I spend a lot of my time reading new research papers and evaluating new investment products. 99% of new investment products fail to meet my high bar as something I will put my client’s money in. My due diligence process is very thorough. It starts by looking at the underlying research for which the strategy is based. I have to be convinced that the research has been fully vetted by the academic community. I also have to be convinced the excess return shown in the research is not the result of data mining and will persist in the future. Once these criteria are satisfied I look to see if the strategy can survive real world transaction costs and management fees. Communication with the portfolio managers is key. I can get a better understanding of the mechanics of the fund and ask questions.
I recently completed my due diligence on a new mutual fund by AQR – QSPIX. Today I will give an overview of the research surrounding this new fund. In a subsequent post I will show how it fits into existing portfolios we build at Grand Street Wealth Management, LLC.
If you look at the portfolios that most people build it is usually some mix of stocks and bonds. The classic one being 60% stocks and 40% bonds. This is the ratio used by most pension funds and other institutions because it has provided about a 5% real return (over inflation) with a standard deviation of about 10%. If you look more closely at what drives the return of the portfolio it is almost completely dominated by the return of stocks. In fact, depending on the time period you examine, 90%-95% of the return and risk of the portfolio comes from the stock allocation. Said another way, even though you have 60% of the portfolio in stocks, 90% of the return of the portfolio can be explained by your exposure to stocks. By owning this portfolio you are basically taking one type of risk: the return of stocks (also called the equity risk premium or beta). The bond portion of the portfolio dampens the volatility of the stocks.
Academia, over the last 40 years, has discovered other drivers of return besides the basic stock and bond returns. Below is a list of the four style premiums that I believe are the most robust and are captured by QSPIX. They have shown persistent long-term performance, across multiple unrelated assets, in different markets and in out-of-sample tests.
Value is the tendency for relatively cheap assets to outperform relatively expensive assets. For stocks the metric used to define “cheap” is usually book-to-market, or the book value of the company divided by its market price. Others metrics use earnings, cash flow or sales relative to price. For bonds, one way to define value is to look at 10-year government bonds (minus the forecast inflation) over the next 12 months. “Cheap” or “bonds with good value” would be bonds with high real yields. In currencies you can use measures of purchasing power, and with commodities you can use 5-year price reversals as measure of value.
Momentum is the tendency for an asset’s recent relative performance to continue in the near future. For stocks the most widely used metric is to define momentum stocks as those that have outperformed over the last 12 months. Other metrics are earnings momentum and changes in profit margins. Similar relative performance can be applied to bonds, currencies and commodities.
Carry is the tendency for higher yielding assets to provide higher returns than lower yielding assets. It is best known among currency traders. A carry strategy goes long or lends in higher yielding currencies and goes short or borrows in lower yielding currencies. For stocks you can look at the dividend yield, which is similar, but not identical to the value premium. For bonds and commodities, carry strategies take advantage of the futures price curve.
Defensive is the tendency for lower risk and higher quality assets to generate higher risk adjusted returns. In stocks it looks at forecast betas (low being good) and measures of quality such as high profitability, low leverage and stable earnings growth. In bonds shorter maturity bonds have shown better risk adjusted returns than longer bonds.
Academics still argue about why these premiums exist. Some explanations use risk based stories, arguing that you are compensated for these returns by taking some specific risk. Other explanations use behavioral stories, saying there is an inefficiency in the market due to human behavior, either self-imposed (irrational behavior) or mandated (not allowed to sell short or use leverage).
Regardless of why they exist, most would agree they have existed and will persist in the future.
At Grand Street Wealth Management we have been capturing these premiums (except for carry) for years, but only in stocks. Up until now there was only one way to capture these premiums, to tilt your portfolio away from market weights towards these other premiums. The first iteration of this was to own a total market index and combine it with a small cap value fund. Next, we combined a large cap momentum fund and small cap value fund to capture the momentum premium. In 2013 we added AQR’s core strategies which builds the funds from the ground up to capture the value, momentum, and profitability (similar to defensive above) premiums simultaneously in one fund.
Since all of these strategies are long only and are fully invested in stocks they also capture the entire equity risk premium as well. Being long only also only allows you to capture part of the other style premiums. For example AQR’s US Large Cap Core portfolio only captures about 22% of the value premium and 18% of the momentum premium, but 100% of the equity risk premium. Because of this, even these portfolios are dominated by the return of stocks. In fact 88% of the return of AQR’s US Large Cap fund can be explained by the equity risk premium. The remaining 12% is attributed to the other style premiums.
So even the portfolios I build are dominated by the return of stocks, albeit much less than most portfolios. QSPIX has the potential to change this. It seeks to capture these premiums in a long/short stock market neutral way. It cleverly simultaneously captures all four premiums above. It does this across multiple asset classes and countries. It looks to invest in 1500 stocks across major markets, 20 country indices from developed and emerging markets, 10-year bond futures in 6 markets, short-term interest rate futures in 5 markets, 19 currencies in developed and emerging markets and 8 commodity futures. To capture the momentum premium, for example, it will go long high momentum stocks/bonds/currencies/etc. and go short low momentum stocks/bonds/currencies/etc. It will do the same capturing the value premium going long value stocks/bonds/currencies/etc. and go short expensive stocks/bonds/currencies/etc. Doing this simultaneously across all four premiums adds efficiently since it eliminates overlap.
Since the four premiums are uncorrelated with each other the ensuing portfolio is well diversified. It should deliver a steady stream of returns. I calculate this portfolio has an expected annual return of about 7% with a standard deviation of about 10% after all costs and fees (the raw academic data shows an annual return of 26%). More importantly, it does this from a completely separate source of returns than stocks and bonds. In fact, the correlation of QSPIX to a total stock index and a total bond index is basically zero.
In my next post I will show how adding this fund to the existing Grand Street Wealth Management portfolios has the potential to keep the same or higher level of returns while further reducing volatility. These new portfolios will also rely much less on the return of stocks as the main driver of returns.