I recently revisited one of my favorite investing websites/tools: www.etfreplay.com. This website has a neat little tool that allows one to backtest a portfolio of ETFs. The most amazing thing about this website is that it completes a backtest in a very short amount of time. What could take an hour or more using specialized backtesting software only takes a few seconds for etfreplay to finish. Unfortunately, the backtester doesn't include an optimizer - the user must supply his or her own ETFs and their weightings.
The ETF backtester's benchmark is automatically set to the S&P 500 (specifically, the SPDR S&P 500 Index ETF (SPY)). One day I had the idea of constructing an ETF portfolio that would outperform the S&P 500 over 1 year, 2 year, and 3 year periods (ending at the close of the most recent trading day, which as of this writing is Friday, March 2, 2012). As part of my requirements, I also wanted my portfolio's volatility to match or be lower than the S&P 500's volatility for each of the 3 time periods. Finally, I wanted to find a way to include Apple (AAPL) into my portfolio. Everyone knows this stock has been a strong performer. While some people doubt that it can continue to climb, the consensus seems to be that it will indeed go much higher. However, this article isn't about the merits of Apple. I only wanted to find a way to include it in my portfolio and capitalize on the company's success.
Before I started my little exercise, I wanted to lay down a few rules (the last two are indirect consequences of the website's use and not being a subscriber):
- Go long only (no shorting of ETFs).
- No ETFs employing leverage should be used. I won't get into the pluses or minuses of using such ETFs, but they can be dangerous if used incorrectly.
- The ETFs must be fairly common and easily recognized (no outside the mainstream, niche ETFs).
- Do the best to diversify across asset classes while also limiting an individual ETF's exposure to no more than 30% of the portfolio.
- The ETFs must have a trading history of at least 3 years.
- Minimize the number of ETFs (they can't be more than 5 - subscribers can play around with 25).
I will explain the use of each of these ETFs and how each affects the portfolio's total return and volatility:
- iShares Russell Midcap Growth (IWP): When I started the entire process, I immediately chose this ETF as one of the core holdings. Over long periods of time, midcap growth stocks are an excellent choice. They are more nimble than large cap stocks and not as volatile as small cap stocks. Additionally, growth stocks have high net margins to keep momentum going. IWP affects the portfolio by increasing both total return and volatility.
- Vanguard Barclays Govt-Credit Mix (6-7yr) (BIV): When it came time to choosing a bond ETF, I decided to go with intermediate term bonds. This avoids the volatility associated with long term bonds and the low yields of short term bonds. It's not the case that low yields are bad, but rising inflation will erode those yields. BIV affected the portfolio by drastically lowering volatility. Additionally, it boosted returns in 2011 - a year that wasn't so good to be in stocks.
- Vanguard MSCI U.S. SmallCap Value (VBR): Adding this into the mix was sort of like adding nitro to a fuel mix: it gives a pop to total return but does so at the cost of increasing volatility. I like how its small, 5% weighting is used to minimize what really comes down to speculation, as small cap stocks are very volatile. At the same time, the value part of this ETF minimizes high volatility by concentrating on stocks with low P/E ratios. The use of this ETF boosts return in good years but magnifies losses in bad years.
- Vanguard MSCI U.S. REIT (VNQ): This ETF is similar to adding more stocks to the portfolio, but with the added insurance of above average dividends. This increases total return and only modestly increases volatility. The effect of adding this ETF is somewhat similar to adding the bond ETF: the return is boosted in 2011. However, as a side-effect, the portfolio suffers when real estate performs sub-par, such as in 2009. The 15% weighting magically doesn't throw off the total return in that year.
- PowerShares Nasdaq-100 Index (QQQ): The use of QQQ is quite an anomaly in that it both increases total return while decreasing volatility, and rather dramatically (also having a drastic effect on improving the portfolio's Sharpe ratio). As a matter of fact, the portfolio's returns would have been astronomical if its entire stock component consisted solely of this ETF. Unfortunately, such a portfolio wouldn't be very diversified and it violates my rule of not having a single ETF comprise more than 30% of the portfolio. I originally used this ETF as a way of incoporating Apple into the portfolio. As a result of QQQ's weighting of 30% in my ETF portfolio and the current weight of Apple in QQQ at 15.65%, Apple's weight in my portfolio is about 4.7% (.3 x .1565 = 0.04695). This tilts the entire portfolio towards Apple, as no other stock component of any other ETF has a higher weighting. Thus, this portfolio truly is an Apple-lover's ETF portfolio.
The weighted average net expense ratio of this portfolio is 0.1865, which actually isn't all that bad. Coincidentally, it is actually just about twice the net expense ratio of SPY at 0.09.
Total return and volatility results are shown in the following charts:
The portfolio's total return is indeed greater than that of the S&P 500 for each of the three time periods.
Volatility is either less than or equal to that of the S&P 500 for each of the three time periods.
Additional portfolio statistics are shown below:
One can see that the "Apple-Tilted ETF Portfolio" has a superior CAGR (compounded annual growth rate) and Sharpe ratio when compared against SPY. Additionally, the portfolio has lower drawdowns. Down below, annual performance numbers are better than SPY's for each year from 2009 to 2012 (with 2009's return calculated from March 2004).
As an added bonus, in 2008 (a horrible year for stocks) the portfolio would have performed better than the S&P 500 in terms of total return and with lower volatility. Additionally, the portfolio would have matched the S&P 500's return over the past 3 months of 2012 and with lower volatility (a period when many investors, including myself, have chosen to be out of stocks).
Of course there is always room for improvement. I wouldn't exchange any of the ETFs for any others, but as an experiment I would try to reduce the weighting of IWP (the midcap growth ETF) and put some of this capital to use in an oil-oriented ETF. I would favor an ETF of stocks instead of an ETF of oil futures. An ETF of stocks would have less volatility than an ETF of oil futures. It would mute the effects of worldwide supply/demand, U.S. inventories, geo-political events, and the occasional pipeline explosion. Moreover, the price of oil will continue to rise indefinitely, and companies will pass on this price increase to their customers. An ETF of oil stocks could also be used as a hedge against the dollar, as oil barrels are denominated in U.S. Dollars.
The exercise I went through in using the ETF backtester shows that one can indeed outperform the S&P 500 while tilting a portfolio in the direction of an individual stock - in this case it is Apple. As a secondary objective, I also learned how each asset class affects a portfolio's performance statistics across different investing environments. Most importantly, I had fun!
Apple stock has been a strong performer and more than likely will continue to rise, so why not take advantage of this fact in one's own portfolio of ETFs? Such a portfolio has a distinct advantage - it allows its owner to have exposure to Apple without having an outright position in Apple stock. However, one must also keep in mind the disclaimer that is on almost every broker's website: "Past performance is not indicative of future results." Superior performance in the long term can't be expected, but one can reasonably expect good results with this ETF portfolio while Apple stock is still going strong. It will be interesting to see how this Apple-Tilted ETF portfolio will perform in the future.
Disclosure: I own AAPL indirectly through my employer as part of the holdings of the State of New Jersey's pension fund.