The Behavior Gap Ch. 5: Designing Against Yourself
阅读中文版 (with Audio)Every defence that depends on willpower during the drawdown has already failed. The ones that work were installed years earlier and require no decision at the moment of maximum pressure.
🔊 Listen to Article (Chinese Audio)
The Behavior Gap Ch. 5: Designing Against Yourself
Investment Background
Four chapters have established that the gap is real, measurable, directional, and concentrated in a few moments when judgment is least available. This chapter is about what actually defends against it.
The defence has one binding requirement, and it eliminates most of the advice normally offered: it must not require a decision at the moment of maximum pressure. Any defence whose activation step is "remember your principles and stay calm" has placed its load-bearing element exactly where chapter 4 showed the structure gives way.
This is why the chapter is called designing against yourself rather than disciplining yourself. The target is not a better operator. It is a plan that demands less of the operator, on the specific days when there is least to draw on.
The Wall Street Translation
The Test Every Defence Must Pass
One question separates defences that work from defences that sound good:
Does this require me to do something, at the worst moment, that I will not want to do?
If yes, it is not a defence. It is a hope with a procedure attached.
| Defence | Requires action during the drawdown? | Holds? |
|---|---|---|
| "Stay the course" as a principle | yes — active refusal, repeatedly | ✗ |
| Automatic contributions already running | no — continues by default | ✓ |
| A rule to rebalance on a fixed date | minimal, and mechanical | ✓ |
| Checking the portfolio less often | no — a setting, not a decision | ✓ |
| "Re-enter when things stabilise" | yes — a judgment call, at the worst time | ✗ |
| Written policy statement, signed and dated | no — it was written years earlier | ✓ |
| Resolving to be more disciplined | yes — entirely | ✗ |
The pattern in the working row is consistent: the decision was made once, in advance, by someone calm, and the drawdown does not present an occasion to revisit it. misbehaving chapter 4 owns this principle in general form. This chapter applies it to the one decision that costs the most.
The Four That Actually Work
Automatic contributions, established and left alone. The single highest-value defence in this book. Contribution continues without any decision being made, which means the most valuable purchases of the entire plan — the ones made during the decline — happen by default rather than by resolve. The critical property is that stopping requires an action. Inertia, which works against the investor in most contexts, is finally working for them.
Reduced observation frequency. Checking a portfolio daily produces many more opportunities to experience a decline than checking quarterly, and each observation is an occasion for a decision. This is not ignorance — the annual review still happens. It is the removal of the daily prompt. your-money-and-your-brain chapter 6 owns environment design in general; this is its highest-value single application.
A written investment policy, dated and signed. One page, written while calm, stating the allocation, the contribution schedule, the rebalancing rule, and — most importantly — what would and would not justify a change. Its value is not that it is legally binding. It is that it proves to the person in the drawdown that a calmer, better-informed version of themselves considered this exact scenario and reached a different conclusion. thinking-in-bets-duke chapter 5 owns the decision journal; this is its narrow application to a single recurring decision.
Mechanical rebalancing on a date, not on a judgment. Rebalancing on a fixed calendar sells what has risen and buys what has fallen without requiring any view about which should be doing what. The mechanism does the countercyclical thing automatically. The word "mechanical" is load-bearing — a rebalancing rule with discretion attached is a forecast wearing a rule's clothing.
What Does Not Work, and Why It Is Still Recommended
Three popular defences fail the test, and it is worth being specific about why, because all three sound responsible.
"Educate yourself about behavioural biases." misbehaving chapter 5 owns the finding directly: knowing about a bias does not disarm it. Education produces an investor who recognises what they are doing while they do it. That is genuinely better than nothing — but recognition is not prevention, and treating a reading list as a defence leaves the structural work undone.
"Have the discipline to stay the course." This is the conclusion, restated as though it were the method. It names the required outcome and supplies nothing that produces it. Every investor who sold at the bottom intended to stay the course.
"Work with an advisor who will talk you out of it." This one is different — it can genuinely work, and is the strongest of the three — because it is external and does not depend on the investor's own state. The caution is that it introduces the party customers-yachts-schwed spends six chapters on, and an advisor compensated on transactions or assets has interests that may or may not align at that exact moment. The defence is real; the incentive structure has to be checked first.
Division of Labor With the Rest of the Library
| Question | Book that owns it |
|---|---|
| How do defaults and nudges bind a future self? | misbehaving ch04, ch06 |
| How do I design an environment around my biology? | your-money-and-your-brain ch06 |
| How do I record my reasoning before the outcome? | thinking-in-bets-duke ch05 |
| Why can a system with an override not be a system? | man-who-solved-the-market-simons ch05 |
| What are an advisor's incentives? | customers-yachts-schwed |
| Which defences survive the moment they are needed? | This book ch05 |
Executable Trading Rules
- Automate contributions so that continuing is the default and stopping requires a deliberate action. This is the highest-value item in the book and it is a one-afternoon task.
- Reduce how often you look. Quarterly is sufficient for a long-horizon core. Each additional observation is an additional opportunity to act.
- Write a one-page policy statement, sign it, date it, and store it where you will find it in a crisis. State explicitly what would justify a change — and note that a large decline is not on that list.
- Rebalance on a date, not on a view. Put the date in the calendar in advance and let the mechanism be countercyclical on your behalf.
- Apply one test to any defence you are considering: does it require me to act, at the worst moment, against what I will want? If so, replace it with one that does not.
Relevance to a Retirement Portfolio
Everything in this chapter is more valuable the longer the horizon, which makes it most valuable to the reader who has just started and least reversible for the reader who has not yet installed it.
For the reader this site is written for — holding a low-cost, diversified core — this chapter is the whole practical payload of the book. The core is already correct. Nothing here proposes changing what is held. Every recommendation is about the machinery around the holding: how contributions arrive, how often it is observed, what is written down, and when it is rebalanced.
None of these are investment decisions in the ordinary sense, which is exactly why they are effective. They are the parts of the plan that keep working when the operator is at their least reliable — which chapter 4 established is the moment that determines what the whole plan actually returns.