Dollar-cost averaging is a well known term across the investment community, however, for those not familiar with this investment strategy, Investopedia.com defines dollar-cost averaging as:
"The technique of buying a fixed dollar amount of a particular investment on a regular schedule, regardless of the share price. More shares are purchased when prices are low, and fewer shares are bought when prices are high."
Therefore, dollar-cost averaging reduces risk and potentially increases return by averaging into equity market prices over time. The most common dollar-cost averaging decision for many investors is whether to contribute 401(k) savings at the beginning of the year vs. dollar-cost averaging evenly throughout the year.
The best way to validate a long-standing theory is to look at the data. Therefore, I ran a Monte Carlo simulation (400,000 iterations) using two different 401(k) contribution approaches over a 40-year projected period:
Front-loaded / lump-sum contributions -- contribute 100% of contribution limit beginning of each year (e.g., 401(k) bonus election, 100% paycheck deductions)
- Monthly contributions (dollar-cost averaging) -- contribute 1/12th of contribution limit the beginning of each month (e.g., less than full paycheck deductions)
Before I dive into the analysis, I'll explain the assumptions and their overall impact on the decision between front-loaded and dollar-cost averaging (high / medium / low):
- S&P 500 (SPY) Volatility [high] -- the advantage of dollar-cost averaging increases as equity market volatility increases. This is the primary argument for dollar-cost averaging, as investors want to avoid fully investing before the S&P 500 declines significantly only to recover over the following quarter. By not having cash on hand, investors miss out on the ability to buy at lower prices. Therefore, I flexed this assumption to demonstrate the sensitivity that volatility has on portfolio returns depending on contribution method
- Contribution Limits / Inflation [medium] -- inflation impacts future contribution limits. I assumed that the Federally allowed 401(k) contribution limit would grow at 3% rounded to the nearest $500. However, run-away inflation occurring in the last several years before retirement would make dollar-cost averaging more attractive as contributions would represent a larger percentage of total portfolio
- S&P 500 Expected Return [low] -- I've assumed 11% annualized nominal returns based on the S&P 500 average annual total return (capital appreciation and dividends) since 1945. This assumption has low impact on the decision assuming the long-term S&P 500 expected return is positive since front-loaded contributions have more invested days than monthly contributions. Therefore, front-loaded contributions returns should be greater than dollar-cost averaging (holding volatility constant). For example, front-loaded contributions made at opening price on the 1st day of the year have 365 invested days. Dollar-cost averaging contributions (1/12th monthly) made at opening price on the first day each month have 198.5 invested days
The following table shows the median percentage difference of portfolio values between front-loaded and dollar-cost averaging strategies by year (5% monthly volatility is assumed "base case" scenario based on historical monthly volatility):
(click to enlarge)
The above results are similar across different volatility scenarios and years because the median strips away the effects of outliers. Therefore, these results show the "base scenario", which demonstrate that front-loaded contributions outperform dollar-cost averaging by approximately 5%, assuming an expected annual S&P 500 total return of 11%.
However, investors want protection against negative outlier events, therefore, I've also looked at the "worst-case scenario" by taking the bottom 5th percentile outcomes.
The following table shows the bottom 5th percentile difference of portfolio values between front-loaded and dollar-cost averaging strategies by year (5% monthly volatility is assumed "base case" scenario based on historical monthly volatility):
The above results show that dollar-cost averaging works best over short periods of high volatility with large drawdowns. However, over a long time horizon, the benefits of dollar-cost averaging disappear.
Alternatively, the "best-case scenario" (95th percentile outcomes) occurs when large returns occur early in the year followed by near zero returns for the remaining of the year, as shown in the table below (5% monthly volatility is assumed "base case" scenario based on historical monthly volatility):
These 5th, 50th and 95th percentile outcomes suggest that an investor's time horizon is an important determinant for choosing between front-loaded or dollar-cost averaging:
- Long-term investors (front-loaded) -- investors with a time horizon greater than 10 years can avoid dollar-cost averaging under "base case" volatility expectations. They are better off saving their yearly contributions as soon as possible by contributing their bonus or 100% of their paycheck at the beginning of the year.
- Short-term investors (dollar-cost averaging) -- investors with a time horizon less than 10 years and current contributions representing a large percentage of their overall retirement account can benefit from dollar-cost averaging under the "worst case" scenario. However, I would argue that investors close to retirement generally should not be allocating to a 100% equity portfolio but diversify across asset classes, which reduces some of the necessity for dollar-cost averaging.
For 401(k) investors with a long time horizon (greater than 10 years) there is little to no benefit of dollar-cost averaging throughout the contribution year. However, don't be too upset if you've been dollar-cost averaging for the past 40 years, as the difference in portfolio value for a 2012 retiree was less than 4% in favor of front-loaded investing in the S&P 500 total return from 1972 to 2012.
As a result, long-term investors should focus on maximizing the amount saved each year, preferably earlier in the year. However, based on the above results, front-loaded investing does not significantly outperform dollar-cost averaging under many scenarios. Therefore, earlier saving should not impair an investor's ability to meet other financial obligations, as the added benefit of front-loaded contributions does not outweigh the ability to meet life's financial obligations (mortgage, car payments, etc).