Home/Systematic Trading/Return Expectations and Capital Market Assumptions

Return Expectations and Capital Market Assumptions

portfolio-constructionLevel 2 — Intermediate

Prerequisites

Edge IdentificationRegime Detection

What It Is

Return expectations is the discipline of constructing forward-looking estimates of asset class returns using earnings yields, yield spreads, and valuation measures — forming the foundation for strategic asset allocation and cross-asset comparisons.

Correct Execution

  • Use cyclically adjusted earnings yield (CAPE inverse) as the primary long-term equity return predictor, adjusted for payout ratios and structural changes in business composition
  • Compare expected returns across asset classes on the same basis: real arithmetic returns, not nominal geometric
  • Account for the distinction between rational risk-based and behavioral/irrational explanations for high valuations — survey data on investor expectations helps adjudicate
  • Cross-sectional comparison (US vs non-US) is more reliable than time-series CAPE comparison because structural changes affect levels but not relative rankings as severely
  • Recognize that return prediction is not market timing — it informs long-run portfolio construction, not monthly rebalancing

Progression Levels

Diagnostic Tree

Coaching Cues

  • "Simplicity is a merit. Weigh the benefit of simplicity against the potential improvement from complexity — and always guard against backfit bias when adding complexity." — Victor Haghani, FWM S7E13
  • "Cape was never meant to be a regression variable against 5-year returns. It was designed for 10-year horizons. Don't ask it to do something it wasn't designed for." — Antti Ilmanen, FWM S7E21
  • "Put your hand over the y-axis. The absolute level doesn't matter. The relative differences are what the optimizer cares about." — Jim Masturzo, FWM S3E4

Common Errors

  1. Using CAPE as a short-term market timing tool: CAPE forecasts 10-year returns, not next quarter → Never use CAPE as a monthly signal; use it for strategic allocation decisions only
  2. Not adjusting for payout ratio when using earnings yield: Companies retaining earnings will grow faster; raw earnings yield understates expected return → Add payout ratio adjustment as standard practice
  3. Not communicating uncertainty: A 3% expected return for US equities could easily be 1% or 5% → Always present expected returns as ranges, not point estimates; communicate the sources of uncertainty explicitly

Edges

⚡ Conventional Wisdom Is Wrong

Cross-Sectional Valuation Comparisons Beat Time-Series Comparisons — Every Time

portfolio-constructionreturn-expectations →

Return expectations frameworks almost universally compare current CAPE or earnings yield to a long historical average (time-series comparison). This approach fails when structural factors shift the equilibrium — as happened when the US equity market became technology-heavy post-1990. Cross-sectional comparison (US earnings yield vs European earnings yield vs Japanese earnings yield, standardized for industry composition) is more robust because it controls for structural change — both markets face the same secular forces simultaneously. The residual spread after industry standardization is genuine valuation differential, not noise from comparing apples to historical oranges.

What most people do
Compute current CAPE; compare to 30-year average CAPE; conclude "market is expensive/cheap" based on deviation from the historical average.
What the best do
Standardize all markets to the same industry composition; compare earnings yields cross-sectionally across geographies; identify valuation differentials that survive the industry-neutral adjustment.
Why it's an edge: Cross-sectional comparisons generate actionable relative value trades (underweight US, overweight Europe or EM) that are more reliable than the absolute timing signals that time-series CAPE produces. Time-series CAPE has been bullishly wrong on non-US markets and bearishly wrong on the US for 15 years simultaneously.
How to exploit: For each major equity market (US, Europe, Japan, EM), compute the industry-composition-adjusted earnings yield by: (1) breaking the index into industry buckets; (2) reweighting each market to global industry average weights; (3) computing earnings yield on the reweighted index. The resulting cross-sectional spread between markets is a durable tactical allocation signal with a better track record than time-series CAPE.
Cross-domain parallel
In sports betting, relative line comparison (Team A implied probability vs league-average for similar game conditions) is more informative than absolute line comparison (Team A vs historical base rate), because the current market structure has already adjusted for many structural factors.
Victor Haghani, FWM S7E13, 2024-12-09; Antti Ilmanen, FWM S7E21, 2025-09-15
🔑 Hidden Causal Lever

Payout Ratio Adjustment Rescues the Earnings Yield Framework From Its Biggest Blind Spot

portfolio-constructionreturn-expectations →

The simple earnings yield framework assumes companies pay out all earnings and the investor receives them directly. In practice, companies retain a substantial fraction of earnings and reinvest them — producing future earnings growth. This means raw earnings yield understates expected return: a company with 5% earnings yield that retains 50% of earnings and reinvests at 15% ROE is generating additional future return that the raw yield doesn't capture. The payout ratio adjustment is not a refinement — it is the correction that makes the framework mechanically correct.

What most people do
Use earnings yield or CAPE inverse directly as the expected return forecast; omit the reinvestment return component.
What the best do
Adjust expected return upward by the retained earnings contribution: expected return = earnings yield + (retained earnings fraction × expected ROE on reinvested capital). Use this adjusted figure for all capital market assumption work.
Why it's an edge: Practitioners who apply the payout ratio adjustment get return expectations that are closer to realized outcomes. Practitioners who use raw earnings yield consistently underestimate equity returns — which biases them toward under-allocating to equities in TAA frameworks and produces worse investor outcomes.
How to exploit: For any equity market, compute: (1) current earnings yield; (2) current dividend payout ratio; (3) long-run ROE estimate. Compute adjusted expected return = earnings yield × payout ratio + earnings yield × (1 - payout ratio) × ROE / book value adjustment. Compare this to the unadjusted earnings yield. The difference is typically 100-200bps for major equity markets — enough to change a tactical allocation conclusion.
Victor Haghani, FWM S7E13, 2024-12-09 — "the machine keeps producing at the same rate — but if the company re-invests earnings, the machine gets bigger."
⚡ Conventional Wisdom Is Wrong

Return Expectations Are for Decades — Using Them at Shorter Horizons Destroys Their Value

portfolio-constructionreturn-expectations →

Earnings yield and CAPE are validated as 10-year return predictors with meaningful correlation (R-squared ~0.4-0.6). Their predictive power at 1-3 year horizons is near zero. Investors who use CAPE as a basis for quarterly or annual tactical decisions are using a 10-year forecasting instrument as a 1-year clock — the equivalent of using a geological map to predict tomorrow's weather. The misapplication of this tool is so common that CAPE's proponents spend as much time defending against misuse as promoting use.

What most people do
Cite current CAPE to justify reducing equity allocation; revisit the allocation quarterly; interpret any short-term underperformance or outperformance as validation/refutation of the CAPE signal.
What the best do
Use CAPE-based return expectations exclusively for strategic portfolio construction decisions (target equity allocation over 10+ year horizon); pair it with a momentum/trend signal for any decision at horizons under 3 years; never use CAPE alone for tactical moves.
Why it's an edge: Investors who correctly time-horizon-match their signals make better decisions at both horizons — they use CAPE for the right purpose and use other signals for shorter-term decisions rather than using CAPE for everything and getting poor results at both horizons.
How to exploit: Add a horizon test to any capital market assumption meeting: "Is this signal being used at its validated time horizon?" For any CAPE-derived conclusion, require that the resulting portfolio change be intended to remain in place for at least 5 years with minimal quarterly review. If someone wants to revisit quarterly, shift the conversation to the momentum signal which IS valid at shorter horizons.
Cross-domain parallel
In algorithmic trading, a 200-day moving average is not a 5-day signal. Using a long-period indicator as a short-term trigger produces a high number of false signals while reducing its structural usefulness. Horizon-matching signals to decisions is a universal principle.
Antti Ilmanen, "Understanding Return Expectations," FWM S7E21, 2025-09-15 — "CAPE was never meant to be a regression variable against 5-year returns. It was designed for 10-year horizons."

Sources

  • Antti Ilmanen, "Understanding Return Expectations," Flirting with Models S7E21, 2025-09-15 — CAPE critique and defense; structural changes in equity returns; survey data on investor expectations; equity bond premium analysis
  • Victor Haghani, "The Last of the Tactical Allocators," Flirting with Models S7E13, 2024-12-09 — payout ratio adjustment; industry composition adjustment; practical implementation of return expectations framework
  • Jim Masturzo, "Tactical Asset Allocation," Flirting with Models S3E4, 2021-04-10 — capital market assumptions construction; importance of relative vs absolute comparisons