Many investment fund managers face a dilemma: they need to achieve true diversification with their allocations and non-correlated sources of return, but it’s an elusive goal if you don’t correctly assemble the alternative investment components of your portfolio.
Getting the mix right is crucial for funds as future liabilities will continue to grow and the outlook for the assets that cover them is dim when seen through the lens of a portfolio consisting only of equites and fixed income. To cover funding gaps, many funds have been looking to alternatives, often taking on more risk.
“It’s very difficult to get genuine diversification when only equities and bonds are on the menu,” says Philip Seager, head of alternative beta strategies at CFM, an asset management company whose investment strategies are based on a quantitative and scientific approach to financial markets. As an illustration, many investors – especially pension funds – have significant exposure to equities, and often credit, but credit has a high correlation to equities.
Getting the mix correct
Alternatives come in many forms and combining them correctly can produce robust and stable outcomes. The key, says Seager, is to identify various risk premia and non-risk premia, and combine the strategies that use them in a non-correlated way to diversify sources of returns and to dampen risk. “Outside of bonds and equities, many risk premia exists, and as many as possible should be implemented in institutional portfolios to improve the profile of future returns,” he says.
Risk premia describes strategies that reward negative-skew risk – large, infrequent downside returns but overall long-term gains. Many well-known strategies are negatively skewed – traditional investments in bonds and equities, for example, along with short delta-hedged option insurance strategies, foreign-exchange carry, corporate credit, relative value bond funds, convertible arbitrage funds, and others.
Investors tend to assume that non-traditional sources of risk premia may not help to drive returns and could increase losses in the next crisis – but it’s simply not true that negatively skewed strategies deliver performance only until the next crisis. Rather, they deliver in proportion to the amount of negative skew in the returns, losing occasionally but gaining commensurately. The losses are not sufficient to offset the long-term gains. “To maximize the probability of achieving positive excess returns, investment in a well-diversified and risk-controlled mix of risk premium strategies is, in fact, essential to building a portfolio that remains robust through good times and bad,” says Seager.
Being rewarded for risk
The insurance industry provides an apt analogy for understanding negative-skew risk. Infrequent but potentially highly disruptive risks to a business require higher premiums. Frequent but far less impactful risks have smaller premiums. It’s assumed that after premiums are paid to the insurer the average expected gain for the insured must be negative and that for the insurer, therefore, correspondingly positive.
Negative skewness is also known more colloquially as fat left tailed returns – a slight-to-likely probability of a moderate or extreme outcome of more than three standard deviations from the mean, and in conjunction with a positive Sharpe ratio, the measure of return adjusted for risk, one obtains the ingredients to what CFM calls true risk premia. “You can think of the premium in the insurance analogy as being proportional to the amount of weight there is in the fat left tail. The fatter the left tail of one’s returns get, the riskier and more uncomfortable the strategy becomes, requiring a larger premium to be paid, which translates to better excess returns and a higher Sharpe ratio,” Seager says. The higher the skewness, the more asymmetric those returns are, which tends to result in a higher Sharpe ratio. As the probability of large downside risk increases, investors are rewarded for assuming the risk in a fat left tail. “The return improves to entice investors in such a way that returns are large enough to compensate for the risk assumed,” says Seager.
The correlation of payouts is another element. An insurer would rather cover a number of detached houses, for example, rather than a whole apartment block, because when the apartment block burns down, everyone makes a claim at the same time. “This is all relevant to risk premia investing because we look for those strategies that do not correlate in the tails. When they draw down, they don’t all draw down together,” Seager says.
Constructing diversified sources of returns
Correlation and the lack of it are underlying principles in the insurance industry, too, but an insurer will still enter a contract when it sees correlated risks, because most of the time there will not be a required payout. With negative skews “you start by looking for uncorrelated payouts – uncorrelated left tails – and even if there is correlation in the tails, a payout usually is not triggered under normal circumstances,” Seager explains. The tails could be correlated by return, but in a normal market, when there are no payouts, there is diversification in the heart of the return distribution, and thus higher risk adjusted returns are available, as long as one carefully manages exposure to tail events.
A positive-skewed, positive Sharpe ratio strategy is not a risk premium, but when bundled with multiple negative-skewed strategies, they provide effective diversification, driving risk-adjusted returns. Trend following is an example of such a positive skew. In the insurance analogy, the premiums collected by the insurer provide positive performance – but it periodically pays out, i.e. a negative skew. Conversely, the market is willing to pay a premium for downside protection, which implies a negative performance for the insurance buyer but a positive skew in the event of a payout. Positive-skew, positive Sharpe ratio strategies should not exist in this risk premia framework and their positive performance is more likely due to a behavioral anomaly.
The importance of diversification is paramount when employing these types of strategies. Adding many risk premia together builds up a higher Sharpe ratio and decreases the fatness of the left tail of returns. “When constructing the overall strategy, we start by combining as many fat left tails together as possible, which mechanically reduces the negative skew and improves the Sharpe ratio,” says Seager. This includes strategies like carry trades, shorting delta hedged options. “Then we think about combining it with a fat right tail to improve the skewness profile further, all the while improving risk adjusted returns,” he says. This includes trend following and certain equity market neutral strategies such as quality and value, which do not have a fat left tail.
At CFM, the strategies are collectively run at predetermined risk and ongoing rebalancing. The portfolio has passive exposure to many risk premia, but maintaining consistent levels of risk requires a great deal of oversight and active trading. “Something like a carry trade or systematic equity short strategy is something that is quite complicated to implement with a constant level of volatility, because you constantly have to readjust your position,” says Seager. In assembling risk premia portfolios, thinking about fat left tails is a different starting point. “We allocate equally to sources of skew in the way allocations are made to risk in a risk-parity approach,” he says. “You can think of it as ‘skew-parity.’”
Any description or information involving investment process or allocations is provided for illustrations purposes only.
Any statements regarding correlations or modes or other similar statements constitute only subjective views, are based upon expectations or beliefs, should not be relied on, are subject to change due to a variety of factors, including fluctuating market conditions, and involve inherent risks and uncertainties, both general and specific, many of which cannot be predicted or quantified and are beyond Capital Fund Management's control. Future evidence and actual results could differ materially from those set forth, contemplated by or underlying these statements. There can be no assurance that these statements are or will prove to be accurate or complete in any way. All figures are unaudited.