Long-Term Dollar Cost Averaging Investment Strategy In S&P 500 (2024)

Overview

Much has been written about the attractions of buying low-cost passive index funds. Active fund managers (Warren Buffett, in his 2013 letter to shareholders), index fund managers (John Bogle, Founder of the Vanguard Group), and academics (Jeremy Siegel, Professor at the Wharton School and author of Stocks for the Long Run) have all encouraged investors who have limited time or expertise to buy and hold a broad diversified index like the S&P 500. This is effectively a long-term play on the vitality of the American economy, which Buffett advised investors in his 2020 letter to shareholders to "never, ever bet against".

According to CNBC, over 64% of large-cap active fund managers underperformed the S&P 500 over a 1-year period and 85% underperform over 10 years. The percentages of mid- and small-cap fund managers that have under-performed their benchmarks are even higher, suggesting that the average investor is likely better off investing in benchmark indexes like the S&P 500.

Figure 1: Percentage of fund managers underperforming their benchmarks over the last 10 years

Source: CNBC

For those who subscribe to the advice of Messrs. Buffett, Bogle, and Siegel, the SPDR S&P 500 Trust ETF (SPY) (informally called Spyders) is an efficient way to get exposure to the S&P 500 due to its exceedingly low (0.09%) fees, although taxes on dividends is one of the factors that may cause total returns to be slightly lower.

The S&P 500 is on the cusp of another all-time high (figure 2), so this is not the time for cash-flushed investors to throw caution to the wind, back up the truck, and load up on the index. A dollar cost averaging and long-term hold strategy seems far more prudent. Furthermore, as most 401(k) plans offer participants a low-cost S&P 500 index alternatives (e.g., the Fidelity Spartan 500 Index Fund, Vanguard 500 Index Fund, Schwab S&P 500 Index Fund, among others), this strategy can potentially work equally well for those who made the commitment to set aside a fixed amount from their monthly paychecks to fund their retirement plans.

Figure 2: Chart of the S&P 500 index

Source: finance.yahoo.com

But how has this dollar cost averaging strategy performed over past market cycles, and how much volatility (i.e., drawdowns or unrealized losses) would an investor have to endure through severe market downturns?

I decided to do some back-testing to find out.

What is dollar cost averaging?

According to Wikipedia, dollar cost averaging is an investment strategy that aims to reduce the impact of volatility on large purchases of financial assets such as equities.

Suppose an investor has $100,000 to deploy. The investor could attempt to invest the entire amount in one lump sum by timing the market, but the investor's returns would be highly dependent on the market level at the point in time the investment was made.

Alternatively, the investor can mitigate market timing risk by deploying the $100,000 in equal amounts into the market at regular intervals (for example, $1,667 every month, for a total of $20,000 each year, over 5 years).

Wikipedia notes that:

[dollar cost averaging] seeks to reduce the risk of incurring a substantial loss resulting from investing the entire lump sum just before a fall in the market. Dollar cost averaging is not always the most profitable way to invest a large sum, but it is alleged to minimize downside risk."

The technique is said to work in markets undergoing temporary declines because it exposes only part of the total sum to the decline. The technique is so called because of its potential for reducing the average cost of shares bought. As the number of shares that can be bought for a fixed amount of money varies inversely with their price, dollar cost averaging effectively leads to more shares being purchased when their price is low and fewer when they are expensive. As a result, dollar cost averaging can possibly lower the total average cost per share of the investment, giving the investor a lower overall cost for the shares purchased over time."

Back testing the strategy

I ran a model to calculate both the value of the portfolio and the maximum drawdown of a hypothetical investor who deployed $100,000 in equal monthly amounts over 5 years beginning in January 2007. I then ran two additional sensitivity analyses: one using different start dates of deployment and a second one by varying the amount of capital deployed each month.

Deployment beginning in January 2007

Suppose our hypothetical investor committed to invest $100,000 in the S&P 500 index through Spyders in equal amounts of $1,667 at the end of each month over 5 years beginning in 2007.

The first purchase would be made at the end of January 2007 (figure 3, line 1) when the S&P 500 was at 1437.5. As each SPY share trades at 1/10 of the S&P 500 index, the investor is able to buy 11.59 shares ($1,667 / $143.75 per share) of Spyders. The following month, on February 28, 2007, the S&P closed at 1409.3, so the investor is able to buy 11.83 additional shares with the $1,667. At the end of February, the investor will have a total of 23.42 shares, which when combined with the cash balance of 966,667, gives the overall portfolio a value of $99,967 (line 2).

After buying shares at the end of the month for the entirety of 2007, the investor will own 135.4 shares, which when combined with the remaining cash of $80,000, gives the portfolio a value of $99,795 (last line).

Figure 3: Investor's shares and cash balance in 2007

Source: created by author using publicly available stock prices

The number of shares the investor buys is inversely proportional to the level of the S&P 500 - if the S&P 500 declines for the month (figure 4, solid red line), the value of shares the investor owns declines, but the investor will be able to purchase more shares for the $1,667 deployed that month (red dotted line). Conversely, if the S&P 500 rises for the month, the value of shares the investor owns appreciates, but the investor's $1,667 will buy fewer shares that month.

Figure 4: S&P 500 level vs shares acquired

Source: created by author using publicly available stock prices

As the investor maintains the strategy of buying $1,667 of Spyders every month for 60 months, the $100,000 will be fully deployed at the end of 2011, at which point the investor would have acquired 838.4 shares (figure 5, red dotted line, right y-axis).

For the investor who began acquiring shares in January 2007, figure 5 also shows the total value of the SPY shares owned (solid blue line, left y-axis), the cash balance (solid red line), and the total combined value of the SPY shares and cash (solid green line).

Before I proceed further, two points of note: (1) the SPY shares pay dividends (currently yielding ~1.3%), but I assumed the investor would need the money for expenses so I did not factor them into the calculations; and (2) after 2012, the cash balance drops to zero, so the value of Spyder shares (blue line) is the equal to the total value of the portfolio (green line).

To a risk averse investor, two metrics are relevant:

1. the total value of the investor's portfolio shares plus cash (if any), and
2. the maximum drawdown (i.e., unrealized losses) in a market downturn

Total value: On September 30, 2021, the last month-end before this article was written, the S&P 500 closed at 4,531, which values the investor's portfolio at about $360,000 (solid green line), or 3.6 times the original cost of $100,000.

Maximum drawdown: During the sharp market decline during the global financial crisis, the total value of the portfolio fell to a low of $82,310 in February 2009 (trough of green line in 2009), representing a maximum drawdown or unrealized loss of $17,690 or -17.7%. The value subsequently recovered and appreciated to the $360,000 at the end of October 2021, but this is the case only for investors that had the discipline and intestinal fortitude to stay the course and continued investing $1,667 in Spyders at the end of each month until the cash balance reached zero.

Figure 5: Investor's capital deployment schedule and values

Source: created by author using publicly available stock prices

Sensitivity analyses on date of initial deployment

The combined value of the SPY shares and cash and the maximum paper loss are highly dependent on the starting date and pace of capital deployment.

If the investor had deployed the entire amount in one lump sum on January 30, 2007, the total appreciation would have been 3.0x (figure 6, line 2), 0.6x less upside than the dollar cost averaging strategy (line 1), and the maximum drawdown of -48.6% would have been far higher than the dollar cost averaging strategy's maximum drawdown of -17.7%.

If the investor were able to pick out the March 31, 2009 market trough and deploy all the capital that day (which I argue would be just about impossible to do), the portfolio would have appreciated to 5.8x by September 30, 2021 with no drawdown (line 3). Conversely, if the investor were unlucky enough to deploy the entire amount in one lump sum on October 31, 2007-the pre-global financial crisis market peak-the portfolio would have appreciated by 2.8x, but the maximum drawdown would have been a gut wrenching -52.2% at the March 31, 2009 market trough (line 4).

Figure 6: Comparison of returns and maximum drawdown for different strategies

Source: created by author using publicly available stock prices

To find out how the start date of a dollar cost averaging strategy can affect returns and maximum drawdown the strategy, I ran a backtest using start dates for every month since January 2007. Figure 7 shows, for each start date of deployment (represented on the x-axis), the highest NAV achieved and the NAV as of September 30, 2021 (solid red and solid blue lines, left y-axis), and the maximum drawdown (dotted red line, right y-axis).

A few interesting and noteworthy observations:

1. An investor who began deploying the $100,000 using the dollar cost averaging strategy on Jan 30, 2007 would have a higher current NAV today ($360,000) than if the investor had begun deploying the capital on March 31, 2009, the month the S&P 500 bottomed out ($340,000). This is because the average cost per Spyder share for an investor who began deploying capital monthly for five years beginning on Jan 30, 2007 was $119.2 ($100,000/838.41 shares), compared to the higher average cost per SPY of $126 ($100,000/793.1) had the investor started deploying capital beginning on March 31, 2009 because of the subsequent market run up.
2. An investor who began deploying the $100,000 using this dollar cost average strategy on November 30, 2017 would have suffered a relatively modest 4.1% drawdown on March 30, 2020-the bottom of the COVID-19 market downturn-and will have $130,000 today. (I picked November 30, 2017 as it is the deployment start date with the deepest drawdown since the end of the global financial crisis.)
3. An investor who invested the entire amount in one lump sum on the pre-COVID month-end peak of December 31, 2019 would have suffered a 20% drawdown on March 30, 2020 but will have $133,000 today. (The 20% drawdown is probably lower than what one would have expected, which can be explained by the fact that I am using month-end numbers, and the S&P 500 index of 2,577.5 on March 31, 2020 was a full 15% higher than the pandemic low of 2,237.40 set on March 23.)

Figure 7: Returns and maximum drawdown for different dates of initial deployment

Source: created by author using publicly available stock prices

Sensitivity analyses on amount of capital deployed each month

To determine if there is any benefit to deploying more capital in a down month and less capital in an up month, I ran a backtest scenario in which the investor would invest $2,500 in a month (1.5x the $1,667) if the S&P 500 closes below the previous month's level, and $833 (half the $1,667) if it closes above the previous month's level. As expected, there is more variability in the numbers of share purchased each month under this scenario (figure 8, dotted red line).

Figure 8: Shares acquired each month depending on the S&P 500 level from previous month

Source: created by author using publicly available stock prices

Interestingly, varying the amount of capital deployed according to this strategy resulted in consistently lower returns (figure 9, solid red line) compared to the original strategy of deploying the same amount of capital every month (crossed red line).

Figure 9: Comparison of returns between stable and variable deployment pace

Source: created by author using publicly available stock prices

However, more counter-intuitive was the finding that employing a variable pace of capital deployment actually resulted in greater drawdowns (figure 10, crossed red line) than a stable deployment pace (solid red line) .

Figure 10: Comparison of maximum drawdowns between stable and variable deployment pace

Source: created by author using publicly available stock prices

Concerns

A major concern is a market that goes sideways for the 5 years while the capital is being deployed, but then drops sharply after all the capital has been deployed. Under this scenario, the investor would not have had the opportunity to enjoy unrealized gains or cost average down before the sharp drop, nor would the investor have the capital to buy more shares at the depressed valuations. Mitigating factor: The S&P 500 has not exhibited the behavior of going sideways for 5 years over the last several decades, so this is hopefully a low probability event.

Summary

The use of a long-term dollar cost averaging strategy to invest in Spyders may not be the most profitable way to invest if one were absolutely certain that the market is on an up-trend. However, for those of us who lack the prescience, the strategy improves our odds of a satisfactory outcome while potentially dampening gut-wrenching drawdowns.

My key learnings from the book, Thinking in Bets by Annie Duke, a world leading poker player and decision science author, is that we need to make the best possible decision in the face of incomplete information. Unfortunately, we can still get a bad result in the short run in spite of having the best quality decision because we got "unlucky" (as Seattle Seahawks Coach Pete Carroll did in the closing seconds of Super Bowl XLIX in 2015). However, if it remains the best strategy we have, we should resist the temptation to jump ship just because a few hands didn't turn out well in the short run, and be prepared to stay the course for the long haul.

I strive to unearth less obvious, overlooked, or under-appreciated but intriguing and potentially profitable data-driven insights into companies of service to society. I write to understand, identify deficiencies in, and share my thinking, and would be most appreciative if you call out blind spots, flaws, or gaps in my observations or reasoning.Hope you enjoy my contributions, but please do not take them as investment advice!