Dollar-cost averaging: How often should one use it? What criteria to use when choosing stocks to apply it to?

Dollar cost averaging is a great strategy to use for investment vehicles where you can't invest it in a lump sum. A 401K is perfect for this. You take a specific amount out of each paycheck and invest it either in a single fund, or multiple funds, or some programs let you invest it in a brokerage account so you can invest in virtually any mutual fund or stock. With annual or semi-annual re-balancing of your investments dollar cost averaging is the way to invest in these programs.

If you have a lump sum to invest, then dollar cost averaging is not the best way to invest. Imagine you want to invest 10K and you want to be 50% bonds and 50% stocks. Under dollar cost averaging you would take months to move the money from 100% cash to 50/50 bonds/stocks. While you are slowly moving towards the allocation you want, you will spend months not in the allocation you want. You will spend way too long in the heavy cash position you were trying to change. The problem works the other way also. Somebody trying to switch from stocks to gold a few years ago, would not have wanted to stay in limbo for months.

Obviously day traders don't use dollar cost averaging. If you will will be a frequent trader, DCA is not the way to go.

No particular stock type is better for DCA. It is dependent on how long you plan on keeping the investment, and if you will be working with a lump sum or not.

EDIT: There have be comments regarding DCA and 401Ks. When experts discuss why people should invest via a 401K, they mention DCA as a plus along with the company match. Many participants walk away with the belief that DCA is the BEST strategy. Many articles have been written about how to invest an inheritance or tax refund, many people want to use DCA because they believe that it is good. In fact in the last few years the experts have begun to discourage ever using DCA unless there is no other way.

How often should one use dollar-cost averaging?

Trivially, a dollar cost averaging (DCA) strategy must be used at least twice! More seriously, DCA is a discipline that people (typically investors with relatively small amounts of money to invest each month or each quarter) use to avoid succumbing to the temptation to "time the market". As mhoran_psprep points out, it is well-suited to 401k plans and the like (e.g. 403b plans for educational and non-profit institutions, 457 plans for State employees, etc), and indeed is actually the default option in such plans, since a fixed amount of money gets invested each week, or every two weeks, or every month depending on the payroll schedule. Many plans offer just a few mutual funds in which to invest, though far too many people, having little knowledge or understanding of investments, simply opt for the money-market fund or guaranteed annuity fund in their 4xx plans. In any case, all your money goes to work immediately since all mutual funds let you invest in thousandths of a share.

Some 401k/403b/457 plans allow investments in stocks through a brokerage, but I think that using DCA to buy individual stocks in a retirement plan is not a good idea at all. The reasons for this are that not only must shares must be bought in whole numbers (integers) but it is generally cheaper to buy stocks in round lots of 100 (or multiples of 100) shares rather than in odd lots of, say, 37 shares. So buying stocks weekly, or biweekly or monthly in a 401k plan means paying more or having the money sit idle until enough is accumulated to buy 100 shares of a stock at which point the brokerage executes the order to buy the stock; and this is really not DCA at all. Worse yet, if you let the money accumulate but you are the one calling the shots "Buy 100 shares of APPL today" instead of letting the brokerage execute the order when there is enough money, you are likely to be timing the market instead of doing DCA. So, are brokerages useless in retirement fund accounts? No, they can be useful but they are not suitable for DCA strategies involving buying stocks. Stick to mutual funds for DCA.

Do people use it across the board on all stock investments?

As indicated above, using DCA to buy individual stocks is not the best idea, regardless of whether it is done inside a retirement plan or outside. DCA outside a retirement plan works best if you not trust yourself to stick with the strategy ("Ooops, I forgot to mail the check yesterday; oh, well, I will do it next week") but rather, arrange for your mutual fund company to take the money out of your checking account each week/month/quarter etc, and invest it in whatever fund(s) you have chosen. Most companies have such programs under names such as Automatic Investment Program (AIP) etc. Why not have your bank send the money to the mutual fund company instead? Well, that works too, but my bank charges me for sending the money whereas my mutual fund company does AIP for free. But YMMV.

Dollar-cost averaging generally means investing a fixed amount of money on a periodic basis. An alternative strategy, if one has decided that owning 1200 shares of FlyByKnight Co is a good investment to have, is to buy round lots of 100 shares of FBKCO each month. The amount of money invested each month varies, but at the end of the year, the average cost of the 1200 shares is the average of the prices on the 12 days on which the investments were made. Of course, by the end of the year, you might not think FBKCO is worth holding any more. This technique worked best in the "good old days" when blue-chip stocks paid what was for all practical purposes a guaranteed dividend each year, and people bought these stocks with the intention of passing them on to their widows and children.

Dollar cost averaging can be done in a retirement plan, and can be done for individual stock purchases, as this will increase your returns by reducing your risk, especially if you are buying a particular stock for the first time.

How many time have I purchased a stock, bottom fishing, thinking I was buying at the low, only to find out there was a new low. Sitting with a thousand shares that are now down $3-$4K. I have a choice to sell at a loss, hold what I've got or double down. I usually add more shares if I'm thinking I'll recover, but at that time I'd wished I'd eased into my investment. That way I would have owned more shares at a smaller cost basis.

Anything can happen in the market, not knowing whether the price will increase or decrease. In the example above a $3,000 loss is equal to the brokerage cost of about 300 trades, so trading cost should not be a factor. Now I'm not saying to slowly get into the market and miss the bull, like we're having today with Trump, but get into individual stocks slowly, being fully invested in the market.

Also DCA means you do not buy equal number of shares per period, say monthly, but that you buy with the same amount of money a different number of shares, reducing your total costs. Let's say you spend $2000 on a stock trading at $10 (200 shares), if the stock rose to $20 you would spend $2000 and buy 100 shares, and if the stock dropped to $5 you would spend $2000 and buy 400 shares, by now having amassed 700 shares for $6,000. On the other hand and in contrast to DCA had you purchased 200 shares for $2000 at $10/share, then 200 shares for $4000 at $20/share, and finally 200 more shares for $1000 at $5/share, you would have amassed only 600 shares for $7000 investment.

Note:

I came across this question and I was not satisfied with the answers available, I wished for some sort of mathematical "proof". Luckily, later I came across the recent paper of Kirkby, J. Lars, Sovan Mitra, and Duy Nguyen. "An analysis of dollar cost averaging and market timing investment strategies." European Journal of Operational Research 286.3 (2020): 1168-1186., which focuses precisely on this topic. I will mention some ideas, but please read it for more details.

Definitions:

• Dollar Cost Average (DCA): investment strategy in which a fixed dollar amount is invested in a risky asset (such as a mutual fund or exchange-traded fund (ETF)) at regular time intervals, and over a predefined holding period.

• Lump Sum (LS): investment strategy where all the money is invested initially over a predefined holding period.

Ideas from the literature:

• A potential advantage of DCA is that it is deemed to reduce the volatility of investments by smoothing out the effects of volatile movements in asset prices with regular, periodic asset purchases. This has also been referred to as time diversification, in that it reduces the risk from purchasing at the “incorrect” time.

• In an empirical study, Williams and Bacon (1993) compares the annualized returns from various DCA strategies with those produced by LS investing in the S&P 500 from 1926 to 1991. For all time periods and averaging strategies investigated, LS investing produced superior returns to DCA, and in all but one instance.

• Dubil (2005), Trainor (2005) showed that DCA reduces shortfall risk, which is a major factor in retirement saving and retail investing. Grable and Chatterjee (2015) documented that a DCA strategy provides a way to outperform during a bear market (rather than a bull market) for clients who have less tolerance for financial risk. Luskin (2017) reported that DCA outperforms LS investment during the period when cyclically adjusted price-to-earnings (CAPE) ratio is high. Moreover, using mutual fund data, ( Israelsen, 1999 ) argued that LS investing does not always yield superior returns over dollar-cost averaging, especially if the volatility of mutual funds is low.

Results from the paper:

• They find theoretically and with historical data of the S&P 500, that the expected returns of the DCA strategy decrease with the number of times the money is divided along the time window. That is, in general is better to follow a DCA strategy yearly instead of monthly or daily.

• They find theoretically and empirically that the LS investment will have higher expect return as compared to that of a DCA investment, but at the cost of a higher variance of wealth. This means that in general the LS strategy will have higher returns, but there is also a higher chance to have low returns. The trade-off, which they measure by the Sharpe Ratio, favours either the LS strategy or a DCA strategy where money is invested yearly.