compressed, +0.4σ below mean) + DXY Neutral (within +/-0.6σ) = aligned Bull. All three asset classes are voting the same way, which is the highest-conviction regime call you can make. The size you take on aligned regimes can be meaningfully larger than the size you take on regimes where one asset class is dissenting.

Stocks tell you what is happening. Bonds tell you why. The dollar tells you who is paying for it. Credit tells you whether the bill is collectible.

5.4.2The Rates Regime Layer

The rates regime is defined by the level and slope of the yield curve, not by absolute yield. The two states that matter are the flattening risk-off state (a bull flattener: long end falling faster than the short end — classic flight-to-safety) and the steepening risk-on state (a bull steepener: short end falling faster than the long end — classic policy easing into recovery). Within those, a flat or inverting curve is its own state with strong implications for equity multiples.

The two-input rates model uses the 10Y yield (level) and the 10Y-2Y spread (slope). A persistent rise in 10Y combined with a steepening spread is a classic risk-on rates regime — the bond market is pricing growth and is willing to demand term premium for it. A persistent rise in 10Y combined with a flattening spread is a tightening regime, which is hostile to multiples for high-duration tech and growth equities. A falling 10Y combined with a steepening spread is the easing regime that historically precedes equity bull markets after recession bottoms.

The rates regime today reads risk-on but tightening: 10Y near 4.4%, the curve mildly positive, and the trend over the last 30 days slightly higher in yield. That is a rates state that is supportive of equity broadly but selectively hostile to the long-duration side of the index. A trader running a rotation playbook gets a different signal from this rates state than from the headline equity regime call.

5.4.3The FX Regime Layer

The dollar is the global liquidity variable. A strong dollar tightens financial conditions for everything denominated in dollars or borrowed in dollars by foreign entities, which is most of the world. The FX regime layer uses DXY momentum on a 60-day window and reads three states: strong dollar (DXY breaking out, +1σ or more from 60-day mean), weak dollar (DXY breaking down, -1σ or more), and range (within +/-0.6σ).

The implications for equity are layered. A strong-dollar regime is hostile to multinationals (foreign earnings translate at lower rates) and to commodity producers (commodities priced in dollars become more expensive in local terms). A weak-dollar regime is the classic supportive backdrop for emerging markets, commodities, and US large-cap multinationals. The range regime is neutral and is the most common state — the dollar is most often a non-event for equities and only becomes an active variable at the breakout boundaries.

The current DXY is within the range state at +0.4σ from the 60-day mean. That confirms the equity regime call rather than dissenting from it — if the dollar were breaking out at +1.5σ while equities were Bull at 0.89, the divergence would force a sizing reduction. As it is, the FX layer is neutral-supportive.

5.4.4The Credit Regime Layer

Credit is the asset class that prices solvency. The HYG/LQD ratio captures the market’s vote on whether high-yield issuers will be able to service their debt relative to investment-grade issuers. When HYG outperforms LQD (ratio rising), the credit market is risk-on. When LQD outperforms HYG (ratio falling), the credit market is risk-off and is pricing in stress.

The credit regime layer uses two inputs: the level of the HYG/LQD ratio relative to its 90-day mean, and the change over the last 5 trading days. A ratio that is above its 90-day mean and rising is the cleanest risk-on credit signal; a ratio falling sharply below its mean is the cleanest risk-off signal. In SomerQuant’s backtests, the credit risk-on-to-risk-off turn has tended to precede equity regime transitions by roughly 5 to 15 days — though credit has also fired without any equity transition following, so treat the lead as probabilistic, not mechanical. This is why credit is the highest-priority leading indicator in the multi-asset stack.

The current credit state reads risk-on with the HYG/LQD ratio 0.4σ above its 90-day mean and modestly rising over the last 5 days. That confirms the equity Bull call. If the ratio were 0.8σ below its mean and falling, the dissent would be a leading indicator that the equity Bull was on borrowed time.

5.4.5Building the 3-Asset Composite Score

The composite is a single score from -3 to +3 that aggregates the three non-equity asset classes (rates, FX, credit). Each gets a vote of +1 (confirms equity Bull), 0 (neutral), or -1 (dissents from equity Bull). The composite score is the sum.

The interpretation is direct. A composite of +3 means all three non-equity asset classes are aligned with the equity regime call — this is the regime in which to take maximum size on regime-consistent setups. A composite of 0 to +1 means the equity call is uncorroborated; size at half. A composite of -1 or worse means the dissenting asset classes are voting that the equity regime is mispriced; reduce sizing meaningfully and treat the equity regime call with skepticism. A composite of -3 with the equity regime still labeling Bull is the classic late-cycle setup — equities have not yet caught up with what the rest of the markets are pricing.

The current composite is +1 (rates: 0, neutral-supportive; FX: 0, the range state is neutral by definition; credit: +1, risk-on). That is weak corroboration — full intent on quality setups only, per the table below.

5.4.6Common Mistakes

• Treating equity regime as the whole regime. Equity is downstream; rates, FX, and credit lead.
• Reading the composite as a hard rule rather than a multiplier. Composite -2 in a regime-matched trade still gets taken, just smaller.
• Confusing absolute yield levels with rates regime. The slope and trend matter more than the level.
• Ignoring the FX layer because it "does not affect my book." DXY is a global liquidity gauge that affects all dollar-denominated multiples.
• Acting on credit dissent in isolation. Credit is the strongest leading indicator but still benefits from concordance with rates or FX.
• Updating the composite intraday. The composite is a daily-close instrument; intraday updates introduce noise without adding signal.