Which of these biases is costing you the most?

Biases are not distributed evenly. Some people trade far too much and hold their losers perfectly well; others are patient to a fault and only ever buy after a rally. Generic advice therefore under-performs a personal diagnosis — and a personal diagnosis cannot be produced by introspection, because the instrument you would introspect with is the one under investigation.

So use your records instead. Your statement is a confession you did not know you were writing.

The five-line audit

Pull twelve to twenty-four months of transactions and compute:

1. Holding-period asymmetry. Average days holding winners versus losers. A large gap in favour of losers is the disposition effect, quantified.
2. Turnover and friction. Number of completed round trips × average size × your round-trip cost, expressed as a percentage of the portfolio. Count round trips, not executions — a buy and its matching sell are one of them, and the round-trip cost already covers both sides.
3. Entry timing. What fraction of your buys occurred within a week of a strong move up in that asset? That is herding and recency, measurable.
4. Source balance. Of the last ten things you read about your largest holding, how many argued against it? Zero is a finding.
5. Your behaviour gap. Time-weighted return minus money-weighted return, from your broker. The single most complete number available to you.

What an audit looks like when it bites

A real-shaped result on a €50,000 account:

• 42 completed round trips in twelve months — 84 executions; average position €5,000; round-trip cost 0.4% → €840 of friction, about 1.7% of the portfolio, every year, forever.
• Average winner held 34 days. Average loser held 190 days. A textbook disposition effect — winners get about one month of patience, losers get six.
• Money-weighted return 1.8 points below time-weighted.

Nothing here required a market view or a valuation opinion. Three numbers, one afternoon, and the two largest leaks in the portfolio are identified — and both are fixable with rules from the previous lesson rather than with better forecasting.

Unit checkpoint ahead

You now have the six mechanisms, the number that measures their combined cost, the rules that blunt them, and a method for finding your own worst one. That completes the behavioural half.

The second half changes one variable and the stakes with it: the money belongs to someone else. Every bias in Unit 1 still operates — but now a bad decision is not a private expense, it is a breach of a duty owed to a person who trusted you. Unit 3 asks what that duty actually requires.

Try it now

1. Run all five lines of the audit on your own records. Write the results down; an audit you did not write down did not happen.
2. Rank your biases by cost, not by how bad they feel. Then write one rule — just one — aimed at the most expensive.
3. Diary a date three months out to re-run the same five lines. A bias audit is a measurement, and one measurement is not a trend.