Historical data going back 65 years on how equities perform, show proven risk premia or factors. This assertion from academics now employed by Janus Capital in its strategies is tantalizing, but also provocative for investors sceptical of being told there is a simple and easy way of choosing equities.
In a passionate discussion at the Conexus Financial offices in Sydney, Andy Weisman, a former hedge fund manager whose good calls in the heady days of 2008 gave him a small but positive return, makes a persuasive case for these newly discovered paradigms. Now the chief investment officer for liquid alternatives at Janus Capital, the starting point for his beliefs come from widely available research into inefficiencies in market indexes.
A turning point
Weisman sees the information available to investors as a game changer. “We’re at a point in history where people are finally figuring out they have to use these techniques and not simply rely on a single dominant [equity] risk factor,” he says.
Some of the research comes from work done by Kenneth French, a US academic, who has tracked the returns from stocks over the last 65 years and come to the conclusion that small cap stocks and value stocks tendto outperform others. Together with fellow academic Eugene Fama, he has created what is known as the Fama-French three factor model, which claims to explain 90 per cent of a diversified portfolio’s returns.
In this context, Weisman posits: “When you’re constructing a portfolio, do you just simply want to leave it to the winds of the equity risk premium over time, or do you want to make use of things that are a little bit more statistically independent? Such as size being somewhat more statistically independent? Or owning baskets of stocks that are cheap relative to their fundamentals and being short baskets of stocks which are rich relative to their fundamentals?”
Weisman breaks down equities into a set of risk factors, with statistically proven rewards to investors who take note of them. He lists the bias towards value, momentum and smaller size stocks as positions that can all pay off without any great application of stock picking or skill.
“Is there a disutility transfer taking place in the marketplace that is obvious that I can point to, that substantiates why I deserve to be compensated for taking that risk,” said Weisman. “That’s really a fundamental question you have to ask in the context of every risk premium.”
Joshua Bloom, portfolio manager at Sunsuper, asked whether this disutility transfer was something that was either predictable or repeatable.
Richard Lindsey, chief investment strategist, liquid alternatives, Janus Capital said this could not be known with certainty. “You have to look and say someplace there’s a transfer occurring. Where does that transfer exist? And why should it exist? And if you can come up with a reasonable explanation, one that you can believe in, then you can have more faith that that’s a fundamental risk.”
The premium on size
Weisman explains the workings of the small stock risk premium as follows. “If you just had large cap and small cap stocks and we formed a universe and then you had a series of time dependent stochastic processes. Every once in a while one of the small caps goes deviating out of small cap land and one of the large caps is going to come down.”
He explains the stocks with the same expected rate of return that are promoted from the small cap index to the large cap index, are effectively sold the stock at a high price, while large stocks going in the opposite direction are sold at a low price.
“So systematically, purely as a function of a rebalancing exercise, small cap sells high and buys low and large cap actually sells low and buys high,” he said. “So if you have some guys in your office who are handy with Matlab, they can program this up and simulate return outcomes over time, which are completely consistent with neutral assumptions about the returns, with the size risk premium over time.”
He adds that the process can be approached in another way. An investor could pick the least volatile stocks in a large cap index; the ones least likely to leave the index. Doing this would also lead to a systemic outperformance of the index. “It’s not really that you’re all that wonderful and smart, it’s just that you’re not subjecting yourself to the leakage associated with the index construction techniques of large gap index providers,” he said.
To illustrate this, he suggests that if 15 years ago an investor had used an investment bank to create a total rate of return swap contract that was $100 long the Russell 2000 index and $100 short the Russell 1000 index, they would now have one of the best ever long short equity track records.
This claim prompted Dr Ben McCaw, portfolio manager, capital markets research at MLC, to ask “if more investors had known about this 10 years ago, would the strategy still have worked?”
Weisman said that academic research on value, size and momentum is not fully put it into use by funds. He cited research of equity mutual fund managers which showed that only between 20 and 30 per cent applied these known factors in a statistically significant sense in their portfolios. The research found 80 per cent of equity mutual fund managers under-performed their benchmarks, while 65 per cent of managers that applied these risk factors outperformed their benchmarks. “There is a vast amount of research out there and the guys that use it have a huge statistical headwind,” he said.
Richard Lindsey, chief investment strategist, liquid alternatives, Janus Capital explained it another way. He said it was important to not see these risk premia as arbitrage opportunities, but rather as premia that paid off over the long run. He boiled it down to the fundamentals of not being paid for investing in assets, but being paid for the assumption of risk. From that basis, he said, these risks should be as diversified as possible.
Weisman offered an alternative perspective. “Where there are transaction costs, frictions, heterogeneous information, not everyone knows everything at precisely the same time,” he said. “In other words, in the real world the equilibrium in the marketplace is by definition inefficient – not efficient.”
McCaw questioned this. “If you ask a security analyst say ‘Why should I pay $100 for this company as opposed to $50?’. They’ll say ‘Well it can grow revenues. It’s got a certain profit margin. It’s got a dividend yield.’ In other words, there’s a fundamental basis for the price and you’re net long economic growth. But when targeting a particular factor, how do you know you’re actually getting a reasonable valuation and risk return trade off?”
Lindsey said the Janus Capital process did not forecast the returns associated with its 11 sets of risk premia. Rather it forecasts the variance co-variance of the premia to put together a portfolio. “We set the marginal contribution to risk, equal for each of those risk breakdowns. It’s based on our forecast of variance co-variance at the margin of what’s the right amount to put into each. And you target an overall portfolio of volatility.”
Oscar Fabien, chief investment officer of VicSuper, asked if Janus believed in mean reversion and if so how did they get to the mean?
Lindsey agreed the market environment would dictate how the risk premia would be obtained, citing how the equity momentum risk premium worked in this situation. “In the upmarket there’s basically a trend, so you tend to go further with the trend,” he said. “So you lever it up. In a down market, there tends to be mean reversion, so you want to buy losers and sell winners in that kind of market.”
What is the risk of investing in risk premia?
Kevin Wan Lum, portfolio manager equities and property at Ibbotson Associates, queried whether these strategies would provide protection in a 2008 GFC scenario. Weisman admitted that in a crisis there was a tendency for certain risk factors to become more highly correlated and that daily monitoring and scaling back would be needed.
McCaw questioned the use of leverage to back these positions. “If you’ve got a risk premium that you know accrues over a long period of time albeit with gaps that are hard to stomach, surely that creates a headwind to performance if you have to be short something in that particular market?” He expressed the same concern another way. “If I give you two options here, one is long equities in a very naive sense and two, long short equities, then I have to be buying some sort of skill from somebody to compensate me for the fact that the short part of my portfolio is expected to lose over time.”
Both Lindsey and Weisman repeated the assertion that by investing in proven risk premia they were not taking any big bets. And Weisman added how firms like Bridgewater were completely agnostic with respect to expected return and rather spend the bulk of their time thinking about volatility and dependence structure without attempting to forecast returns.
He added, he was very happy with the Janus Capital performance figures this year which put them in the top quartile, but he qualified this by stating: “Everyone else that’s in that top quartile has about 0.7 or higher correlation to the S&P, in ours it’s in the 0.2-0.3 range and at some point the world is going to separate into guys that are effectively writing equity beta and people who are doing things that are more diversified.”
Fabien questioned how much Janus were prepared to back these views. He pointed out that VicSuper’s alternatives program represented 6 per cent of assets could perform brilliantly one year, but it would not have a huge impact on overall returns.
Weisman said funds that had peer risk were often limited in how much they could back new ideas such as this. He pointed out that the largest sovereign wealth funds had been able to take bold steps with factor investing as they did not have to worry about peer risk. He cited the Norges fund – the biggest sovereign wealth fund in the world, as the best example of this, adding that the China Investment Corporation, Korea Investment Corporation were moving in this direction.
He also provided an example of how a fund might get round their sensitivity to peer risk. The Ohio Public Retirement System has taken a step by step approach to factor investing, starting out at a 2 per cent allocation in their portfolio, reviewing it after three years with their board and then six months ago moving up to a 10 per cent allocation.
Gerard Parlevliet, chief investment officer, Commonwealth Bank Super, saw the attraction of building portfolios that are truly diversified using liquid strategies where possible. Whilst many funds have looked to diversify through the use of unlisted real assets, this strategy has its limits, he said, as the lack of liquidity can result in sub optimal behaviour in times of crisis. The GFC showed us that the lack of liquidity can lead to situations where portfolios can become unbalanced as the more liquid assets, such as equities are the only assets that can be easily sold to provide liquidity. This can also have the flow on effect of exacerbating the falls in equities in such conditions.
Dr Ben McCaw, portfolio manager, capital markets research, MLC
Joshua Bloom, portfolio manager, Sunsuper
Gerard Parlevliet, chief investment officer, Commonwealth Bank Super
Kevin Wan Lum, portfolio manager equities and property, Ibbotson Associates
Oscar Fabien chief investment officer, VicSuper
Andrew Weisman, chief investment officer, liquid alternatives, Janus Capital
Richard Lindsey, chief investment strategist, liquid alternatives, Janus Capital
David Rowley, editor, Investment Magazine