Stock-Bond Indicator 2.0: A Transparent, Adaptive Signal for Tactical Equity-Bond Allocation

Key Conclusion

The Morgan Stanley Global Investment Committee's rebuilt Stock-Bond Indicator (SB 2.0) delivers a more consistent, transparent tactical allocation signal than its predecessor (SB 1.0) by adopting dynamic weighting, a broader indicator pool (30-60 inputs), and a clear three-month forward forecast window. Backtest evidence shows SB 2.0 significantly outperforms SB 1.0 in recent out-of-sample periods (last 5 and 10 years), where the static version’s effectiveness decayed. The dynamic-weighting mechanism prevents persistent underperformance across market regimes and yields stable hit rates (directional accuracy) in all three tested decades (2000-2009, 2010-2019, 2020-2025). Market participants may underestimate the value of adaptive weighting to capture evolving market structures; static historical relationships that worked in 2000-2010 have weakened since 2010, and SB 2.0’s construction explicitly addresses this fragility.

---

Evidence Chain

1. SB 2.0 Outperforms SB 1.0 in Recent Out-of-Sample Periods with More Stable Directional Accuracy

• Conclusion: SB 2.0’s performance advantage is concentrated in the last 5–10 years, where SB 1.0’s static assumptions broke down; SB 2.0 shows uniform hit rates across all three decades, whereas SB 1.0’s directional accuracy declined sharply after 2010.

• Evidence: In the 2000–2009 period, SB 1.0’s hit rate was comparable to SB 2.0. In 2010–2019 and 2020–2025, SB 1.0’s directional agreement fell materially (Exhibit 10 from the evidence pack). SB 2.0 maintained stable hit rates across all periods. In a direct head-to-head comparison of cumulative returns (Exhibit 8), SB 2.0 outperformed SB 1.0 in both 5- and 10-year trailing windows.

• Investment Implication: Investors relying on static indicators run the risk of using a tool that was calibrated for an earlier regime. SB 2.0’s multi-decade stability suggests it is better suited for ongoing tactical use, especially given that market structures (e.g., PMI–equity return relationships) have shifted.

2. Dynamic Weighting Avoids Long-Term Underperformance and Improves Regime Consistency

• Conclusion: Dynamic weighting—applied at both the indicator and category level with upper/lower bounds—provides modest long-run return improvement over equal weighting, but more importantly yields consistent results across different market environments without any sustained periods of underperformance.

• Evidence: Comparison of dynamic-weighted vs. equal-weighted versions (Exhibit 9) shows that dynamic weighting prevents persistent drawdowns relative to the equal-weight baseline. The bounds prevent any single indicator or category (e.g., valuation, sentiment) from dominating the signal based on short-term efficacy. The category-level dynamic weights also allow the model to shift emphasis from, say, macro growth indicators to liquidity measures as conditions change.

• Investment Implication: The added design complexity is justified by the reduction in regime sensitivity. For a tactical allocation signal meant to be applied systematically, avoiding long-term drift or prolonged periods of negative relative performance is critical for user confidence and portfolio stability.

3. Signal Behavior During Key Market Stress and Recovery Periods Matches Tactical Intent

• Conclusion: SB 2.0 turned underweight equities during the 2008 financial crisis and the 2022 bear market, and overweight during the 2009–2010 recovery. Portfolio simulations applying the signal (with a ±40% maximum deviation around a 60/40 benchmark) generated competitive cumulative and risk-adjusted returns, placing above the static 60/40 on the efficient frontier.

• Evidence: Exact signal states at these turning points (Exhibit 11) match the intuitive tactical call: low risk appetite when forward excess returns are most negative, high risk appetite when they improve. The month-by-month portfolio simulation (Exhibits 12–13) shows that a simple rebalancing rule based on SB 2.0—adjusting equity exposure between 20% and 100%—outperformed the static 60/40 in total return and Sharpe ratio over the full 2000–2025 sample.

• Investment Implication: The indicator passes the “does it make sense at extremes” test. This behavioral validity supports its use as a systematic cross-check for qualitative macro views. When the signal diverges from a team’s conviction, it flags a risk that may not be fully priced into current positions.

---

Key Risks

1. Unstable relationships: Macro and market indicators may change their predictive relationship with forward stock–bond excess returns, even with dynamic weighting. Periods of decay should be expected as market structure evolves.
2. Implementation sensitivity: The signal depends on choices of normalization windows, lookback periods, drift constraints, and threshold mapping. While these have been tested, they may require recalibration if hit rates deteriorate or category-level attribution becomes inconsistent.
3. Not a substitute for fundamentals: SB 2.0 is a quantitative starting point for tactical adjustments, not a standalone recommendation. It cannot replace bottom-up analysis, qualitative judgment, or risk management.
4. Backtest limitations: Historical performance does not guarantee future results. Out-of-period behavior could differ.

---

Trading Implications

• Signal output: Overweight, neutral, or underweight equities relative to bonds on a weekly basis.
• Application: Tactical deviation around a strategic benchmark (e.g., 60/40) with a maximum ±40% adjustment, implying an equity allocation range of 20%–100%.
• Use cases: (a) Around-benchmark tactical rebalancing, (b) new-money deployment pacing, (c) cross-check against macro views to highlight risks not yet reflected in portfolio weights.
• Relevant horizon: The 3-month forecast window is the primary target, but the framework’s insights extend to the 1–12 month range.

---

Appendix: Methodological Framework Summary

Indicator Category Framework (illustrative, not exhaustive): Profitability, Macro Growth, Inflation & Prices, Rates & Liquidity, Relative Valuation, Sentiment & Technicals, Positioning.