I'm not a big fan of Dollar Cost Averaging. If you have a lump sum amount you were going to invest, do it fully at once and let all of the money work. If you divide it and invest over a period of time, you are (1) letting the un-invested money be idle. (2) chasing a market downturn to invest the rest in hopes of "averaging down". Remember, market goes up more often than it goes down, so you will likely end up increasing your cost basis than if you had invested at once.

Bitcoin is highly volatile. If you do a lump sum payment you might end up joining the market at the top of a bubble where the value of Bitcoin is 12x times its long term value. Sure it will not crash 12x down but going 70% down after the bubble could be enough to cause a panic sale and thus lose your money.

>If you divide it and invest over a period of time, you are (1) letting the un-invested money be idle

No, that's not always true. For example. When you earn $5000 in excess income per year you can decide to invest it monthly. That's dollar cost averaging and you didn't let your uninvested money sit idle.

You can save up $5000 until the end of the year and then invest it. That's a lumpsum and you did let your uninvested money sit idle.

Investing your excess income as you get it isn't really Dollar Cost Averaging the way firms talk about it. It's called DCA on forums and such, but it's really just investing your excess lump sum each time as soon as you get it.

Lump Sum vs DCA comparisons are usually made in the context that you have your lump sum now. Not that you're accumulating a lump sum and trying to time the market for when to invest it. The page is titled "How to invest a lump sum of money" after all.

> Suppose you received a windfall. Someone gave you a gift or you inherited a lot of money. Maybe you hit the lottery jackpot or got a huge bonus.

> Here's the question you face: Should you invest it all right away or in smaller increments over time, a strategy known as dollar-cost averaging?

Thanks for the correction, I didn’t know DCA was meant for lump sums. That makes it really dumb, I think, as it assumes down market. This whole time I thought DCA was a reflection of constantly investing chunks over time as money comes available, a good idea I think.

The Wikipedia article [0] has quite a few links to studies showing this is worse than lump sum.

I’m trying to think of any scenarios where this would be smart, but they all seem to assume market knowledge and just adjusting a lump sum payment. Even if you “knew” it was a down market and didn’t know when it would pick up it is bad because you would miss the biggest upswings at the bottom of the bear.

I think this is just one of those platitudes that retail advisors give people because it sounds good, but actually isn’t.

[0] https://en.wikipedia.org/wiki/Dollar_cost_averaging

> Bitcoin is highly volatile. If you do a lump sum payment you might end up joining the market at the top of a bubble where the value of Bitcoin is 12x times its long term value.

I should have been clear, I'm not talking about Bitcoin at all. Just traditional index investing.

Research generally shows that if you have a lump sum of cash, it's better to invest it all at once rather than average it out.

https://investor.vanguard.com/investing/online-trading/inves...

Time in the market > Timing the market, and all that jazz.

Notice how it says "Expected Return". The problem is that it is better "on average" to invest it all at once (and doesn't even talk about variance). But an individual is not an average. They will lie somewhere on that spectrum. I'd rather take a small hit to the expected value of my returns in order to reduce risk of being on the wrong side of that average.

> I'd rather take a small hit to the expected value of my returns in order to reduce risk of being on the wrong side of that average.

This is a common fallacy. That is not actually the tradeoff you are making, in fact you are taking a hit to the mean value, and also greatly increasing the chances of being on the wrong side of the average[1]. Dollar cost averaging increases variance and decreases the expected return in proportion to your averaging period. You earn less and have a much lower floor for almost no increase in the ceiling.

There are nice visualizations here: https://ofdollarsanddata.com/lump-sum-investing/

[1] at least according to history. Who knows what the future holds.

Good post thanks for sharing. It's true, the market grows more often than it shrinks and I could see how that skews the distribution in favor of lump sum.

You’re still an individual either way. So it’s still better to use expected return.

Maybe you could compare and choose an option with less variability or standard deviation even though the mean return is worse. But these studies don’t compare the size of average losses and gains. I expect that the variance is also better for lump sum.

The page you link to does admit, though, that dollar cost averaging minimizes possible downside risk.

Agreed that you shouldn't try to time the market, but it is of course possible that you invest that lump sump at a high, and then see a significant drop soon after. Sure, over time that drop will almost certainly rectify itself, but psychologically you might feel better about buying in over a longer period of time.

Living with slightly lower long-term returns in exchange for peace of mind is a perfectly valid trade off that people can choose to make.

That page is inaccurate, according to historical data the downside risk is always greater with dollar cost averaging: https://ofdollarsanddata.com/lump-sum-investing/

This is because there are always two downsides, the risk of investing just before a huge crash, but also the risk of holding out while a boom passes you by. People always psychologically overweight the former due to loss aversion, but the latter is more pernicious.

That's what the chart says but what does human behavior say? Does everyone truly ride the crash down and back up or will some of them panic and leave the market before it recovers?

One of those risks a 70% loss the other one merely risks unrealized gains.

> Does everyone truly ride the crash down and back up or will some of them panic and leave the market before it recovers?

This is an orthogonal issue. If you DCA for 12 months and then a crash immediately occurs, you suffer the exact same fate (you are worse off than someone who had 12 months of investment returns prior to the crash). Does this mean you should increase DCA horizon to 24 months? The only way your framing makes sense is if you have prior knowledge that a crash will occur during the DCA period.

You should choose your asset allocation to fit your risk-reward preferences. Once you've done that it's suboptimal to further reduce risk at the expense of reward by holding more cash for longer.

So what has the better expected return between a) invest lump sum all at once into a low-volatility portfolio b) invest the lump sum progressively into a slightly higher-volatility portfolio, such that the expected regret remains the same?

It's (a).

The optimal strategy can't possibly be (b), because it would have the bizarre implication that two investors who have exactly the same preferences and who currently have the same amount of wealth would want to allocate their investments differently depending on how far in the past they started investing.

Edit: I'm quite sleepy right now, but I think the way the mathematics works out is that the risk arising from an asset is a convex function of the amount allocated to it. So by Jensen's inequality it's less risky to invest 50/50 in cash and stocks for two years than to invest for one year in cash and one year in stocks, for the same reward.

This stops being generally true if you assign a non-linear utility to future dollars.

Above some point more dollars are less valuable to you, below some point they are much more valuable to you.

If your investment objectives will be very likely satisfied from the returns of the DCA and the lump sump has a higher probability of failure -- why wouldn't you DCA even though it has lower expected returns?

> the lump sump has a higher probability of failure

For this to be true, you must be a successful market timer. Whatever your DCA period is, at the end of the period, you are fully invested -- a crash at that point in time would hurt you just as much as a crash would hurt a 100% equities from day 1 investor. DCA'ing is a bet that a crash will occur sometime before you are fully in. In every other scenario, the DCA is strictly worse. If you don't happen to know the timing of future crashes, it's better to put your money to work now than later, come what may.

To account for non-linearly increasing dollar utility, it's better to choose a less volatile asset allocation, not to delay your investment period.

And yet, back tested it shows significantly lower variance. (of course, that isn't a guarantee... but it's the same kind of analysis that says lumpsumming is better, in terms of expected return)

> In every other scenario, the DCA is strictly worse.

Yes. And so what? It will always be the case that a lower variance approach will be better in the cases where the higher variance approach did poorly and worse otherwise.

> it's better to choose a less volatile asset allocation

These are not mutually exclusive. One can lump sum all into the lowest risk allocation and then risk on over time.

It doesn't help that right now there is not much freedom in asset allocation at all. Your choices are pretty limited to higher risk things and things with a negative real yield.

> back tested it shows significantly lower variance

Not sure what you're using as the measure of variance, but Morningstar found DCA has higher variance in IRR: https://www.morningstar.com.au/learn/article/the-dollar-cost...

> These are not mutually exclusive. One can lump sum all into the lowest risk allocation and then risk on over time

I never said they were mutually exclusive, just that it is suboptimal.

yeah but cost averaging is often a psychological technique. It's easier to commit yourself to cost average than it is to commit yourself to a lump sum investment.

historical performance in equities shows that lump sum investing has beaten dollar cost averaging over almost any time period

I’m aware of the macroeconomic policies that coincide with that observation, but people will lean on those papers

In higher volatile assets or periods, dollar cost averaging is the better move, so people that come from more volatile markets are correct to lean on that