Prepare, Don't Predict Portfolio
Build for several plausible futures instead of betting on one forecast.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 96%
The Prepare, Don't Predict approach begins by accepting that investors must act on the future without being able to know it. Instead of selecting one forecast and constructing the theoretically perfect portfolio for that outcome, investors map several plausible environments and seek a portfolio that can perform reasonably across them. This sacrifices maximum upside in any single scenario in exchange for avoiding catastrophic dependence on one prediction. Decisions rely on present evidence—valuations, yields, risk premiums, balance sheets, and investor behavior—rather than precise calls about recessions or market direction. The result is a resilient portfolio whose exposures are deliberately diversified and whose potential losses remain tolerable even when the investor's preferred scenario does not occur.
Origin
Extracted from Masters in Business, where Howard Marks contrasted forecast-dependent optimization with portfolios designed to withstand a variety of possible futures.
Core principles
- 01The future cannot be known reliably.
- 02A perfect portfolio for one forecast is fragile.
- 03Present observations are more dependable than predictions.
- 04Resilience across outcomes matters more than optimization for one outcome.
How to run it
- 1
Admit the forecast limit
State explicitly which future variables cannot be known with dependable precision. Avoid disguising opinions as facts.
Pro tip Record confidence ranges rather than point forecasts.
Watch out Confidence and accuracy are not the same thing.
- 2
Map plausible environments
Define several materially different scenarios, such as growth, recession, inflation, disinflation, calm, or crisis.
Pro tip Include at least one scenario that contradicts the prevailing consensus.
Watch out Do not create minor variations of the same base case.
- 3
Test portfolio behavior
Assess how each position and the total portfolio could perform under every scenario. Look for hidden concentrations and correlated failure modes.
Pro tip Focus on potential impairment, not only short-term volatility.
Watch out Historical correlations may break during stress.
- 4
Build broad resilience
Combine exposures so the portfolio can do well in several environments and avoid doing badly in any plausible one.
Pro tip Balance return opportunities with liquidity and defensive capacity.
Watch out Do not optimize so aggressively for resilience that no meaningful return remains.
- 5
Adjust from present evidence
Use current prices, yields, fundamentals, and sentiment to tilt exposure without claiming to know exactly what happens next.
Pro tip Make larger adjustments when present conditions are extreme.
Watch out Do not let a tactical tilt become an unacknowledged all-or-nothing forecast.
In the wild
An institutional investor cannot know whether inflation will persist or growth will collapse. Rather than concentrating in the assets favored by one forecast, it holds discounted credit with strong coverage, liquid reserves, selective equities, and inflation-sensitive assets, sizing each so no plausible macro path threatens solvency.
→ The portfolio gives up the maximum return from a perfect forecast but remains viable across divergent outcomes.
Common mistakes
Constructing for one certain future
A portfolio optimized for one forecast can fail badly when events take another path.
Treating diversification as random variety
Owning many assets does not create resilience if they share the same underlying risk.
Ignoring current price
Scenario resilience still depends on whether each asset offers suitable compensation at today's valuation.
Is it for you?
Best for
Long-term investors managing capital through uncertain economic, market, and geopolitical conditions.
Not ideal for
Strategies explicitly designed to make concentrated, short-duration bets on a forecasted event.
From the transcript
“You can't predict, you can only prepare.”
“I don't believe the future is predictable, you know.”
“I try to put together portfolios that will do well in a variety of environments, but without trying to make a prediction about here's what's…”
From the episode
Interview with Howard Marks: Masters in Business (Audio)