A sovereign wealth fund sets five-year CMEs. Six months later, every number is wrong, labor force surprised upward, rates stayed higher, the equity risk premium compressed 150 bps. The asset allocation built on those stale expectations is now positioned for a world that does not exist.
Capital market expectations (return, risk, and correlation forecasts for each asset class) feed every optimization. Small changes in expected returns (even 50 bps) can shift optimal allocations dramatically because the optimization surface is flat near the top. The quality of the CME process constrains every downstream investment decision.
HIGH-FREQUENCY: The exam tests the framework for forming CMEs, not your ability to produce a specific forecast. Know the steps, pitfalls, and tools.
- Data limitations. GDP data is revised, employment figures get benchmark revisions, inflation indexes change composition
- Measurement errors and biases, survivorship bias inflates historical equity returns; appraisal smoothing understates PE and real estate volatility
- Historical estimates describe regimes that may not recur. Bretton Woods, 1970s inflation, post-GFC zero rates
Common mistakes
- Using arithmetic mean for multi-period wealth projections. Geometric mean = arithmetic mean minus one-half variance. The arithmetic mean is correct for single-period optimization; the geometric mean is correct for compound growth and reporting.
- Anchoring EM equity returns to the historical 20-year average. Survivorship bias, favorable sample periods, and one-time repricing inflate historical EM returns. Forward-looking model-based estimates are lower and more reliable.
- Double-counting inflation in Grinold-Kroner. Nominal earnings growth already includes inflation. Adding an inflation term separately overstates the expected return.
Bottom line
- CMEs are the most consequential and most error-prone input to asset allocation.
- Nine recurring challenges include data limitations, survivorship bias, non-stationarity, data mining, and anchoring.
- GDP decomposition (labor force growth plus productivity growth) anchors long-run equity return estimates.
- Grinold-Kroner: expected equity return = dividend yield + buyback yield + nominal earnings growth + repricing.
Exam shortcut
When the vignette gives you historical returns and asks you to form CMEs, the answer is almost never "use the historical average." The exam rewards candidates who adjust for biases and use model-based approaches. For Grinold-Kroner, the three components to remember are income (yield + buybacks), growth (nominal earnings), and repricing (P/E change). Repricing averages near zero over long horizons, so sustainable equity return is roughly yield plus growth.
The full lesson (about 5,078 words, 34 min read) adds 2 worked examples, all 12 common mistakes, a self-check, free in the app.
Learning objectives
- cme part 1
Browse all free CFA L3 Portfolio Mgmt lessons or jump into free CFA L3 Portfolio Mgmt practice questions.