Does dollar cost averaging boost or hinder performance?

By Dan Hallett, CFA, CFP

In response to many fee-bashing articles of late, I recently urged both the media and the investment industry to broaden the focus to not only fees but also other less obvious performance drags that are tougher to control.

To explain one factor I stated that dollar cost averaging (i.e. regularly investing a set amount) usually hinders performance due to volatility. Since that’s contrary to the industry’s multi-decade teachings, my over-simplification and over-statement triggered a lot of confusion and many questions. Accordingly, I thought the topic deserved a column of its own.

Elusive advertised returns

Investors have long been frustrated by their mutual fund returns because they almost always lag the performance advertised for the identical funds. To have experienced advertised performance, you’d have to have bought at the beginning of the period advertised and held through to the end with no interim trading. But the vast majority don’t invest this way.

Most people don’t have the luxury of a single large sum to invest today. The best most can do is invest money when it is available. As David Chilton first popularized twenty years ago, pay yourself first with regular fixed-dollar investments – i.e. dollar cost averaging.

Dollar-cost averaging (DCA)

DCA’s main benefit is the prioritization of saving that it imposes on investors – key to long term wealth accumulation. Many have also long portrayed DCA as a no-brainer way to exploit market volatility – a notion that I challenge.

When prices fall, your fixed dollar investment buys more shares or mutual fund units. In periods of rising prices, the same dollar amount buys less. DCA is usually framed as a way to realize better returns than simply betting a large amount on one purchase date – i.e. lump sum investing – but this rarely happens over the long term.

DCA investing is a terrific way to build capital and I use it myself. DCA produces better percentage returns when the average price paid for an investment is less than the price paid with a lump sum. (Note, however, that even in this scenario, lump-sum investing often results in a higher dollar accumulation because it keeps more money invested for a longer period of time.) Given enough time and a growing investment, DCA investing is likely to underperform – but this isn’t always the case.

The performance difference between DCA and lump-sum investing is influenced by time, volatility, the direction of returns and the sequence of returns.

Investor returns

The first table below (click to enlarge) shows return calculations for lump-sum investing (LS) vs. dollar-cost averaging (DCA) for three asset classes over selected time frames.

Since Canadian stocks and bonds produced roughly the same total returns for the five years through October 31, 2011 – but with very different volatility levels – this is a good starting point for a comparison.

Note how investing in volatile Canadian stocks through a DCA strategy did not work well over this recent five-year period. More specifically, DCA investing fell short of LS investing by 2.1 percentage points annually. By contrast, DCA investing in Canadian bonds over the same period yielded the opposite result – a return of 70 basis points annually in excess of investing a lump sum five years ago.

But notice how the third scenario in the above table – U.S. stocks for the decade through October 31, 2010 – saw DCA investing produce 2.9% per year in excess returns (compared to lump sum investing) despite high volatility. This is due to the fact that this time period for U.S. stocks began with more than two years of mostly negative returns, ending that decade roughly where it began.

Sequence of returns

The sequence of investment returns is an often-cited risk when drawing down investments (i.e. in retirement) but it also impacts wealth accumulation. The sequence of returns doesn’t impact a lump-sum investing strategy but it can make or break a DCA strategy. The table below replicates the above scenarios but with monthly returns ranked in order from lowest to highest for the same period. This scenario makes DCA investing shine. While unrealistic, this illustrates the power of such an extreme return path.

Similarly, the next table shows the opposite – i.e. the same five-year period but with monthly returns ranked from highest-to-lowest. This makes DCA investing a train-wreck of sorts, though again reality is unlikely to ever be so extreme.

While sequence of returns is a strong force, volatility gives it its power. In a zero volatility investment (i.e. same return every day, week, month, etc.) the order of returns has zero influence on investor returns since both LS and DCA investing yield the same percentage return.

But add volatility to the mix and the differences grow significantly. And the higher the volatility the more likely DCA investors are to fall short of lump-sum investors. Finally, the longer the investment time horizon, the more likely the investment may be to rise significantly enough to make DCA’s percentage returns inferior to lump-sum investing.

Academic support

The notion that DCA investing results in a long-term performance drag was validated in “Buy High, Sell Low: Timing Errors in Mutual Fund Allocations“, by Stephen Nesbitt, Fall 1995, The Journal of Portfolio Management. While it wasn’t his main focus, Nesbitt’s paper contains a table showing returns of a LS strategy and a DCA strategy for 17 fund categories. Fourteen of the seventeen categories saw lower returns for DCA scenarios than for a lump-sum (LS) investment from December 31, 1983 through August 31, 1994.

DCA investing remains the best way for the vast majority of investors to accumulate wealth. An awareness of the above factors can help avoid the performance drag that often results from DCA investing so that investors and their advisors can work to enhance investors’ long-term returns in the face of so many performance hurdles.