now that many asset classes have become expensive
Diversification - defined as a risk management technique that mixes a wide variety of investments within a portfolio - is the alpha and omega of portfolio construction. The rationale behind this technique contends that a portfolio of different kinds of investments will, on average, yield higher returns and pose a lower risk than any individual investment found within the portfolio. i
Diversification in itself relies on the return correlation between different asset classes. The correlation varies between -1 and +1 and measures the co-movement between different investments. Funds that tend to rise and fall in tandem will have a positive correlation; if one asset falls in price while the other rises, the correlation will tend to be negative.
Since most asset classes are not perfectly correlated with each other, adding more asset classes to a portfolio increases the risk-adjusted returns. If, for example, a given bond portfolio is expanded to accommodate an additional 15% in stocks, the Sharpe Ratio will increase to a level that is higher than each of the Sharpe Ratios of the two asset classes. This is remarkable, since, at first sight, the Sharpe Ratio could (naively) be expected to be the weighted average of the stocks and the bonds. Thus, as illustrated, the magic of correlation can push the Sharpe Ratioii upwards to an astonishingly high level.

It is obvious that diversification has played a key role in the past. The negative correlation between stocks and bonds has been one of the fundamentals of portfolio construction. In times of risk aversion, bonds increased in value since investors switched out of stocks. When investors' risk appetite improved, the inverse relation prevailed.
But will this negative equity-bond correlation continue to behave as before?
The quantitative easing of many of the world's central banks has increased the valuation of many asset classes around the world. Yields on government and corporate bonds have reached all-time lows, while, at the same time, equity multiples have impressively increased in recent quarters. At the same time, many asset classes have simultaneously increased, raising the probability that they will fall all together in the future. If this happens, the equity-bond correlation could turn positive and hence the diversification benefits will disappear!
Intuitively, it is easy to understand why bonds might no longer dampen the volatility of a multi-asset portfolio. At yield levels close to, or even below, 0%, bonds cannot be expected to increase much in value. Hence future bond returns can be seen as asymmetric: limited upside potential but having large downside risks.
Some historical episodes might illustrate the effect of a sudden change in correlation. The diagram below shows some bond market corrections during which the equity-bond correlation turned all of a sudden positive. The first half of 1994 was marked by a bond crash and declining stocks. The turmoil in June 2013 was caused by the so-called 'Fed Taper Tantrum'. Since mid-April 2015, bond yields have sharply increased while equity markets have dropped. In all these episodes, the classic diversification assumptions, based on negative equity-bond correlations, broke down. Naive investors were surprised by the losses on the multi-asset portfolios due to the unexpected positive equity-bond correlation.

It is indeed worrisome that the equity-bond correlation has been creeping upwards since the actions of the European Central Bank. The next graph illustrates the correlation over three years of some bond indices with the MSCI EMU equity index. It shows that the correlation between Eurozone government bonds and Eurozone equities has turned positive. This has not happened since before the new millennium.

Additionally, a negative equity-bond correlation should not be taken as a given. In reality, it is a relatively new phenomenon. Looking at historical data in the USA over the last 60 years, negative equity-bond correlations have been observed only in the last 15 years. Before that, positive equity-bond correlations were the norm.

The crucial question now is what preparations can be made to handle instability in the equity-bond correlation.
One of the most obvious answers would be to forecast the future equity-bond correlation. While perfectly rational, this writer is rather sceptical regarding the predictability of correlations. Predicting returns is already a challenge; forecasting co-movements between asset classes undoubtedly isn't much easier …
In order not to be surprised when correlations prove disappointing, the investment processes of institutional portfolios should be adjusted. The seven summarized recommendations below should prove helpful in preparing for the future:
- Consider cash as an asset class While cash is not a natural component of a long-term investment portfolio of, for example, pension funds or life insurance companies, it is the only asset class with almost zero correlation with other asset classes. Thus, when all asset classes fall in value, cash might be the only safe harbour where investors will not lose money. To this end, the investor locks in part of the unhoped-for past returns on bonds and sets ammunition aside to benefit from future potential higher yields.
- Improve your risk calibration By now, it is widely known that return distributions are not Gaussian. They tend, however, to be asymmetric and also have fat tails. Hence we strongly advise replacing volatility by Conditional Value-at-Risk (CVaR), replacing simple correlations by Copula functions and maximizing expected portfolio returns for a given portfolio CVaR.
- Integrate stress tests & scenario analysis into your investment decision-making process Stress test & scenario analysis illustrates the impact on the portfolio in specific scenarios that have not yet materialised. However, since the past is not always a reliable indicator of the future, any stress test will complement the information on which well-informed investment decisions should be taken.
- Beware of simple risk-parity multi-asset portfolios Simple risk-parity multi-asset approaches are based on historical risks and correlations. If correlations formerly negative turn positive, these portfolios will accumulate large losses. Additionally, simple risk-parity portfolios tend to be excessively exposed to bond markets based on the high Sharpe Ratios of bonds over the past 15 years. However, given today's yield levels, this approach might be an accident waiting to happen once yields start to rise …
- Do not blindly follow bond benchmarks As bond benchmarks simply include the outstanding amounts of all existing bonds, they are, in essence, driven by lenders who issue longer-term debt when yields are low, and inversely. This translates into a higher duration when yields are low. This is exactly what a rational bond investor should avoid since he is getting less return for an increased level of risk. We advise searching for flexible bond funds that do not merely stick to the bond benchmark.
- Analyse the possibility of having negative durations When bond yields increase, the only way to make money is by having a negative duration. Although many institutional investors are not familiar with negative durations, investors are advised to look into Total Return Bonds or hedge funds, which are generally flexible enough to hold negative durations.
- For those investors that do not trust future correlation and diversification effects, a Downside Risk Control (DRC) mechanism can be implemented This tool will aim to protect the value of the portfolio by integrating the recent portfolio performance into the equation. Based on the estimated portfolio CVaR, the probability of breaching a predefined maximum portfolio loss will determine an eventual deleveraging of the portfolio. When this occurs, the maximum loss on a portfolio is controlled, even in times when correlations turn wild …
iSource: Investopedia.com iiThe Sharpe (named after William F. Sharpe) is a way of examining the performance of an investment by adjusting for its risk. The ratio measures the "excess return per unit of deviation in an investment asset or a trading strategy", typically referred to as "risk".
Fixed
Income
News