The Right Answer for a Robot

A lump of money lands in your lap. An inheritance, the proceeds of a flat you sold, a bonus bigger than usual. You know, roughly, that it should go into the market and be left there for twenty years. And then you do nothing with it for eight months, because you can’t decide whether to put it all in at once or feed it in slowly, and both feel like a way to get it wrong. The money sits in a savings account earning a pittance while you wait for a sign that never comes.

If you took the question to a financial theorist, the answer would be clean: put it all in immediately. This is called lump-sum investing, and the case for it is genuinely strong. Because stock markets rise more often than they fall, and by more, the money you invest today is, on average, money that starts compounding today rather than next spring. Vanguard ran the historical numbers across decades and three countries and found that investing a lump all at once beat spreading it out about two-thirds of the time, by a bit over two percent in the first year for a typical portfolio. Every month you hold the money back as cash, you’re sitting out of a game whose expected score is positive. The name of Vanguard’s paper says it with a shrug: dollar-cost averaging just means taking risk later.

So the maths is settled, and it points at lump sum. And yet telling most people “the maths says all at once, so do that” is bad advice, and understanding exactly why is the whole point of this.

The trouble is that the clean answer is the right answer for a robot. It assumes an investor who feels nothing when a number on a screen drops, who will look at a fresh twenty percent loss on money they invested last Tuesday and calmly do nothing. You are not that investor, and neither am I. We have a quirk, well measured, that a loss hurts about twice as hard as an equal gain feels good, so a big immediate drop on newly invested money doesn’t register as “temporary volatility,” it registers as “I have just destroyed a fortune I was trusted to look after.” And the thing that actually wrecks real portfolios is not bad timing. It’s what people do after bad timing: they panic, they sell at the bottom to make the pain stop, and they turn a paper loss into a permanent one. The single most expensive event in an investing life is going all-in the day before a crash and then capitulating at the trough.

Seen that way, spreading the money in over several months, dollar-cost averaging, is not a worse bet that innumerate people make. It’s insurance. Not insurance against low returns, insurance against yourself. If you drip the money in over, say, a year and the market falls, only a fraction of your capital is exposed to that first drop, and, crucially, the story you get to tell yourself flips from “I lost a fortune” to “good, the next instalment buys in cheaper.” That reframing is worth more than it sounds, because it’s the difference between a person who keeps investing through the fall and a person who slams the door. You give up a couple of percent of expected return to buy a much lower chance of the catastrophe. For a lot of people, that’s a trade worth making, and it’s not irrational at all, it’s just priced in a currency, peace of mind, that the spreadsheet doesn’t track.

Which of the two is right for you comes down mostly to one thing people rarely name: how big the sum is relative to everything else you have. A ten-thousand bonus landing on top of a half-million portfolio is a rounding error; if it drops twenty percent you’ll barely feel it, so put it straight in. An inheritance that doubles your net worth is a different animal entirely, because now a bad month isn’t a rounding error, it’s a shock large enough to make you do something stupid, and against that, feeding it in slowly is cheap protection. The other inputs are your temperament and your timeline. If you’ve held stocks through a real crash and didn’t flinch, and you won’t need the money for a decade or two, lump sum, and don’t overthink it. If you’re new to this, or you lost sleep during the last wobble, respect that, it’s information about you, not a character flaw, and feed it in. And if there’s any chance you’ll need the money within three to five years, this whole debate doesn’t apply, because that money shouldn’t be in the stock market in the first place.

There’s a clean test that cuts through all of it, one honest question to ask before you commit: if I put it all in tomorrow and the market fell twenty percent next week, would I sell? If the truthful answer is “yes,” or “I’d feel sick and I don’t know,” then you are, by definition, a drip-feed case, whatever the maths says about optimal returns, because for you lump sum leads straight to the worst outcome there is. If the answer is a genuine “I’d do nothing, or I’d buy more,” and the sum is small enough that losing half of it wouldn’t blow up your life, then you’re the disciplined investor the theory is written for, and you should take the two percent.

Now, two warnings, because the ways people get this wrong are more expensive than the choice itself. First, if you do drip the money in, do it on a fixed, automatic schedule, the same amount every month until it’s done, and then stop. What you must not do is turn “averaging in” into “waiting for a better moment,” because that’s just market timing wearing a sensible cardigan, and market timing is a skill almost nobody reliably has. So much of the market’s entire long-run gain arrives in a tiny handful of days that missing just the ten best days over a couple of decades can roughly halve your total return, and nobody knows in advance which days those are. A machine schedule protects you from the one thing you’re worst at, deciding in the moment.

Second, and this is the mistake that quietly costs the most: the real enemy is neither lump sum nor averaging. It’s the eight months of paralysis we started with. Money sitting in cash while you search for the perfect entry earns nothing and loses ground to inflation every single day, a small guaranteed loss you’ve chosen over an uncertain one. Both strategies beat that, easily, because both share the one feature that matters, they make you actually invest. If drip-feeding is the version you’ll actually go through with because it scares you less, then drip-feeding is the correct answer for you, full stop, even at a cost of a couple of percent, because the alternative on offer isn’t the flawless lump sum, it’s the money rotting in a current account for another year.

If you want a compromise that satisfies both the maths and the nerves, there’s an obvious one: put a big chunk in now, a third, a half, so you’re not left on the sidelines if the market climbs, and feed the rest in over the next six to twelve months. It captures most of the time-in-the-market advantage and takes the edge off the all-or-nothing dread, and there’s nothing shameful about it.

Because here is the thing the “optimal” framing misses, the reframe worth keeping. You are not, in the end, trying to squeeze out the last two percent of a first-year return. You’re trying to stay invested for twenty years, and the biggest threat to that isn’t the entry point, it’s you, bailing out after a crash or never starting at all. Both of those cost far more than two percent. The right strategy is simply whichever one you will actually stick to through a bad year, because the asset you’re really managing here isn’t the market. It’s your own behaviour, and that’s the one thing no expected-value calculation knows how to price.