Seven-Year Mean-Reversion Forecast
Estimate returns by projecting income, growth, and normalization over seven years
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 97%
Begin with the premise that markets spend little time at equilibrium but pass through normal conditions often enough to make normalization useful. Select a long horizon, such as seven years, and estimate three return components: income received, sustainable fundamental growth, and the gain or loss caused by valuation and profitability reverting toward normal. Because equilibrium itself is uncertain, calculate more than one scenario. A traditional scenario can use historically normal valuation multiples and real bond yields, while a partial-mean-reversion scenario can assume permanently lower interest rates and therefore higher allowable valuations. The output is not a point prediction of where markets will trade on a particular date. It is a consistent set of prospective real returns that helps an allocator compare assets, expose assumption sensitivity, and identify markets whose current prices require unusually optimistic outcomes.
Origin
Ben Inker described GMO's forecast methodology, first published in the mid-1990s, on Masters in Business.
Core principles
- 01Markets rarely look normal but revisit normal conditions over time
- 02Expected return combines income, fundamental growth, and valuation change
- 03Profitability and valuation should be normalized separately
- 04Equilibrium assumptions must reflect plausible interest-rate regimes
- 05Forecasts are scenarios rather than precise predictions
How to run it
- 1
Set the horizon
Choose a horizon long enough for abnormal conditions to plausibly normalize. GMO uses seven years as an explicitly uncertain future point rather than a precise timing claim.
Pro tip Apply the same horizon to every asset being compared.
Watch out Do not interpret the horizon as a forecast of the exact turning date.
- 2
Define normal conditions
Specify normalized valuation, profitability, and real interest rates. Make each equilibrium assumption explicit and economically defensible.
Pro tip Use long-run evidence rather than the latest market regime alone.
Watch out A single historical average may be inappropriate after a structural change.
- 3
Estimate income and growth
Project the dividends, coupons, or other income the asset should deliver, then add sustainable earnings or cash-flow growth.
Pro tip Separate fundamental growth from changes in valuation multiples.
Watch out Avoid extrapolating unusually high recent growth indefinitely.
- 4
Calculate normalization effects
Estimate the gain or loss produced as current valuation and profitability move toward their assumed normal levels.
Pro tip Show valuation and profitability reversion as separate return contributions.
Watch out Do not conceal a weak fundamental forecast behind an assumed multiple expansion.
- 5
Run alternative scenarios
Repeat the calculation under another plausible equilibrium, such as permanently lower interest rates with higher fair valuations.
Pro tip Focus on whether the asset remains attractive across scenarios.
Watch out Do not label only the preferred scenario as reasonable.
- 6
Rank prospective returns
Combine income, growth, and normalization for each scenario, convert the results to real returns, and compare opportunities across assets.
Pro tip Use return ranges when assumptions materially affect the ranking.
Watch out Expected-return rankings do not remove implementation or path risk.
In the wild
An allocator estimates income and sustainable earnings growth for both markets, then calculates how current valuations would change under traditional and permanently low-rate equilibria. US equities suffer a normalization loss in both scenarios, while emerging markets remain near fair value in the low-rate case. The allocator does not predict an immediate US decline; instead, the forecast identifies the market with the stronger prospective real return over the common horizon.
→ Capital is tilted toward the market with better scenario-resistant expected returns.
Common mistakes
Treating seven years as a deadline
The horizon is a modeling convention for an uncertain future normalization point, not a promise that reversion will occur on schedule.
Using only one equilibrium regime
A single normal-value assumption can make the forecast falsely precise when interest rates or other structural conditions may have changed.
Extrapolating recent profitability
Current margins and growth can be cyclical or exceptional, so projecting them unchanged distorts prospective returns.
Is it for you?
Best for
It is best for long-horizon investors comparing broad markets or asset classes under uncertain valuations and interest rates.
Not ideal for
It is not ideal for short-term market timing or assets whose value cannot be anchored to future cash flows.
From the transcript
“Everything looks normal, So the p of the market looks normal, profitability looks normal UM”
“the return we're going to get between now and then is going to be driven by whatever earnings growth will occur, whatever income we're going…”
“We now call it partial mean reversion.”
From the episode
Ben Inker on Value and Asset Allocation (Podcast)
Ben Inker