Track 7 · Claims · lesson 9
Why investors trail their own investments
11 min
Track 7 · Claims · lesson 9
11 min
Suppose a fund reports a return of eight percent a year over a decade. The figure is accurate, audited and not misleading in any way.
Now suppose you could go and ask every person who held it during that decade what they actually earned. There is no reason for the average of their answers to be eight percent, and a structural reason for it to be lower. Nobody has lied and no fee has been hidden. The two numbers are answering different questions.
A fund's reported return measures what one unit of money would have earned if it had been present for the whole period, ignoring the size of the pot at each moment.
An investor's return measures what their actual money earned, given how much of it was there for each stretch.
Those diverge whenever money arrives and leaves at uneven times — which is always, because money arrives when people feel good and leaves when they do not.
Take a fund with a very simple history. In its first year it returns 50%. In its second it falls by a third.
Over the two years together, one unit present throughout ends where it started: up a half, then down a third, back to level. The fund reports a two-year return of zero, correctly.
Now put an investor into it. They start with 100 at the beginning of year one and end that year with 150. The performance has been excellent and it has been noticed, so they add 900 — a perfectly ordinary human response to a good year. They now hold 1,050 going into year two, which falls by a third to 700.
They contributed 1,000 in total and hold 700. The fund returned nothing and its investor is down almost a third, and no charge, error or misconduct is involved anywhere in the story. Most of the money was not there for the good year, and all of it was there for the bad one.
Not the same thing
What one unit of money would have earned by being present for the whole period.
What the money actually put in actually earned, weighted by how much was present when.
Which is which? Put each one on a side.
The figure used to rank funds against each other
The number that determines whether you can afford the thing you were saving for
A return that improves when a holder adds money just before a good stretch
A return unaffected by anybody adding or withdrawing anything
The pattern is not stupidity, and describing it as stupidity is both unkind and useless, because it suggests the remedy is trying harder.
Good performance is the advertisement. A holding is most visible, most discussed and most recommended immediately after a strong run. That is when money arrives — not because people are chasing, exactly, but because that is when they hear about it.
The bad stretch is when the money is needed. Falls cluster with the events that cost people their income, which is the same clustering from the liquidity lesson. Withdrawals happen at the bottom partly because that is when withdrawals are necessary.
Holding through a fall is genuinely unpleasant. Watching a number decline for eleven months is not a small ask, and treating it as one is how plans get written that nobody can follow.
Predict
Three separate things get bundled into the phrase, and separating them is useful because only one of them is behavioural.
Timing of flows. The arithmetic above. Money present for the bad stretches and absent for the good ones.
Switching between holdings. Selling one thing after it has done badly and buying another after it has done well, which is the same effect applied across funds rather than in and out of one, plus the costs of doing it.
Costs of the movement. Every switch has a transaction cost and, where gains are realised, brings a tax charge forward from later to now. The previous lesson showed what bringing a charge forward does over a long period.
Only the first two are about behaviour. The third is a mechanical consequence of the first two, and it is why the gap is larger than the timing alone would produce.
Check