By: Larry MacDonald
Passive index investing may have lost some of its luster during the bear market of 2008. It seems proponents may have focused too much on relative returns and not enough on volatility and absolute returns.
One way to get a smoother ride besides having a higher allocation to bonds (like the One-Minute Portfolio) may be to use sector indexes as building blocks instead of market indexes. With sector indexes, an investor can bring the risk-reduction process of rebalancing and lowering correlations down to a finer level of detail.
Consider a passive index investor just starting out. They can buy and hold the oldest and most liquid of exchange-traded funds (ETFs): the S&P 500 SPDR (SPY). Compared to other major indexes, the S&P 500 has better diversification and lower volatility. But it would have been even better to have owned an equal-weighted portfolio of ETFs tracking the nine sectors making up the S&P 500:
Materials SPDR (XLB)
Energy SPDR (XLE)
Financials SPDR (XLF)
Industrials SPDR (XLI)
Consumer Staples SPDR (XLP)
Utilities SPDR (XLU)
Technology SPDR (XLK)
Consumer Discretionary SPDR (XLY)
Health Care SPDR (XLV)
The investor would have enjoyed lower volatility/higher returns, as Tristan Yates illustrates in Enhanced Indexing Strategies (John Wiley & Sons, November 2008), a book reviewed in my last column. His table on page 49 shows SPY earned 3.9% annually with 14.1% annualized volatility from 1999 to 2006 while the portfolio of sector ETFs earned 7.2% annually with 13.5% annualized volatility (Yates’ figures are dividend adjusted). This result occurs because of rebalancing, i.e. keeping sector proportions equal over the years — as I understand Yates to say.
Yet, the nine-sector portfolio is not optimal in the sense the ETFs were selected so as to minimize volatility and correlations amongst each other. If only sector ETFs with low volatility/correlations are included, that leaves four: Consumer Staples SPDR, Energy SPDR, Financials SPDR and Utilities SPDR. This basket returned 8.8% per year with volatility of 12.7%, on average. “The results show that it’s better to build [your own indexed portfolio] than to buy [an off-the -shelf package],” writes Yates.
Taking Yates’ line of thought further, someone with a high aversion to volatility might conceivably index the equity component of their portfolio to a basket of non-cyclical ETFs, notably Consumer Staples, Healthcare, and Utilities. Over the five years to Jan. 9, they individually outperformed SPY. Two of them show a positive return for the period and the third a smaller loss than the S&P 500, which was down over 15%.
3 Responses to “ Passive indexing revisited ”
“The results show that it’s better to build [your own indexed portfolio] than to buy [an off-the -shelf package]”
Really? It looks more to like the results say that with the benefit of hindsight you can put together a portfolio to do just about anything you’d like it to do. That a particular strategy worked will over the course of 7 years is no basis for saying X is better than Y. It’s a basis for saying X worked better than Y for a specific interval of history.
The strategy is guaranteed to do one thing over the next 7 years: Cost you trading fees. It may or may not outperform buying the SPY and holding. It may be more volatile and it may be less volatile. It will cost more to manage.
By Meander on Jan 9, 2009
Agreed with Meander. I’d rather see some reasoning beyond back-testing the strategy over such a small window — especially one that is dominated by the dot-com bubble and finishes with a once-in-a-century market crash.
By Patrick on Jan 12, 2009
Meander, Patrick
As you say, the performance may be different depending on the time period (I think Mr. Yates just intended his example as an illustration, not a proof). But something not dependent on the time period, I believe, is the risk reduction one can achieve by rebalancing the relative weights of the sector ETFs. If one sector grows to a large weight, then one’s portfolio gets riskier. Rebalancing at the sector level controls this drift. It adds some additional rebalancing cost — which some investors may prefer to avoid while other, more risk-adverse, investors may deem acceptable (as opposed to owning the broad market and being exposed to the volatility that can result from changes in sector relative weights).
I personally am intrigued by the idea of indexing to non-cyclical sectors (which can likely be done with little or no rebalancing). Warren Buffett Steven Jarislowsky, and other seasoned long-term investors seem to have a preference for companies from these sectors. They should be a smoother ride and there is less interruption in the compounding effect. It may or may not be better than indexing to the broader market in the end but it’s easier on the nerves.
By Larry MacDonald on Jan 12, 2009