All ETF Portfolio Performance – The Power of Dollar Cost Averaging

Back in August 2008, I designed a diversified, all-ETF portfolio before the collapse of Lehman Brothers triggered the huge sell-off of the stock market. The goal I had in my mind when I built the portfolio was to have a portfolio that covers a wide range of asset classes such that it gives me the diversification I need, with both domestic stocks and foreign equities. The funds I chose for the portfolio and their respective allocations are:

  • Domestic stocks: SPDRs (SPY): 40%
  • Foreign stocks: iShares MSCI EAFE Index (EFA): 30%
  • Domestic REIT: Vanguard REIT Index ETF (VNQ): 5%
  • Foreign REIT: iShares S&P World ex-U.S. Property Index Fund (WPS): 5%
  • Precious metal: SPDR Gold Shares (GLD): 10%
  • Domestic bonds: Vanguard Intermediate-Term Bond ETF (BIV): 5%
  • Foreign bonds: SPDR Lehman International Treasury Bond (BWX): 5%

As you can see, the ETFs I picked for the portfolio are all index funds (of course, back then only passive, index ETFs are available. Now there are also actively managed ETFs). The reason for choosing index funds is obvious: To have as many securities in the portfolio as possible.

Looking for a cheap stock broker to trade stocks, ETFs, or options? Check out

Now, one year later, after what the market has gone through in 2008 and early this year, I am curious about the performance of my portfolio had I followed my plan in making regular investments into each funds. Like I have been doing with my mutual fund investments for years, if possible, I wanted to buy each fund every month on the 5th, but the amount of each new purchase will be based on the pre-determined allocation, not the same amount for all funds, as I do with mutual funds. For example, I would invest $1,000 per month in this portfolio, then $400, or 40%, would go to SPY, and 30% would go to EFA, and so on. There’s no particular reason why I buy on the 5th of the month (that’s my date of making investments) and purchasing based on asset allocation is to have each asset in the portfolio stays as close to its allocation as possible (of course, rebalancing is still needed over time).

So how’s my all ETF portfolio doing? From August 05, 2008 to September 08, 2009, the total invested amount is $14,000 in 14 investments. As the following table shows, of the 7 funds, only 3 have positive returns over the past 14-month period, with the best performance being the gold ETF GLD. The worst performer is VNQ, the domestic REIT fund, which lost more than 33%. However, when purchases were made regularly, the picture changed completely.

Fund Begin
Price
End
Price
Fund
Return
Shares Market
Value
Gain Portfolio
Return
Current
Allocation
SPY $124.4 $102.45 -17.64% 60.453 $6193.46 $593.46 10.60% 38.69%
EFA $64.22 $53.54 -16.63% 93.873 $5025.97 $825.975 19.67% 31.40%
VNQ $57.93 $38.51 -33.52% 20.522 $790.33 $90.33 12.90% 4.94%
WPS $37.04 $30.89 -16.60% 29.249 $903.50 $203.50 29.07% 5.64%
GLD $86.08 $97.43 13.19% 16.266 $1584.85 $184.85 13.20% 9.90%
BIV $73.49 $79.24 7.82% 9.326 $738.99 $38.99 5.57% 4.62%
BWX $54.37 $57.54 5.83% 13.382 $770.01 $70.01 10.00% 4.81%

With my plan of investing, the total market value on September 8th when I made last purchases was $16,007.15 without dividend reinvestment. With a total of $14,000 invested, the overall return is 14.34%, better than any single fund return. And looking at each fund’s return in the portfolio with monthly purchases, all but one had double-digit returns from August 2008 to September 2009 (the Portfolio Return column in the above table).

Why such a huge difference? It is because of the dollar-cost averaging (DCA) method used in making the investments. I have discussed dollar-cost averaging in details in the past when I looked at a much longer period of the broad stock market. My conclusion at that time was that DCA isn’t really the good way to invest as far as the performance is concerned because in a up market, shares purchased through DCA become less and less than through a lump-sum. However, it’s a different story in a down market, as the market in the past 14 months. In this case, DCA let me purchase more and more shares as the market kept going down until March when it started to rebound. And once the market starts to recover, the shares purchased when the market was tanking will have a bigger impact on the overall return of the portfolio, as it is clearly shown in the above table.

Of course, using dollar-cost averaging isn’t all about performance. Rather, it’s more about making regular investments a habit, in good time and bad, instead of trying to find out when to get in and when to bail out based on the market condition at the moment :)

BTW, it seems that making purchases based the asset allocation of each fund works pretty well. At the end of the 14-month period, the allocation of each fund in the portfolio only drifted a little bit from the target allocation, despite each fund’s drastic performance.

This article was originally written or modified on . If you enjoyed reading this post, please consider subscribing to my full RSS feed. Or you can also choose to have free daily updates delivered right to your inbox.


Author Info

This post was written by Sun You can find out more about Sun and his activities on Facebook , or follow him on Twitter .

One Response to “All ETF Portfolio Performance – The Power of Dollar Cost Averaging”

  1. Markus |  Oct 13, 2009 at 12:27 am

    I’d drop SPY to about 15%, drop GLD entirely, and get some small-cap international (VSS) and small-cap international value (DLS) and small-cap value domestic (DSV or VBR). You’ll get better returns than you’ve gotten. You also will have MUCH more diversification than you’re getting now.