Designing around yourself

The last chapter of the subject, and the only one that asks you to do something.

Why knowing is not enough

The uncomfortable finding running underneath these ten chapters is that the biases are found in people who can describe them. They appear in professionals, in high-IQ investors, in academics who study them. Reading chapter 3 will not stop you holding a loser.

The reason is that these are not errors of knowledge. In the moment, the loss-averse choice does not feel like a bias; it feels like patience. The overconfident trade feels like conviction. You do not experience a bias from the inside as a bias — you experience it as a reason.

So awareness is necessary and insufficient, and the useful move is to stop trying to be a better decision-maker in the moment and instead reduce how many decisions the moment contains.

The four mechanisms that work

Automation. A standing instruction that invests on a date regardless of circumstances removes the recurring decision entirely. It is the single most effective device available to a household investor, not because the timing is good but because there is no timing.

Pre-commitment. Decide the rule while calm, in writing, and make the writing specific enough to bind. "Rebalance when any holding is more than ten points from target" is a rule. "Rebalance periodically" is a wish.

Friction, correctly placed. Chapter 5's finding is that costs rather than selection drive underperformance, which means anything that reduces trade frequency has positive expected value. A mandatory waiting period between deciding and acting costs nothing when you are right and saves a great deal when you are not.

Defaults. Whatever happens if you do nothing should be the thing you would have chosen. Most people's defaults are set by whoever designed the product, which is a decision made by someone whose interests differ from yours.

What does not work

Resolving to be disciplined. Discipline is the thing in short supply at exactly the moment it is required.

Monitoring more closely. More frequent observation increases the number of moments in which a loss is salient, and salient losses drive action. Checking a long-term portfolio daily makes it harder to hold, not easier.

Keeping the rule in your head. An unwritten rule is revised silently in your favour, and you will not notice the revision.

Working the problem

Three behaviours and three rules.

1. Trading too much (chapter 5). No purchase or sale executes on the day it is decided. Write the reason first; act the next day at the earliest.

Evasion: decide on Monday to act on Tuesday, repeatedly, which satisfies the letter. Does it matter? Less than it seems. The rule's work is removing trades driven by a moment's salience, and a trade you still want tomorrow is a different animal. The rule fails against sustained bad judgement, which it was never going to fix.

2. Holding losers and selling winners (chapters 2 and 3). Before any sale, state what the holding is worth and whether you would buy it today at this price. If the only reason to keep it is the purchase price, that is not a reason.

Evasion: produce a justification, since one is always available. Does it matter? Yes, and this is the weakest of the three rules. It is partly rescued by requiring the answer in writing beforehand, because a written forecast can be checked later and a remembered one cannot.

3. Chasing what is salient (chapters 7 and 8). New holdings come only from a written watchlist assembled at least a month earlier. Nothing enters the portfolio in the same month it entered your awareness.

Evasion: add things to the watchlist as you notice them, so the month passes while the enthusiasm survives. Does it matter? Not much — the delay is doing the work, and an idea that is still interesting a month later has at least survived the news cycle that produced it.

The general finding from this exercise: every rule is evadable, because you are both the author and the subject. That is not a reason to skip the rules. A rule that is merely inconvenient to break still changes behaviour at the margin, and the margin is where the money is. What a written rule reliably does is make the evasion visible to you — you cannot quietly revise what is written down, you can only knowingly override it, and knowingly overriding your own considered judgement is harder than drifting.

On making rules too costly: a genuinely unbreakable arrangement — a long lock-in, an illiquid product — removes your ability to respond to real changes in your circumstances, and real changes do happen. The aim is friction, not imprisonment.

The honest summary of the subject

Ten chapters, and the conclusions are modest.

The documented tendencies are real, directional, and measured in millions of accounts. They are not curable by understanding them. The largest single cost is activity, which makes the most valuable habit the least impressive-looking one. And the parts of the field that promise to tell you what others will do next are far weaker than the parts that tell you what you will do next.

Which is why the subject is finally about your own conduct, and why it belongs next to a subject on ethics and regulation: both are about what happens when an incentive, yours or someone else's, points away from your interest.

The point

Biases persist in people who can name them, because from the inside a bias presents itself as a reason rather than as an error — so the remedy is not better judgement in the moment but fewer decisions in the moment. Automation, written pre-commitment, deliberate friction and well-set defaults all work; resolving to be disciplined and watching more closely do not. Every self-imposed rule is evadable since you are both author and subject, but a written rule converts quiet drift into a knowing override, and that alone changes behaviour where it matters.