After enduring two equity market crashes in the past decade, most investors are now approaching risk management with renewed emphasis. In this new reality, pension plan sponsors are forced to assess the evolution of risk management techniques and current approaches to managing pension funding volatility.
The bear market of 2000–2002 was a wake-up call to investors, who had enjoyed almost uninterrupted double-digit equity returns since the early ’80s. It also brought home the realization that the gospel of traditional mean-variance optimization—the framework for portfolio construction in the global investment industry for more than 40 years—was premised on the key assumption that asset class correlations remained relatively unchanged over time. Instead, investors found that equity market correlations effectively “went to one” in a crisis as global markets fell in tandem.
The mindset quickly set in that equities are risky and that this risk had to be reduced or hedged. This provided the perfect entry point for hedge funds and other alternative asset managers to approach the institutional marketplace en masse and begin targeting pension funds.
Interest in hedge funds
According to Hedge Fund Research Inc., from 2000 to 2008, hedge fund assets are estimated to have quadrupled from about $400 billion to close to $2 trillion. This newfound interest in hedge funds was not unwarranted, as these funds had successfully weathered previous equity bear markets and provided returns similar to equities. And there was once again evidence that hedge funds offered important downside protection when they significantly outperformed stocks from 2000 to 2002. Confidence was shaken, however, in 2008, when returns approached a disappointing -20%—despite the fact that this represented about half of the loss in stock markets. Investors became acutely aware that not all hedge funds truly hedged and that they held both residual equity exposure and non-traditional exposures to currency, interest rates, commodities and volatility.
As equity markets recovered from 2003 to 2007, memories of the previous crash faded and the pressing need to hedge was forgotten, evidenced by the ever-narrowing credit spreads on corporate bonds. But while equity markets staged an impressive comeback, the financial health of pension funds improved at a much slower pace.
According to Mercer estimates, from January 2003 to mid-2008 (when Canadian equities peaked), the solvency ratio of the typical Canadian pension plan fell by about 10%—despite the 100% trough-to-peak stock market recovery. The problem lay within the declining long Canada bond yields used to mark to market a Canadian pension plan’s liabilities on windup. The decline in interest rates that started in the early ’80s continued unabated throughout the ’90s and the new millennium, and this trend has steadily increased liabilities to levels that were difficult to foresee.
And this phenomenon was not limited to Canada: mark-to-market accounting using corporate bond yields began to have an impact on plan sponsors’ earnings globally, focusing attention on hedging interest rate risk in pension liabilities. The realization that liability risk was as important as asset risk led to a philosophy that investment portfolios should be liability driven. Liability driven investing (LDI) then became the buzzword of the years leading up to the Great Recession in 2008 and still lingers today.
Growth of LDI
LDI means different things to different investors. For many Canadian pension funds, the basic application was to restructure the existing (mid-term) universe bond allocation to be longer in duration, with an appropriate allocation, if required, to inflation-linked real return bonds. For those who took this step, liability risk remained largely unhedged, given the average asset allocation of 60% equities and 40% bonds. Interest rates, therefore, would continue to wreak havoc, though the pace had slowed somewhat. Other funds were interested in the LDI concept but were convinced that long-bond yields would inevitably increase. However, as of Sept. 6, 2011,10-year bond yields in Canada fell below 3% and to 1.9% in the U.S., closing in on all-time lows.
Several Canadian pension funds were early adopters of a more thorough and creative approach to LDI. With the objective of fully eliminating liability-related interest rate and inflation risks, these funds implemented an interest rate hedging bond overlay, in addition to restructuring physical bonds. Using this structure reduced equities while maintaining the total fund expected return, as the interest rate hedging overlay was expected to provide a positive return, albeit with some volatility. This philosophy represented a more rational balancing of equity and liability-related economic risks. In years when interest rates increased, the overlay return could be low or negative, but liabilities would decrease by roughly an equivalent amount.
Funds that implemented this approach still retained significant “unmatched” asset risk (around 45% to 50% of liabilities), though meaningfully less than the previous 60% to 65% exposure. The worst-case scenario was one in which equities fell drastically and financing costs on the overlay rose in a similar fashion. The impact of this scenario would be magnified, as the economic exposure of these funds was usually equivalent to 150% of liabilities (50% in each of stocks, bonds and the overlay). This did happen for a short time in 2008, as the three-month LIBOR (London Interbank Offered Rate) exceeded 5%, reflecting the severity of the global liquidity and banking crisis.
The 2008 credit crisis served as a stark reminder that risk mattered and that extreme market events were occurring much more frequently than expected. Liquidity temporarily disappeared for virtually all assets except government bonds, as the financial crisis deepened. Investment-grade credit spreads in Canada increased to more than 300 basis points, and the resulting flight to safety spared only government bonds. In the aftermath, investors concluded that other risks besides equity and interest rate risks were not only present but significant enough to impact results.
Tail risk events
Today, market cycles such as these have prompted a new orientation among pension plan sponsors toward the risk of catastrophic loss. There has been increased interest in hedging extreme tail-risk events, leading to the identification and evaluation of individual risks through a risk factor analysis/stress testing framework. Initial research suggests that there are three groups of tail-risk hedging strategies: option strategies; assets with low correlation to equities; and volatility and credit hedging strategies.
1) Option strategies: These strategies, by and large, use put options on equity indexes at various strike prices in a stand-alone or collar (sell a call and buy a put) set-up. One advantage to this strategy is that options are a perfect hedge with an identifiable payoff pattern. The key disadvantage is that options are expensive and represent a high fixed cost if used permanently. At-the-money options (in which the strike price is the same as that of the underlying security) can cost 3% or more per annum—roughly equivalent to the equity risk premium over government bonds—which implies that using them is the same as changing the asset allocation. An alternative is to use them tactically, which requires expertise in timing markets.
Despite these drawbacks, options can be used effectively. The low or no-cost collar structure can eliminate the impact of modest market declines of 10% or less.
2) Assets with low correlation to equities: The table below shows the recent behaviour of a variety of assets and strategies during the credit crisis. Certain ones stand out, including gold (20%), managed futures (17%) and global macro (5%) strategies. Managed futures are interesting because unlike options, which represent a potentially expensive one-way bet, they have shown strong up and down market performance as well as cumulative performance over 10- and 20-year periods. Managed futures funds are liquid, well diversified and able to invest in any asset class, subclass, sector or geography.
Managed futures are also able to vary their market exposure from net long to net short, and this ability has allowed them to track trending up and down markets remarkably well. An investor can diversify manager styles in the managed futures area across trend-following, fundamental and trading strategies. Managed futures have the additional advantage of being divergent price rather than convergent price strategies, as they will follow price trends that fundamental analysis may miss, given that markets may overshoot on both the upside and the downside. Fees in this hedge fund strategy are high and funds may impose a lock-up, but it is the one portfolio insurance approach that generates returns in both up and down markets (though they may have difficulty in volatile markets with no clear trend). However, fiduciaries that are not familiar with hedging may need time to reach the required comfort level with this strategy.
3) Volatility and credit hedging strategies: The VIX (CBOE) Volatility Index and bond credit spreads are not only a reading of current market prices for options and non-sovereign debt but also a current and forward-looking indicator of investors’ views on the riskiness of economic and market environments. As investor fear rises, the price of options is generally bid up—as is the implied volatility of the S&P 500, which the VIX measures. Similarly, investors require higher yields on non-government debt when the economic outlook is negative (ignoring, for the moment, the fact that many corporations currently appear to be much safer bets than several sovereigns—particularly in Europe).
These approaches can work effectively but require the use of derivatives and may be expensive. Also, as with options, constant use may entail periodic losses and ongoing cost. Finally, their complexity may make them suitable only for seasoned investors.
Understanding risk
For pension plan sponsors, risk factor analysis may be the best exercise in understanding the true nature of non-demographic funding risk. This analysis involves identifying key risks to which a fund may be exposed—including equity, inflation, credit, interest rate, currency and commodity risks—and evaluating the marginal loss from extreme moves in each area. This is a complex but extremely useful analysis best performed using stochastic asset-liability modelling. Information produced by such analysis provides excellent insight into both individual risks and the interaction of different factors, such as interest rates, inflation and asset and liability fluctuations.
When considering potential changes to investment policies, hedge funds—and particularly managed futures strategies—merit consideration. Artfully employed option strategies may be valuable in truncating potential losses, and perhaps gold will lose its stigma as an institutional asset class to avoid if governments globally further debase currency through unsustainable debt build-up. Bond overlay strategies can also be effective in hedging liability-related risks, but some of the benefits may be lost given the current level of long bond yields.
In today’s evolving risk management climate, the first step is to evaluate and deal with potential extreme market moves. Awareness of both the importance of risk management and the techniques used to understand and quantify risk have advanced considerably over the past 10 years. But the nature of risk continues to change, and investors will likely face significant challenges over the next decade, including sovereign debt, structurally high unemployment, climate change and environmental, social and governance expectations, as well as slowing developed market economic growth. As plan sponsors evaluate these new risks, it is critical to retain the lessons of the past while developing new solutions for the future. And as we do so, it will again feel as though we are experiencing risk for the first time.
Frank Belvedere is partner, investment consulting, with Mercer Canada. frank.belvedere@mercer.com
Get a PDF of this article.