Fundamentals key to handling tail risk

Updated with correction

Tail risk — the risk of extraordinarily large losses — has become a major concern for institutional investors.

Many seek to insure against tail risk directly by steadily purchasing put options or similar derivative strategies.

Unfortunately, experience and financial theory suggest that the long-term cost of such insurance strategies will be larger than the payouts. No surprise, since the expected return for perpetual buyers of insurance is negative and conversely positive for insurance sellers (see the entire insurance industry). It is reasonable to assume that there is a premium from bearing tail risk, so attempting to benefit from traditional assets while passively removing the tail is an expensive proposition. Worse, the long-term cost of insurance means that investors might abandon the strategy after a long period of calm markets — potentially the worst time to bail out.

Equities are the greatest source of tail risk in most institutional portfolios, and to address this risk, investors would be better served by a fundamental approach, one that changes the portfolio structure itself. Consider the following approaches, which we think are most effective in combination.

One of investors' main takeaways from the financial crisis was that “diversification failed.” We think that the implementation of diversification failed, not diversification itself. Most institutional portfolios are not truly diversified. Although they might hold different asset classes, these portfolios are overwhelmingly concentrated in equity risk simply because the risk of equities is higher than for most other asset classes. (For example, a 60/40 stock/bond portfolio has almost 90% of its risk in stocks.) This concentration means that a bad year for equities will be a bad year for the entire portfolio. Other asset classes — bonds, credit, commodities and alternatives — won't really matter much, even if they have extraordinary results.

Our recommendation is to create a portfolio that is better risk-balanced by reducing equity in favor of other return sources, including nominal interest rate exposure; exposure to inflation-sensitive assets such as commodities and inflation-linked bonds; and a range of credit exposures including corporate and emerging markets. Diversification can also come from adding truly uncorrelated alternative strategies. Importantly, adding new risk sources does not mean that total portfolio risk increases; rather the portfolio's concentration risk is reduced. In other words, the portfolio is better diversified.

Most portfolios are rebalanced to a target asset allocation. But this approach fails to reflect the changing riskiness of assets. When the volatility of a given asset spikes, investors who rebalance to their previous capital allocation (or who don't rebalance at all) are essentially doubling down. Dollar exposure may stay about the same, but risk exposure — e.g., the amount of money likely to be made or lost from that asset on a given day — has increased.

Our research suggests portfolios that maintain steady risk (or even reduce risk) when forecast volatility is high might earn greater risk-adjusted returns. This is in part because the risk-adjusted return of equities (and other assets) often falls when volatility increases, leaving investors with larger exposure to an asset with a smaller risk-adjusted return. But the greater benefit of volatility targeting is risk management. The round-trip gain or cost of the approach will vary from occasion to occasion, but it's a near certainty that a volatility-targeted portfolio is easier to stick with in tough times. We believe that a good portfolio you can stick with through a market cycle is better than a perfect portfolio that you're forced to abandon during a tail event.

Because equities are the dominant risk in most institutional portfolios, another approach to reduce tail risk is to lower the intrinsic risk of the equity investments themselves. “Beta” describes how a security's price varies with the market. A stock with a beta of 1.5 has historically been expected to gain or lose 1.5 times what the market gains or loses. Investors seeking ambitious returns tend to favor high-beta stocks, those with greater risk and greater expected reward.

Because high-beta equities are the preferred way to get higher expected returns, their prices tend to be bid higher than they otherwise should be. Conversely, low-beta equities are relatively unloved, and with less demand, trade at lower prices relative to their fundamentals. This market distortion creates opportunity: a portfolio of low-beta stocks that has roughly the same return as the overall market, with less risk. Low-beta stock selection underlies many “defensive equity” strategies, which seek to reduce tail risk while preserving much of equities' upside.

Ultimately, these three approaches to tail-risk management lead to better-constructed portfolios. By contrast, a myopic focus on tail risk often leaves investors worse off. Buying insurance in perpetuity is rarely the right long-term investment policy for several reasons: the amount to insure against is not a static figure; the cost of insurance is a drag to long-term performance; and costs tend to increase when insurance is most needed and most in demand.

Market returns are not normally distributed and tail risks do exist. A fundamental approach to addressing these risks — ideally one that includes all three elements described above — may leave investors better positioned for the long term.

Adam Berger is head of the portfolio solutions group, Lars Nielsen is a principal and Daniel Villalon is an associate of AQR Capital Management LLC, Greenwich, Conn. Their commentary is adapted from their paper “Chasing Your Own Tail (Risk).”