Dollar cost averaging is insurance, and it is not free

Two different things share the name

Investing part of each paycheque as it arrives is not dollar cost averaging in any decision-relevant sense. It is investing money at the moment you have it, and there is no alternative to compare it against. The genuine version is holding a lump today, in cash, and choosing to deploy it over the next six or twelve months. Only that second case involves a choice, and only that case has a cost.

Conflating the two is how the strategy acquired its reputation. The paycheque version is obviously good, so the deliberate version inherits the endorsement without ever being examined.

Why the lump wins more often than not

The asset you intend to own is expected to return more than cash. Every month the money sits waiting is a month you have declined that premium. Because markets rise in more periods than they fall, the delay usually costs rather than pays, and the cost scales with how long the averaging window is. Across long United States samples the lump sum finishes ahead in roughly two thirds of rolling twelve-month comparisons; the exact share moves with the market and the period, which is what the calculator is for.

Notice what this argument does not require. It does not need markets to be efficient, or valuations to be reasonable, or the next year to be good. It only needs the asset to have a positive expected return relative to cash, which is the reason you were buying it.

What averaging actually buys

A narrower distribution of outcomes. The best case gets worse, the worst case gets better, and the middle barely moves. If your lump lands one month before a 35% decline, averaging over a year turns a catastrophe into a bad start, and that is a real benefit even though it lowers the expectation.

The second benefit is behavioural and it is the one that usually decides the question honestly. A strategy you abandon in month three has a realised return of whatever you locked in when you abandoned it. If an immediate large drawdown would make you sell, then the theoretically optimal plan has a lower expected return than the plan you would actually stick to. That is not a rationalisation, it is a correct comparison between two things you might really do.

Three questions, then decide

Our position: if the horizon is over ten years and the money is not needed in the meantime, invest it now. If you cannot, cap the averaging window at six to twelve months, because past that point you are no longer buying insurance, you are making a directional bet on the next two years and calling it prudence.

And if the answer to the second question is that a 30% fall on day one would genuinely change your behaviour, the position size is what needs adjusting, not the entry schedule. Averaging into a position that is too large for you produces the same problem eleven months later.

• Would you sell this position after a 30% fall in the first month.
• Is the money needed for anything inside the next ten years.
• If you already owned the asset today, would you sell it to hold cash instead.
• Any window you choose, write down the dates in advance and do not renegotiate them.