As of February 17, 2012, the S&P 500 index has risen more than 23% off of its October 3, 2011, closing low.

Some investors, who were convinced that history would not repeat itself by seeing the “500” advance 23% in the six months after concluding a near-miss or baby-bear market, still have a lot of money on the sidelines. They are probably now asking themselves, “Should I throw in the towel and invest it all, or put my money back to work gradually since the market may correct any day now?” This dilemma is timeless for investors, who are basically asking if they should dollar cost average (invest a fixed amount at equal intervals) or lump-sum invest. The correct answer? “That depends.”

Testing the Approaches

To try to answer that question, I analyzed both approaches using two similar investment vehicles—the S&P 500 index and the S&P 500 Dividend Aristocrats index—from December 31, 1999, through February 17, 2012, to see which would have been the better approach. I had one hypothetical investor make an initial investment of $10,000 in the S&P 500 on December 31, 1999, who then added $1,000 at the start of the subsequent 48 quarters (for a total investment of $58,000). As of February 17, 2012, this investor’s portfolio grew to $75,611, including dividends (Figure 1). The second hypothetical investor, who plunked down all $58,000 on December 31, 1999, and let it ride, now has a portfolio that is worth $67,247. So in the case of the S&P 500 index, it was better for an investor to dollar cost average than it was to make a lump-sum investment.

But the same did not hold true when using a different investment vehicle. When this same dollar-cost-average investor made an initial investment of $10,000 in the S&P 500 Dividend Aristocrats index on December 31, 1999, and then added $1,000 in the subsequent quarters through February 17, 2012, their portfolio grew to $108,441, while the lump-sum investor ended up with $148,332. So in this case, it was better to make a lump-sum investment than it was to dollar cost average.

Figure 1. Dollar Cost Averaging Versus Lump-Sum Investing

Source: S&P Capital IQ.

Why the difference? At first blush, the answer lay in the benefit of compounding higher dividend yields, since I also found that the lump-sum approach outperformed the dollar cost average approach for such other high-yielding S&P indexes as the S&P 500 High Quality Rankings index, the S&P 500 Low Volatility index, and the S&P U.S. Preferred Stock index.

Comparing Sector Performance

However, I began to question a high dividend yield as the sole reason for the outperformance of dollar cost averaging versus lump-sum investing when I crunched the numbers for the sectors in the S&P 500 over the same time frame (Table 1). If a high dividend yield were the sole reason, then why was it better to dollar cost average with the telecommunications services sector, yet better to lump-sum invest with the energy sector?

Upon further analysis, I concluded that the decision had more to do with how well each sector rode the undulating waves of bull and bear markets since 2000 in general, and how big of a hit it took during the bear of 2000–2002 in particular, rather than the standard deviation of monthly returns during the prior 12+ years. In seven of 10 cases, it was better to be a lump-sum investor than to take the more gradual dollar cost average approach, as the difference between the ending values ranged from less than 5% for the financials and industrials sectors to more than 20% for the consumer staples and utilities sectors, and in excess of 45% for the energy sector. By looking at the total returns during each of these beatings and bounce-backs, I found that the dividing line between dollar cost averaging and lump-sum investing, in eight of 10 cases, was determined by a descending sort of all sectors based on their cumulative total returns from 2000–2012. Only in the case of the consumer discretionary and financials sectors was this not true, possibly because of their relative outperformances during the bear market of 2000–2002 and comparative dividend yields.

Table 1. S&P 500 Sector Total Returns: Dollar Cost Averaging Versus Lump-Sum Investing: 2000–2012

S&P 500 Sector	Std Dev	Cur Yield (%)	Dollar Cost Averaging ($)	Lump-Sum Investing ($)	% Dif	Bull & Bear Returns (%)				2000–2012
						2000– 2002	2002– 2007	2007– 2009	2009– 2012	Total Return
						2000– 2002	2002– 2007	2007– 2009	2009– 2012	(%)
Energy	20.88	1.9	132,430	193,781	46.3	-15.7	274.6	-42.7	80.3	226.4
Consumer Staples	11.77	3.0	101,653	126,865	24.8	24.8	61.9	-25.9	76.3	163.8
Utilities	17.17	4.2	92,579	115,042	24.3	-32.1	175.5	-37.7	58.4	84.6
Health Care	14.29	2.3	78,922	91,258	15.6	-6.9	53.7	-34.6	67.8	57.1
Materials	22.84	2.1	99,078	113,156	14.2	-16.7	178.6	-55.6	115.8	122.4
Industrials	19.85	2.4	83,033	87,071	4.9	-29.5	122.9	-58.3	131.7	51.8
Financials	24.14	1.7	45,516	47,157	3.6	-11.2	84.8	-76.5	105.7	-20.6
S&P 500	16.50	2.1	75,611	67,247	-11.1	-43.8	108.4	-50.9	97.1	15.9
Consumer Disc.	19.88	1.7	86,845	76,469	-11.9	-36.3	70.3	-51.0	157.7	36.9
Telecom Services	22.18	5.8	68,499	36,405	-46.9	-73.9	166.9	-41.8	57.1	-36.4
Info Technology	28.49	1.1	77,153	36,144	-53.2	-80.3	151.8	-50.6	123.5	-45.1
Source: S&P Capital IQ. Past performance is no guarantee of future results.

Looking Ahead

So what does this mean going forward? Lump-sum investing is probably the better way to go, in my opinion.

Should the U.S. equity markets continue to suffer through a secular bear market for the next few years, as they have since 2000, the consumer staples, energy, health care, and utilities sectors may again be superior performers, due to their relatively high dividend yields and defensive characteristics. Each of these groups showed better results from the lump-sum approach than from dollar cost averaging since 2000. Yet if we find that the S&P 500 is in the embryonic stage of a new secular bull market, we will probably look back and conclude that it would have been better to lump-sum into the cyclical areas.

So while the answer is still the same—lump-sum investing over dollar cost averaging—the outstanding question is whether a secular bull or bear awaits.

Discussion

Donald from WA posted over 14 years ago:

I seem to be missing something here. As I understand the analysis, the investor had $58,000 on day one. In one case he invested the entire 58K. In the other he invested 10K and hid 48K under the mattress and pulled out 1,000 each quarter and invested it. Wouldn't most investors keep the funds not yet in the stock market in what he considered a safe investment that at least provided some return. I didn't run the numbers but it seems to me that the results could change substantially.

David from KY posted over 14 years ago:

Donald, you make a very good point. At times during that 12 year period, CD and Money Markets were paying pretty good rates. The investor would have received some interest on his/her money. It should be included in the analysis.

Charles from IL posted over 14 years ago:

I corresponded with Sam about questions posed above. He wanted to make to clear the purpose of the study was not to see how much marginal cash an investor could squeeze out of one strategy versus another, but rather to show the power of compounding (as well as the benefit of avoiding of substantial loss) on an investor's choice of investment vehicles. He further noted that in a near zero-interest rate environment, a money market or a savings account would not have offered much more than security of principal. -Charles Rotblut, AAII

Michele from FL posted over 14 years ago:

usually the choice to lump-sum or dollar cost is out of our hands. Most people that I know never have a lump of money but save it from month to month so they generally always only are able to dollar cost average unless they just sold out of a position.

Ted from NJ posted over 14 years ago:

If we dramatically reduced the time series to 1 to 3 years. Then take into effect short term interest alternatives and tax implication Then comapre many iterations over bull and bear markets. Then, maybe we can start to learn something about this question.

Leonard from NY posted over 14 years ago:

I used $ cost averaging to rebalance my portfolio to my % allocation for stocks during and after the stock market decline in 2008. It was more comfortable to do than to invest a large lump sum in the stock market not knowing whether the stock market had reached bottom.

Thomas from CA posted over 14 years ago:

I ran the numbers and found that over 12 years the amount of interest that would be earned on the $48,000 not invested initially would be $6,148 at the average return of the Vanguard Prime Money Market Fund of 1.95% per year over the past 10 years. Since 1975 the Vanguard MMF has earned an average return of 5.79% per year. At that rate the $48,000 would have earned $21,053.

James from Vermont posted over 14 years ago:

Thomas, Thank you for running the numbers. In calculating your returns, did you assume a $1000/quarter withdrawal of principle from the $48,000 put in the MMF's? Clearly adding the 'risk free' money earned by the $48,000 in the Dollar Cost Averaging scenario is the only meaningful way to compare the two approaches. To ignore, as is suggested by Charles' communication with the author seems silly. That may sound harsh, but what is the point of comparing investment strategies other than to "see how much marginal cash an investor could squeeze out of one strategy versus another". I'm going to speculate that the average reader of AAII already understands the concept of compounding. The article used real world returns for one strategy, but ignored real world behavior for the other. I doubt the author would use his own article for any sort of guidance.

William Greene from NY posted over 13 years ago:

To: Charles Rotblut, AAII Do we have anything like Sam Stoval's S&P's Capital IQ database available to us here at AAII?

Charles Rotblut from IL posted over 13 years ago:

William, You can find S&P Capital IQ research on the Markets page at http://www.aaii.com/markets But we do not publish Sam's research on a regular basis. -Charles

Bruce Oakley from MI posted over 13 years ago:

The message seems to be that if you drop your money in the market when it is low you are likely to do well. Picking one time point hardly qualifies as an analysis of the two alternatives.

Alessandro Martelli from CA posted over 13 years ago:

Don't forget VALUE averaging == http://en.wikipedia.org/wiki/Value_averaging == invariably superior to COST averaging (and simulations show it has better risk adjusted returns than big-bag investing, too).

Jon from VA posted over 13 years ago:

This article is totally inadequate and mis-informing. The arbitrary date range chosen has a huge impact on results which is the entire reason to dollar-cost-average in the first place. You don't know what is going to happen in the future as much as you might have reason to expect one thing something else could happen. In addition to the timing of the initial investment the timing of the dollar cost averaging has a big impact as well. Suppose I invested lump sum the week before the most recent major crash. Or I invested 20 grand just before, then waited a quarter and put in 20, then another quarter and 18. The dollar cost averaging would have helped tremendously more than likely. It would be even better likely if you waited 6 months between each investment. By timing the initial lump sum before the peak you would immediately get hammered. Same as if you time it before a big gain you would end up better off lump sum probably. But at least you would buy the most with your 20k when it was cheapest. Not to mention the fact mentioned above that the other money is more than likely not sitting under the mattress and we tend to come into it in small chunks vs large ones. I can't believe the author's conclusion was drawn off of such a pitiful analysis then posted on here for people to "learn" from. The fact of the matter is this is a huge generalization which could produce totally different results based on timing and that is the argument FOR not against dollar-cost averaging. -Disappointed.

Rick Miller from IL posted over 12 years ago:

In our finance discussion group someone mentioned this article, so I just recently read it. As I see it, there are three variables: the starting date (Dec 31, 1999), the length of time for dollar cost averaging(twelve years), and the investment interval (quarterly). Change any of those variables and, more likely than not, you'd get a different result. Yet none of these variables was changed. So we have a conclusion drawn from a data set of one. Not nearly rigorous enough. No wonder they still teach dollar cost averaging in finance texts. Wouldn't broker/dealers like us to lump-sum invest? Wouldn't that benefit them? Any conflict of interest here?

John H from FL posted over 11 years ago:

I think this topic could use much more research. There will likely be many cases where someone receives either a lump sum payout from a pension plan or an inheritance. The question is very important to someone who has received such a payout. If there is more research out there, I would like to read it. Personally, if I were to receive a lump sum, I would be nervous investing all of it in the stock market at one time, especially where the markets are in Sept 2014.

You need to log in as a registered AAII user before commenting.
Create an account

Email Post Tweet Share Save

FREE REPORT

Top Tech Stocks for 2026

Get your free copy of our special report analyzing the tech stocks most likely to outperform the market.

Download the FREE Report Here: