Worth and Wrong Valuation Process
Estimate intrinsic value, then identify exactly where your estimate can fail.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 97%
The process begins by estimating what an asset is worth rather than predicting its next price movement. For equities, GMO combines valuation measures with growth and income, while examining returns on capital and the likelihood of mean reversion. The investor then performs a second, equally important task: identifying where the estimate may be wrong. That means testing whether margins, required returns, economic regimes, or other assumptions have structurally changed. The gap between price and estimated value provides the opportunity, while the error analysis determines confidence and position size. The method therefore converts valuation from a single-point forecast into a disciplined decision process that explicitly joins expected return with risk management.
Origin
Extracted from Masters in Business, where Matthew Kadnar summarized GMO's long-standing valuation and asset-allocation process.
Core principles
- 01Price determines most long-term investment outcomes.
- 02No asset is inherently attractive regardless of price.
- 03Short-term market prediction is unreliable.
- 04Every valuation must include an explicit error analysis.
- 05Confidence should vary with the quality of available evidence.
How to run it
- 1
Define the asset and horizon
Specify exactly what is being valued and choose a horizon long enough for fundamentals and mean reversion to matter.
Pro tip Use a consistent horizon when comparing opportunities.
Watch out A long-term valuation model cannot reliably predict near-term price movements.
- 2
Estimate fundamental worth
Combine relevant valuation proxies with sustainable growth and income. For equities, examine returns on capital rather than relying on one superficial multiple.
Pro tip Use several economically defensible proxies instead of depending on a single ratio.
Watch out A metric's historical mean reversion does not prove that it must mean revert.
- 3
Compare price with value
Measure the gap between the current market price and the estimated fundamental value. Require a larger gap when the model or data is less reliable.
Pro tip Rank assets on both expected return and confidence.
Watch out Cheap assets can remain cheap for extended periods.
- 4
Ask where you can be wrong
Challenge the assumptions behind profitability, discount rates, growth, inflation, and mean reversion. Consider whether a structural regime change could invalidate historical relationships.
Pro tip Invite experienced colleagues with different perspectives to attack the thesis.
Watch out Do not use 'different this time' as either an automatic dismissal or an automatic justification.
- 5
Size for uncertainty
Translate expected return and model confidence into a portfolio weight. Reduce exposure when the opportunity looks attractive but the forecasting evidence is weak.
Pro tip Separate confidence in valuation from confidence in timing.
Watch out A strong expected return does not justify unlimited concentration.
- 6
Review without chasing noise
Update the valuation when fundamentals or assumptions change, while resisting reactions to ordinary short-term volatility.
Pro tip Document the original assumptions so revisions are auditable.
Watch out Changing the model merely to explain current prices destroys its decision value.
In the wild
An investment committee estimates returns using normalized returns on capital, valuation multiples, growth, and income. It then tests a lower-rate regime in which required returns remain permanently depressed. Both scenarios still imply poor prospective returns, so the committee reduces exposure rather than relying on a bullish short-term narrative.
→ The portfolio decision remains robust across more than one plausible economic regime.
An investor finds a market statistically cheap but recognizes that current earnings sit near a cyclical peak. The investor normalizes profitability, tests weaker margins and currency depreciation, and takes a moderate position instead of treating the headline multiple as definitive.
→ The position captures potential upside while limiting damage from an overly optimistic earnings assumption.
Common mistakes
Treating one multiple as truth
A single ratio may be distorted by accounting, margins, or structural change. Use multiple defensible measures and understand their mechanisms.
Ignoring model failure
A valuation without an explicit account of where it can be wrong creates false precision and encourages excessive position sizes.
Using valuation as a timing signal
An asset can move farther away from fair value before reverting, so valuation should guide expected return and risk rather than precise market timing.
Is it for you?
Best for
It is best for long-term investors evaluating equities, bonds, currencies, or other assets with estimable fundamental value.
Not ideal for
It is not ideal for short-term trading strategies driven primarily by momentum, order flow, or market timing.
From the transcript
“figure out what you think something's worth and where you where you can be wrong.”
“valuation is going to determine the vast bulk of your outcomes.”
“your starting valuation, the price you pay for an asset is really going to really determine the vast bulk of your outcomes.”
From the episode
Matthew Kadnar Talks About His Transition From Law to Finance
Matthew Kadnar