With the U.S. TR‑CAPE hovering around 40, this exercise investigates whether systematic factor tilts reshape the valuation–return nexus for long‑horizon, real, annualised returns.
The analysis uses the S&P 500 TR‑CAPE as the valuation anchor and U.S. factor indices (MSCI USA Momentum, MSCI USA Value) for the factor leg. Series are converted to real, annualised total returns by deflating with U.S. CPI and compounding at the appropriate horizon; higher TERs typically associated with factor implementations are incorporated so that results approximate net‑of‑fee realisations.
The empirical design links the TR‑CAPE observed at time t to subsequent 15‑year real CAGR from t to t + 15 using overlapping rolling windows, enabling a comparison of factor payoffs conditional on the valuation. The perspective is explicitly market‑relative: the analysis targets differential performance versus the U.S. benchmark (beta) and quantifies how index construction – reconstitution rules, turnover, and selection mechanics – propagates into realised factor premia under high‑valuation conditions.
- The downward‑sloping fitted lines are consistent with the long‑documented inverse link between valuation ratios like CAPE and long‑horizon equity returns, where higher CAPE implies a lower cyclically‑adjusted earnings yield CAEY = 1 / CAPE and thus lower expected real returns over 10–20 years.
- TR CAPE is a total‑return variant of Shiller’s CAPE that reinvests dividends into the price index and rescales earnings, reducing biases from changing payout policies when mapping valuations to future returns.
- The broad vertical scatter around the trendlines highlights substantial uncertainty around point forecasts; CAPE provides useful average information for long horizons but admits a wide range of outcomes at any given starting valuation.
- The Momentum series reflects an MSCI implementation that selects the top MSCI USA constituents by a momentum score and weights them by market‑cap times momentum, with caps and quarterly reviews, so its cloud represents a momentum‑tilted subset of the parent index.
- The Value series represents large‑ and mid‑cap U.S. equities exhibiting value characteristics defined by book‑to‑price, 12‑month forward earnings‑to‑price, and dividend yield, which can shift its valuation/return coordinates relative to the parent index.
- From an asset‑allocation perspective, the plot aligns with evidence that valuation level is a primary driver of decade‑scale expected returns, while factor tilts (e.g., momentum or value selection rules) modify the distribution of realised outcomes around that baseline.
- The relation is informative for strategic horizons rather than tactical timing; extrapolation outside the observed TR CAPE range or for short‑term horizon is not supported by the empirical literature on CAPE‑style indicators.
- The momentum index exhibits consistently higher returns across all valuation regimes, reflecting the persistent nature of the momentum premium documented in academic literature, which stems primarily from behavioral biases rather than risk compensation.
- The momentum factor shows less sensitivity to valuation levels because it operates on shorter horizons (typically 3-12 months) while CAPE predicts returns over 10-15 year periods, creating temporal mismatch between the momentum signal and valuation-based mean reversion.
- The MSCI USA Value index is defined by book‑to‑price, forward earnings‑to‑price, and dividend yield, which induces structural tilts toward what is cheap by accounting ratios at each review; such accounting‑based tilts can lag when early‑cycle leadership is elsewhere, yielding weaker results even when the aggregate market is cheap.
- The MSCI USA Momentum matching MSCI USA at low valuations is also consistent with the MSCI Momentum index being a long‑only subset of the MSCI USA with weights linked to parent capitalisation and capped at the issuer level, preserving substantial market exposure when trends are broadly synchronised.
- At high valuations, momentum’s design selects the strongest risk‑adjusted price trends (6–12 month horizon) from the MSCI USA parent and implements sizable but controlled turnover, allowing faster migration toward the few persistent winners that typically drive returns when overall market expected returns are muted.
CAPE‑based predictability is statistical rather than mechanical and can be affected by structural breaks across countries and periods, reinforcing that conditional factor payoffs in any CAPE bucket will depend on the particular historical window underpinning the scatter.