Is Dollar Cost Averaging or Lump-Sum Investing better?

Is Dollar Cost Averaging (DCA) or Lump-Sum Investing (LSI) better? This is a question which always leaves investors wondering and hoping for more clarity. Why? Because most of the academic finance theory as well as historical evidence shows that Lump-Sum Investing is better, but yet financial advisors and investors in general practice some form of Dollar Cost Averaging. Are investors always so sure of their own wisdom and stubborn in the face of data and evidence? Even Warren Buffett, never a follower of academic theories, advises:

By periodically investing in an index fund, for example, the know-nothing investor can actually out-perform most investment professionals. Paradoxically, when “dumb” money acknowledges its limitations, it ceases to be dumb.
Warren Buffett, Shareholders’ Letter 1993

So what do we do? While for most of the time as an investor, the default investment method will be Dollar Cost Averaging, since we may invest monthly or quarterly with the savings from our income, there may be times when we have the choice of Dollar Cost Averaging or Lump-Sum Investing. For example, when we inherit a sum of money, or strike the lottery, or when we get access to our mandatory savings upon retirement. What do we do then? Is Dollar Cost Averaging or Lump-Sum Investing better? Or are there circumstances we may want to use one or the other?

The Case for Dollar Cost Averaging. Or is Lump-Sum Investing better?

Why is Lump-Sum Investing Supposed to be Better?

Let’s start with understanding why Lump-Sum Investing is usually considered a better choice. Back in the day in 1949, JF Weston pointed out that Formula Timing Plans (as Dollar Cost Averaging was known back then) did not perform well under all market scenarios. For example, if prices are rising, and will continue to rise, it is better invest in a lump sum instead of dollar cost averaging. Hence, investing in a fixed and inflexible manner without considering the current market or economic situation would arguably be worse than timing investments based on updated information. This is the same point made in Constantinides (1979), which has been quoted extensively in almost all later research. Ignoring updated market information when investing is a suboptimal way of investing.

But that is in theory. What if we couldn’t make head or tail out of the most updated market information? The way that the recent interest rate hikes, equity and bond market movements, and how they seem to have caught many professional investors off guard, all seem to point to the fact that even with updated market information, it is not clear that we can make better investment decisions. So much for theory!

And Dollar Cost Averaging continues to be popular. Hence, the next step was to formulate as to how this can be rational due to behavioural reasons, as in Statman (1995). In a large part, this is surely true. Dollar Cost Averaging takes away a lot of the fear, emotions, and guesswork out of investing, and if this makes investors more comfortable with investment process, then there are grounds for why it continues to be popular.

Does Dollar Cost Averaging Work in Practice?

But does Dollar Cost Averaging really work in practice? We can turn to studies done on historical data to see whether Lump-Sum Investing works better, as the theory predicts. And the results turn out to be pretty uniform in supporting it! For example, Vanguard (2012) finds that Lump-Sum Investing gives higher returns compared to Dollar Cost Averaging over 12 months in almost two-thirds of the time, across the US, UK and Australia, and across stock, bond and balanced portfolios:

Vanguard: Lump-Sum Investing trumps Dollar Cost Averaging in 3 markets historically

Vanguard: Dollar Cost Averaging or Lump-Sum Investing? — Source: Vanguard (2012) *Dollar-cost averaging just means taking risk later*

The same results in favour of Lump-Sum Investing have also been documented across even more markets by PWL Capital (2020). Here, it is interesting to note that Lump-Sum Investing has worked best in the US, and worst in Japan, for obvious reasons. However, the chance of success of Lump-Sum Investing compared to Dollar Cost Averaging is still around 65%.

PWL Capital: Lump-Sum Investing trumps Dollar Cost Averaging in 6 markets historically

PWL: Dollar Cost Averaging or Lump-Sum Investing? — Source: PWL Capital (2020) *Dollar Cost Averaging vs. Lump Sum Investing*

Now, the results in favour of Lump-Sum Investing may not mean much if the returns between the two approaches were not very different At first glance, this is what the historical record tells us. the average difference in performance is around 0.38% per annum, which is a difference of roughly 5% of the long term return. Even if we allow this difference to compound over 10 years (assuming compounded returns of 7% for Dollar Cost Averaging), the difference in end wealth after 10 years is only 3.5% (7.3% after 20 years). Quite small indeed!

PWL Capital: The difference in returns between Lump-Sum Investing and Dollar Cost Averaging is not large historically

PWL: Is dollar cost averaging or lump sum investing better? — Source: PWL Capital (2020) *Dollar Cost Averaging vs. Lump Sum Investing*

But where Lump-Sum Investing also does better historically is on the upside and downside. Even if the mean returns between Lump-Sum Investing and Dollar Cost Averaging are small, on both the upside and downside, Lump-Sum Investing does gives significantly better outcomes!

PWL Capital: Lump-Sum Investing does significantly better than Dollar Cost Averaging historically when we consider both upside and downside outcomes instead of the average

PWL Capital: DCA vs LSI — Source: PWL Capital (2020) *Dollar Cost Averaging vs. Lump Sum Investing*

Finally, when we looked at the theoretical reasons against Dollar Cost Averaging and in favour of Lump-Sum Investing, we argued that it is suboptimal not to take updated market information into consideration when investing. But on the other hand, the market gives a lot of signals, and interpreting them can be quite hard. So, in a situation when we do not know which way the market is going, does it still make sense to do Lump-Sum Investing? It turns out that the answer is a qualified yes:

PWL Capital: Lump-Sum Investing is still as good as Dollar Cost Averaging even during bear markets historically

PWL: LSI vs DCA in Bear Markets — Source: PWL Capital (2020) *Dollar Cost Averaging vs. Lump Sum Investing*

PWL Capital: Lump-Sum Investing is still better than Dollar Cost Averaging even when the US market seemed overvalued historically

PWL: LSI vs DCA when markets are overvalued — Source: PWL Capital (2020) *Dollar Cost Averaging vs. Lump Sum Investing*

In a bear market, not surprisingly, Lump-Sum Investing doesn’t do as well as it did over all periods. the difficulty of being able to invest at the market lows reduces the effectiveness of Lump-Sum Investing, but in general, it still performs at roughly the same success rate as does Dollar Cost Averaging. But for the US at least, even when market look overvalued, Lump Sum Investing still does as well as before, better than Dollar Cost Averaging! So much for staying in cash and waiting for the market to fall!

So with all this theory and historical evidence backing Lump-Sum Investing instead of Dollar Cost Averaging, why are we not doing Lump-Sum Investing all the time?

Why and When Dollar Cost Averaging May Still Work

This is where things get interesting. Conceptually, we know that Dollar Cost Averaging works best in two situations:

When markets are volatile, because DCA allows us to have the chance to “buy on the dips”
When markets are trending down, because DCA allows us to buy when market prices are lower

Historically though, Lump-Sum Investing has usually worked out better. But this can also be because the past history of stock markets in general has been quite unique:

Stock markets have been trending up most of the time, especially since the 1980s
When stocks prices go up, market volatility tends to go down, and vice versa when they move in the opposite direction
As interest rates have trended down since the 1980s until very recently, the returns on holding cash instead of investing in the stock markets have been low

In short, the historical period on which most of the research on Lump-Sum Investing vs Dollar Cost Averaging has been done is uniquely suited for Lump-Sum Investing.

What if these conditions do not hold? Morgan Stanley (2020) investigates this question with a series of market simulations, which we show below. The results are instructive:

Morgan Stanley: The success of Lump-Sum Investing is driven by high stock market returns and low volatility

Morgan Stanley: LSI vs DCA — Source: Morgan Stanley (2020) *Dollar-Cost Averaging Versus Lump-Sum Investing: Behavioral Considerations and Potential Outcomes*

From the table on the left, we see that as market returns come down (the return spread is the difference between stock and cash returns), and when volatility goes up, both LSI and DCA have a roughly 50-50 chance of outperforming the other. In fact, when volatility goes even higher, DCA can become the better option. Having said that, from the table on the right, the difference in returns between these two approaches is likely to be small, smaller than what they have been historically, and perhaps of an order of magnitude similar to the fees on Index Funds or Exchange Traded Funds.

So for investors in markets such as China, Hong Kong or Singapore, and also in tech stocks, where the volatility is high, Dollar Cost Averaging may in fact be a better way of investing. Intuitively, given the volatility in the market, it is hard to know on any single day, whether the lump-sum investment made will be at a higher or lower price. DCA, by spreading the investments over a period of time, allows the investor to better average out the purchase price. Moreover the costs of being wrong (i.e. in the form of lower returns) are fairly low as well. For Singapore especially, the higher volatility of the market (25% for the Straits Times Index vs 15% for the S&P500), and the lower returns (roughly 5%-6% historically) mean that DCA is probably the better option all the time!

And in terms of the situation in the markets today, with:

Downward trending markets
Higher market volatility
Higher returns to holding cash due to higher interest rates

There are grounds to believe that a Dollar Cost Averaging approach to investments may work out as well as the Lump-Sun Investing approach, if not better!

Conclusions

There is an endless stream of articles and research and studies telling us that Lump-Sum Investing is better than Dollar Cost Averaging. And yet investors Dollar Cost Average all the time! Why is this so?

Well, for a start, most studies comparing the two approaches cover a period when stock returns were high, and market volatility was low. In short, the perfect scenario of Lump-Sum Investing. But this may not hold in the future, especially when we are looking at downward trending markets, higher volatility and higher interest rates right now. These three factors are, in fact, the conditions whereby Dollar Cost Averaging may work better. Looking ahead, the chances of either LSI or DCA doing better might actually be 50-50, compared to the 66-34 we have seen historically.

Moreover, the costs in getting the investing approach working, in the form of lower returns, is actually quite small. The costs of investing in bonds/cash instead of stocks, or not letting investments compound over time, for example, are far higher than getting LSI vs DCA wrong. In fact, when in doubt, why not diversify? By having half of the sum of money for investments put to work through Lump-Sum Investing, and the other half through Dollar Cost Averaging. If anything, such an approach would minimise investors’ regret!

In the volatile and uncertain markets today, putting half your money to work through Lump-Sum Investing and half through Dollar Cost Averaging is probably the best way to hedge your bets and minimise investors’ regret!