Portfolio Construction Across Regimes
The 60/40 portfolio isn't broken — it just wasn't designed for every regime. Gold shines in cutting cycles (+2.57%/mo). Financials dominate pauses (+1.41%/mo). The right portfolio depends on where we are in the cycle.
The Regime-Aware Portfolio Framework
Static asset allocation assumes the future looks like the past average. But markets don't live in averages — they cycle through distinct regimes with dramatically different return profiles. The key to portfolio construction isn't finding the optimal static mix; it's understanding which mix works in each regime.
We analyze three Fed policy regimes (Hiking, Pause, Cutting) and three volatility regimes (Low, Normal, High) to build a framework for adaptive allocation.
Asset Returns by Fed Policy Regime
Monthly returns (%) for each asset class, 2005-2025
The Counterintuitive Finding
Cutting cycles are dangerous for equities. When the Fed cuts rates, SPY averages -0.73%/mo, XLF -2.68%/mo, XLE -2.22%/mo. The Fed cuts because something is wrong — and by then, the damage is already happening. Only gold (+2.57%/mo) and bonds (+0.98%/mo) provide shelter.
Complete Asset Class Performance by Fed Regime
| Asset | Cutting | Std Dev | Hiking | Std Dev | Pause | Std Dev | N (months) |
|---|---|---|---|---|---|---|---|
| GLD (Gold) | +2.57% | 5.76 | +1.23% | 4.38 | +0.66% | 4.72 | 251 |
| TLT (Long Bonds) | +0.98% | 5.57 | -0.14% | 3.48 | +0.33% | 3.73 | 251 |
| SPY (S&P 500) | -0.73% | 5.60 | +0.54% | 4.16 | +1.37% | 3.96 | 251 |
| IWM (Small Caps) | -1.28% | 7.78 | +0.28% | 5.53 | +1.36% | 5.27 | 251 |
| VNQ (REITs) | -2.02% | 9.76 | +0.48% | 5.11 | +1.36% | 5.56 | 251 |
| XLE (Energy) | -2.22% | 14.06 | +1.04% | 8.33 | +1.11% | 6.51 | 251 |
| XLF (Financials) | -2.68% | 9.50 | +0.32% | 4.87 | +1.41% | 5.60 | 251 |
Fed regimes: Cutting = rate decreased >0.25% in 3 months; Hiking = rate increased >0.25% in 3 months; Pause = stable
Diversification Fails When You Need It Most
SPY-TLT correlation varies dramatically by volatility regime
Cross-Asset Correlations by VIX Regime
| VIX Regime | VIX Range | SPY-TLT Corr | SPY-GLD Corr | Avg SPY | Avg TLT | Avg GLD | Days |
|---|---|---|---|---|---|---|---|
| Low Vol | VIX < 15 | -0.076 | +0.063 | +0.151% | -0.006% | +0.050% | 1,990 |
| Normal | 15 ≤ VIX < 25 | -0.264 | +0.062 | +0.054% | +0.013% | +0.048% | 2,434 |
| High Vol | VIX ≥ 25 | -0.420 | +0.055 | -0.206% | +0.081% | +0.052% | 868 |
Note: TLT provides best diversification in high vol (correlation -0.42) but loses diversification benefit in low vol (-0.08). Gold correlation stays near zero across all regimes.
Model Portfolios by Regime
Based on the regime analysis, here are three model portfolios optimized for different Fed policy environments. Each is designed to maximize risk-adjusted returns in its target regime.
Pause Portfolio
Current Regime
Optimized for stable Fed policy. Risk assets dominate.
Hiking Portfolio
Tightening Cycle
Balanced with commodity exposure. Avoid long bonds.
Cutting Portfolio
Easing Cycle
Defensive positioning. Gold and bonds dominate.
Portfolio Performance Comparison
Expected monthly returns by regime
Transition Signals: When to Rotate
The hardest part isn't knowing which portfolio to hold — it's knowing when to switch. Here are the signals that historically precede regime transitions:
Pause → Hiking Signals
- Core CPI accelerating above 3% for 2+ months
- Unemployment rate falling below 4%
- Fed dot plot shifting higher
- 2Y Treasury yield rising 50bp+ in 3 months
Action: Reduce TLT, increase GLD and XLE exposure
Pause → Cutting Signals
- Credit spreads widening 100bp+ in 3 months
- Initial claims rising above 250K
- ISM Manufacturing falling below 50
- Yield curve un-inverting after prolonged inversion
Action: Dramatically reduce equity exposure, maximize GLD and TLT
Cutting → Pause Signals
- ISM Manufacturing rising above 50
- Credit spreads tightening below historical median
- Leading Economic Index turning positive
- Fed language shifting to "data dependent"
Action: Begin rotating back to equities, add XLF exposure
Implementation Rules
Don't over-rotate
Regime changes don't happen overnight. Transition portfolios gradually over 2-4 weeks as signals confirm.
Keep a core allocation
Even in the Cutting portfolio, maintain some SPY exposure. Regimes can shift faster than you can react.
Gold is the universal hedge
GLD has positive expected returns in all three regimes. Maintain 10-35% allocation regardless of regime.
XLF is the regime amplifier
Financials swing from -2.68%/mo (Cutting) to +1.41%/mo (Pause). It's the highest-conviction sector bet.
TLT is regime-dependent
Long bonds lose money in Hiking (-0.14%/mo) but provide crucial diversification in Cutting. Size accordingly.
Current Positioning: January 2026
Regime: Pause
The Fed has been on hold for 3+ months. Inflation is moderating but above target. Employment remains solid. This is historically the best environment for risk assets.
Recommended Allocation:
Risks and Limitations
- Regime identification lag: You only know you're in a new regime after it's started. The first month of transition can cause whipsaw.
- Sample size: Only 32 months of Cutting data since 2005. The confidence intervals are wide.
- Correlation instability: The -0.42 SPY-TLT correlation in high vol is not guaranteed to persist.
- Transaction costs: Frequent rotation erodes returns. Only rotate when signals are strong and confirmed.
The Bottom Line
The 60/40 portfolio averages performance across all regimes. But you don't live in averages — you live in a specific regime. In Pause regimes (like now), financials and small caps shine. In Cutting regimes, only gold and bonds provide shelter. In Hiking regimes, commodities lead.
Match your portfolio to the regime. Hold the Pause portfolio while the Fed is stable. Watch for transition signals. When the regime shifts, be prepared to rotate — gradually but decisively.
Key insight: Financials swing 4.09 percentage points between Pause (+1.41%/mo) and Cutting (-2.68%/mo). That's the single biggest regime sensitivity in the asset universe. XLF allocation is your regime conviction bet.