The term ‘smart beta’ has only been around since perhaps 2007, although it seems that we have been hearing about these strategies for much longer. Smart beta sits between passive and active investing and aims to leverage the benefits of both while avoiding the pitfalls. Smart beta strategies have several common characteristics: they are rules-based and transparent at a high level; they claim to outperform the market over the long term; and they have varying levels of quantitative influence. But why should investors consider them, and are they really that smart?
What is beta?
The beta of a portfolio is its exposure to the market capitalisation weighted portfolio (cap weight). This is usually the benchmark like the S&P/ASX200 or MSCI World. Why the cap-weighted benchmark? Well, it is the net position of everyone in the market. Beta is measured by regressing the portfolio returns against the market return. If the market is up 2% then down 3%, and your portfolio is up 1% then down 1.5%, then your portfolio has a beta of 0.5. That is, it moves half as much as the market. If you hold stocks in your portfolio in equal proportion to the cap-weighted benchmark, then you will have a beta of 1. For more information I suggest reading Markowitz (1959) and then Sharpe (1964).
So, really, what’s wrong with cap weight?
The market’s returns are the average of all investors, importantly, before fees and taxes. Fund managers will manage a cap weighted portfolio for very low fees; it has massive liquidity and capacity, so why doesn’t everyone just invest in it? If cap weight is the average portfolio of all investors, and investors can have very different reasons for buying, selling, holding, or not holding a stock, then to invest in cap weight means you are receiving the same investment objectives, both long and short term, as the average investor.
Supporters of smart beta alternatives argue that cap weight will always overweight overpriced securities and underweight under-priced securities, and the effect of this is largest just when you don’t want it. The tech boom and bust is an example where the market can massively misprice entire sectors over long periods of time. Investing in a cap weighted index has you underweight the wrong sectors before the boom, but more importantly, overweight the wrong sectors at the bust. Furthermore, the market can be irrational for long periods of time.
We want more beta
Theory tells us that the risk premium rewards you for taking on more risk, so for decades portfolio managers overweighted stocks with a beta of 1.5 or 2. Simple, hey? Not quite. In his 1972 working paper Robert Haugen first showed that stocks with high volatility (prices jump around a lot) generally had lower subsequent returns. High beta stocks have high volatility. As Haugen’s findings didn’t fit in with the accepted theories, they were ignored for 30 years by many academics and some fund managers.
The beta described above is market beta. We can also measure exposure to other factors such as high dividend yield stocks or high growth stocks. This is the basis of what is known as Arbitrage Pricing Theory, which says that the returns on your portfolio are dependent upon its exposure to many different factors. In the same way that we can increase or decrease our portfolio’s exposure to the market, we can increase its exposure to other factors like energy, gold or changes in interest rates. Thus, you can build a strategy that has your desired exposure to any type of beta.
Some standard smart beta strategies
Historical testing of all commercial smart beta strategies shows that they outperform the cap weighted market over the long term. In the US or Global, this will be in the order of 2-3% per annum over 40-50 years (see Table 1). In Emerging Markets and Developed Small Caps the outperformance is generally 4-5%. All strategies attempt to break the link between a stock’s price and its weight in the portfolio, deliver a diversified portfolio, and give a higher Sharpe Ratio (a measure of the excess return for the risk taken) than the market. Portfolios that have stable target weights through time, such as equal weighted and fundamentals weighted, have low turnover and benefit from a rebalancing bonus.
Equal Weight is the simplest alternative beta strategy; just invest the same amount in each stock, let the weights drift and then rebalance back to equal weight each quarter or year. It sounds simplistic and naïve, but over the long term this outperforms, and in fact it is difficult to build a strategy that outperforms equal weight over the long term. The nice thing about Equal Weight is that it makes no assumptions about expected return or expected risk. The S&P500 Equal Weight Index, launched in January 2003, now has nearly US$4.5 billion invested in related ETFs.
Fundamentals strategies weight companies according to accounting metrics such as total sales and dividends paid, see, for example, Arnott et al (2005). Like equal weight it has slow moving weights to rebalance toward, claiming the rebalancing bonus. However, it also has a moderate tilt toward value and has higher capacity and liquidity. As of December 2012, Research Affiliates had US$74 billion under management in their Fundamental Index strategies.
Equal Risk Weight and Risk Parity make no assumptions about expected returns, but weight stocks according to their historical volatility. Other risk controlling strategies include Low Volatility and Low Semi-Deviation, which have evolved from the work of Haugen. Moving further toward the quantitative end of smart beta are strategies like Minimum Variance, Maximum Diversification and Risk Efficient Index, which build upon the work of Markowitz in the early 1950’s.
Table 1: Long term returns of popular smart beta strategies
1964 to 2012 using 1,000 largest stocks in the United States
|
Strategy
|
Return
|
Standard Deviation
|
Sharpe Ratio
|
|
US Cap Weighted
|
9.7%
|
15.3%
|
0.29
|
|
Equal Weight
|
11.3%
|
18.3%
|
0.33
|
|
Minimum Variance
|
11.8%
|
11.7%
|
0.56
|
|
Maximum Diversification
|
12.0%
|
14.0%
|
0.48
|
|
Risk-Efficient
|
12.5%
|
16.8%
|
0.43
|
|
Risk Cluster Equal Weight
|
11.2%
|
14.6%
|
0.41
|
|
Fundamentals Weighted
|
11.6%
|
15.4%
|
0.41
|
Source: Research Affiliates
Table 1 highlights that all strategies outperform cap weighted over the long term, and that cap weighted is the outlier. Factor analysis shows the strategies outperform for similar reasons: they all have a significant tilt toward value and/or smaller companies (alternatively, cap weight underperforms because it tilts away from value and/or smaller companies). Smart beta is sometimes criticised for simply capturing well known factors. In an industry which has at times valued complexity as superior intellectual thought, simply capturing well known factors in a transparent manner has been undervalued.
So what’s wrong with smart beta?
You will have noticed the phrase ‘outperform over the long term.’ Smart beta strategies will perform differently in different markets environments, sometimes significantly. If there is a flight from risk, then low volatility will outperform, while equal weight might underperform if the flight is toward large cap.
The trap that investors can fall into is ignoring or selling out of a smart beta strategy that has underperformed for a year or two and investing in something else. This locks in your losses, and you might just time it completely wrong so that the next strategy is just starting its own period of underperformance. Recalling that the tech boom lasted years, most strategies that weren’t heavily invested in tech stocks would have underperformed for years. The tech bust, however, would have more than redeemed these losses, assuming that you had fortitude enough to stick to your strategy when others were riding high.
Fund managers will charge more for smart beta than for passive market cap indexing. They would argue that over the long term, after fees and taxes, smart beta outperforms by 1.5% - 2.5% per annum. They would also argue that active managers are in general less transparent and are more focused upon the short term.
Smart beta investing is about understanding the limitations and the timeframe of practical alternative beta strategies, and accepting that outperforming by 2% per annum over 10 years really is world class investing.
Choosing a smart beta strategy
Many large institutional investors have moved away from the core/satellite approach where most of the equity portfolio sits in market cap and the remainder in medium to high conviction active managers. They have allocated less to market cap and active managers and made an allocation in one or two smart beta strategies. When one of the managers or smart beta strategies has exceptional returns the investor rebalances to a strategy that has had market or economic headwinds. But this is done looking over a period of years, where the investor is not concerned whether the profits are moved in a month or in a year.
Smart beta strategies are rules-based and generally transparent, but they all have a quantitative element, with some strategies becoming a little opaque. A fund manager should be able to easily attribute returns, describe why they are increasing weight to a sector, or have sold out of a country. Costs due to trading are important: some smart beta strategies have large capacity, low turnover and are very liquid. The after-tax returns of the strategy also need to be considered, with some strategies having far lower turnover than others.
Smart beta strategies can and will underperform the market cap for periods of time. You need to understand this pattern of performance and have a long term mindset. Measure the performance against similar factors, such as a value benchmark, and check that the claims of the smart beta strategy are holding. The reward is the potential for higher longer term returns in a cost effective manner. Smart beta is a term you will hear more of, and more strategies will come to the market over time.
Adam Randall Ph.D. is a Portfolio Manager at Realindex Investments, an affiliate of Research Affiliates of Newport Beach, CA.