I’ve wondered for some time if investing equally in a broad basket of individual sectors of the stock market could beat the overall S&P 500 index. To test this theory I compared the performance of SPY (a popular S&P 500 ETF) vs a group of ten sector ETFs:
- Financial – iShares US Financials (IYF)
- Healthcare- iShares US Healthcare (IYH)
- Consumer Goods – iShares US Consumer Goods (IYK)
- Consumer Discretionary – Consumer Discretionary SPDR (XLY)
- Industrials – iShares Us Industrials (IYJ)
- Telecom – Vanguard Telecommunications Services (VOX)
- Materials – Materials Select Sector SPDR (XLB)
- Energy – Energy Select SPDR (XLE)
- Technology – iShares US Technology (IYW)
- Utilities – Vanguard Utilities (VPU)
I wanted to run this test during a strong bull market so I started in January 2009, right at the end of the housing bust. The theory is if this methodology can’t beat the market under the best conditions then it wouldn’t be a viable alternative strategy. I tested my hypothesis two ways: 1) uniform contributions each month and a lump sum contribution at the start of 2009.
Test 1: Uniform Contributions
First, for Test #1 a $100 contribution is made on the first trading day of the month for each ETF. Second, a $1,000 contribution is made to SPY on the same date. Below are the cumulative returns:
As you can see, SPY is the winner returning 3.8% more than the combination of ETFs. The table of annual returns below show it wasn’t just one or two years driving the outperformance. SPY was on top six out of the nine periods analyzed:
Test 2: Lump Sum Contribution
For Test #2 an initial $100k contribution is made to SPY and a $10K contribution is made to each of the ten ETFs. Below are the cumulative returns:
Results are quite close, there is less than a 1% difference between the cumulative returns. You can also see that for many years the basket of sector ETFs excelled significantly against SPY. Let’s examine the annual returns to ensure again its not just one or two years powering the results:
Although the cumulative results are close, SPY beat the group of ETFs seven out of nine periods. Test #2 has the same results as Test #1: SPY is the best option.
In both tests SPY bested the basket of sector ETFs. Even with these results I’m not ready to abandon the idea of sector investing just yet. Uniform contributions is a too simplistic approach, I need to add some variation to the system somehow. I’m excited to reengineer this model with some robust rules and criteria to see if I can get a result favorable for sector investing.
I’m aware that this model ignores transaction costs and the fact that you can’t purchase fractional shares of an ETF. My initial goal is to see if I can develop a sector investing system that can consistently beat the S&P 500. If I can construct a system I’ll go back in add the transaction costs and correct for fractional shares and see if the results still stands up.