Stable Behavior Forecasting
Forecast recurring reactions instead of unpredictable events
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 95%
Stable Behavior Forecasting separates unpredictable events from the more consistent ways people respond to them. Recessions, wars, bear markets, and crises differ in their causes and details, making their timing and form difficult to forecast. Fear, panic, risk aversion, storytelling, and eventual adaptation are more persistent. The method starts with an uncertain situation, studies analogous periods, and identifies reactions that recur despite changing circumstances. Decisions are then designed around those behavioral patterns rather than an exact prediction. This does not reveal when the next disruption will arrive, but it can improve preparation for how customers, investors, employees, or competitors are likely to behave once it does.
Origin
Morgan Housel described the model while explaining why recurring human behavior offers a firmer basis for forecasting than unique historical events. Extracted from Masters in Business.
Core principles
- 01Events change, but human reactions remain comparatively stable
- 02Forecast what is repeatable rather than what is merely imaginable
- 03Use history to identify behavioral patterns, not exact future events
- 04Prepare for recurring responses without pretending to know their timing
How to run it
- 1
Name the uncertain event
State the recession, market shift, competitive threat, or other event you are tempted to forecast. Acknowledge which details and timing cannot be known reliably.
Pro tip Phrase the uncertainty as a range of scenarios rather than one prediction.
Watch out Do not disguise a preferred outcome as a forecast.
- 2
Find historical analogues
Review several situations that produced similar uncertainty, even if their causes differed. Look across multiple periods to avoid overfitting to the latest example.
Pro tip Include both severe and moderate cases.
Watch out Do not assume an analogue will repeat event-for-event.
- 3
Extract stable reactions
Identify behaviors that recurred across the examples, such as panic selling, defensive spending, narrative shifts, or excessive optimism.
Pro tip Prioritize reactions appearing across different eras and contexts.
Watch out Separate observed behavior from your interpretation of why it occurred.
- 4
Design around behavior
Build safeguards and opportunities around the likely reactions. Decide what you will do if others become fearful, euphoric, impatient, or risk-averse.
Pro tip Write trigger-based responses before emotions intensify.
Watch out Avoid relying on a single behavioral response.
- 5
Stay flexible on specifics
Update event-specific assumptions as new evidence arrives while retaining useful long-term behavioral expectations.
Pro tip Review which assumptions concern human nature and which concern temporary conditions.
Watch out Confidence in behavior does not justify confidence in timing.
In the wild
An investor cannot reliably predict the next bear market's date or cause. Instead, she studies past declines and expects fear, alarming narratives, forced selling, and pressure to abandon long-term plans. She maintains liquidity, sets rebalancing rules, and documents why panic-driven decisions are usually costly before the decline occurs.
→ The investor is prepared to respond consistently without needing to predict the trigger.
A business leader cannot identify the next economic shock but expects customers to delay discretionary purchases and employees to become anxious when uncertainty rises. The company preserves cash, creates a customer-retention playbook, and prepares transparent internal communications.
→ The organization responds faster because it anticipated behavior rather than the exact crisis.
Common mistakes
Predicting the costume, not the behavior
People focus on the unique cause and appearance of the next crisis while ignoring reactions that recur across many crises.
Treating history as a precise script
Historical patterns illuminate behavior, but they do not provide exact dates, prices, or sequences.
Using only the latest crisis
A single analogue can produce false confidence because its distinctive details may not recur.
Is it for you?
Best for
Investors, leaders, and planners making decisions under uncertainty or preparing for market disruptions.
Not ideal for
Situations requiring precise short-term timing, exact prices, or event-specific operational forecasts.
From the transcript
“history never repeats itself, but man always does.”
“the events never repeat themselves.”
“the behaviors are stable over time,”
From the episode
At the Money: What Never Changes with Money