Warren Buffet famously said to “never invest in a business you can’t understand.” The same advice should undoubtedly be applied to pooled investment vehicles. The advent and rapid growth of the ETF has made it easier than ever for investors to know what they own.
Take the Financial Select SPDR Fund (XLF) as an example. By visiting the XLF homepage investors are able to get a plethora of information on the ETF and its holdings, including:
- Average 5 year earnings growth
- Average price/book ratio
- Average price/earnings ratio
- Average price/cash flow
- Average return on equity
- Average market capitalization
- Median market capitalization
- Smallest market capitalization
- Largest market capitalization
- Premium/discount to NAV
- Performance of the ETF vs. NAV and the underlying index that is tracked by the ETF
- Complete daily holdings data
- Holdings by industry (i.e. financials broken into banks, insurance, REITs, capital markets, diversified financial services, consumer finance, real estate management and development, and thrifts and mortgage finance industries)
All of this information is extremely valuable when it comes to evaluating an investment, understanding where it should fit in a portfolio, and managing the risk of the position on an ongoing basis. And the amount of information provided is certainly better than what has typically been provided with mutual funds – where holdings level data is usually only provided quarterly.
But is it enough to truly know what you own?
With daily holdings transparency, ETF investors do literally know what they own. But is a list of securities with some high level characteristics adequate? We don’t think so. We think that part of knowing what you own is having some sense of how the investment will behave in different economic environments. When we mention behavior of an investment, we mean both individual performance of that security (return, volatility, potential for loss, etc.) as well as performance relative to other exposures in a portfolio.
If we are considering holding both XLF and the Utilities Select SPDR Fund (XLU) , looking at holdings data shows no overlap. Obviously the lack of holdings overlap does not mean that XLF and XLU share no common risks. They both hold large-cap U.S. equities and so will have significant betas to the U.S. equity markets. Looking at the SPDR website, we can make some additional qualitative observations.
XLF’s holdings have a weighted average market-cap of $118 billion compared to just $27 billion for XLU. With smaller stocks, we may reasonably conclude that XLU may be a bit more sensitive to the relative performance of small-cap vs. large-cap equities.
XLF’s holdings have an average PE ratio of 15.8 compared to 17.1 for XLU. Lower PEs may indicate that XLF has more exposure to the relative performance of value vs. growth equities.
While these insights are crucial in portfolio construction, they are incomplete. XLU may have more small-cap exposure than XLF. But how much more? If small-caps outperform large-caps by 2% this year, what will the impact be on XLU performance relative to XLF?
It is also very difficult to evaluate the exposure an ETF may have to risk factors like commodities, interest rates, and credit spreads by examining holdings data.
To provide some more definitive color on the risks that sector investors take, we build a factor-based risk model using an approach similar to that described by Research Affiliates. Since we are focused on U.S. sector ETFs, we decided on 13 factors (detailed below).
We then ran a regression of the performance of each SPDR sector ETF on the 15 factors.
We see that while all nine sectors have consistent positive exposure to broad U.S. equities, certain sectors have large exposures to other risk factors. What may be surprising are the large betas (both positive and negative) to cash and short-term interest rates. For example, Materials has a beta of 1.4 to after-inflation cash returns and a beta of 1.0 to the U.S. equity market. This does not mean that Materials returns are more sensitive to cash than the broad equity market. Comparing betas across factors can be misleading because the volatilities of the factors are not equal. To get a better sense of what factors are actually most important in driving sector returns, we must normalize the betas.
Observations from the normalized betas
Broad U.S. equity performance is the biggest driver of sector returns. Eight of the nine sectors have a beta to U.S. equities of 0.85 of more. The lone exception is energy, which unsurprisingly is driven roughly equally by U.S. equities and commodities.
The low volatility factor is the second biggest driver of sector returns after U.S. equities. Consumer Staples, Utilities, and Health Care – typically considered defensive sectors - have large positive exposures to the low volatility factor (i.e. they benefit when low volatility stocks outperform high volatility stocks). Financials, Materials, Energies, and Technology have large negative exposures to the low volatility factor (i.e. they benefit when high volatility stocks outperform low volatility stocks). Interestingly, the sector exposures to long-term interest rates are very similar with Consumer Staples, Utilities and Health Care being positively exposed (benefit when rates fall) and the other sectors being negatively exposed (benefit when rates rise). Part of this parallel may be due to a positive correlation between long-term interest rates the low volatility factor. However, there may also be a business cycle component.
The value factor is the third biggest driver of sector returns. Materials, Energy, Financials, and Utilities have tended to perform better when value stocks outperform growth stocks. Technology has tended to perform better when growth stocks outperform value stocks.
Consumer Discretionary has the most significant small-cap exposure, which is interesting because it is the fourth largest sector by market-capitalization.
Materials and Energy have significant exposure to the momentum factor.
- Financials have a significant real estate exposure – not surprising since REITs are included in the sector. Energy has a significant negative exposure to real estate. Perhaps higher energy prices, which benefit Energy companies, also constrain household budgets and hurt spending on new and existing home purchases.
In our tactical sector strategies, we rotate around an equal sector weight portfolio. We can run the same type of factor analysis on an equal sector weight portfolio. The results are below.
We see that the equal sector weight portfolio has a beta to the broad U.S. equity market of around 0.95 with small, but positive, exposures to the four main equity factors that have been documented to outperform on a risk-adjusted basis.
In our commentary for the week of April 13th, we answered the question as to why we manage both sector-based and factor-based tactical equity strategies. While the two approaches have some key differences, we see above that our equal sector weight portfolio is actually a close cousin to a factor-based strategy portfolio.
By equal weighting our sector exposures, we get indirect exposure to the same rigorously studied equity factors that we are able to access directly in our U.S. Factor Defensive Equity strategy.
Note: The factors that considered are (1) after-inflation cash return(cash return minus inflation), (2) U.S. equities (S&P 500 minus cash), (3) small-cap (small-cap minus large-cap, (4) low volatility (low volatility minus high beta), (5) value (value minus growth), (6)momentum (high momentum minus S&P 500), (7) short interest rates (1-3 year Treasuries minus cash), (8) intermediate rates (7-10 year Treasuries minus 3-7 year Treasuries), (9 long rates (20+ year Treasuries minus 10-20 year Treasuries), (10) commodities(commodities minus cash), (11) investment grade credit spreads(investment grade bonds minus duration matched Treasuries), (12)high yield credit spreads (high yield bonds minus duration matched investment grade bonds, and (13) real estate (REITs minus large-cap).
In Our Models
Our Tailwinds Conservative model rebalanced this week. The rebalance was mainly the result of changes in our Multi-Asset Income model in prior weeks. Most notably, our positions in bank loans and convertible bonds were increased with the capital coming mainly from reductions in our preferreds and mortgage REIT positions.
DISCLOSURE: The views and opinions expressed in this article are those of the authors, and do not represent the views of equities.com. Readers should not consider statements made by the author as formal recommendations and should consult their financial advisor before making any investment decisions. To read our full disclosure, please go to: http://www.equities.com/disclaimer