Equity risk and bond risk are analogous to two people standing next to each other, one shouting (equities), and the other whispering (bonds). The whisperer is sadly never heard. We draw the analogy between the volume of their voices and their risk as defined by standard deviation. There is no balance in this picture and investing quickly becomes a bipolar experience as investors ride the often hysterical equity market's ups and downs (economic expansions and contractions). So what is the solution to this analogy? There are three: buy the whisperer a microphone, sedate the shouter, or find another person that can holler just as loud as the person shouting.
Buy the Whisperer a Microphone
Buying the whisperer a microphone to amplify their voice is exactly what leverage can do to a financial asset. Take a look at the risk return plots in Figure 1. If an investor had levered a bond portfolio four times over the last 18 years, they would have actually achieved the same risk profile as equities, with a far superior return. In addition, because bonds and equities have historically been uncorrelated, the overall portfolio risk doesn't increase much at all.
Figure 1: The Annualized Return and Risk of Equities and Bonds (Oct 2003 to Oct 2012)
This solution is really smart in principle, but we wouldn't recommend it in practice. Firstly, bear in mind the fact that bonds have been on a multi-decade bull run. Taking the historical return of any asset class that has done well and levering it provides a mouthwatering result. Don't, therefore, be too distracted by the positive outcome. The real problem with this approach is the mirror image of the return axes and it's called a levered loss.
If bonds and equities decide to correlate, an investor can lose a lot of money. In 2008 the S&P 500 was down -38% (using ETF proxy SPY) and the Barclays Aggregate Bond Index was up 5.9% (using ETF proxy AGG). If an investor had levered bonds four times, instead of losing -20% in a 60/40 portfolio, they would have lost closer to -10%. But what would have happened if bonds had lost -10% and the investor had levered them four times? The investor would have lost close to -40% of their capital. Could it happen? Probably not, but we don't like making bets with the potential for large left tails based on "probably not". This solution is too close to high risk, masquerading as low risk.
Sedate the Shouter
In this solution we directly address the shouter to curb their overexcited ways; it's the financial equivalent of giving them a Xanax. This is an approach we do advocate and it is exactly what lifting the constraint on selling short can do to an asset class. If you allow managers to both a) buy the securities they like, and b) sell short the securities they don't like, the resulting risk, as defined by standard deviation, is radically reduced. Being long and short at the same time has a dampening effect. Generally, when the market goes up, the shorts lose money and the longs make money. When the markets go down, the reverse is true. Using this approach it is not difficult to reduce the volatility and drawdowns by approximately one third. In fact, at the extreme end of the spectrum, in an equity strategy referred to as "market neutral", it is possible to morph the risk/return characteristics of equities to look like bonds.
In the traditional paradigm, equity exposure is slowly eradicated because the need for capital preservation becomes paramount. That's unfortunate because whilst performance hasn't been great over the last ten years, generally the equity risk premium has shown itself to be a powerful return generator over time. Long/short equity strategies allow the investor to have their cake and eat it, because they provide capital growth without compromising capital preservation. Put simply, an investor can keep a larger allocation of equities for longer by using equity long short.
Find Another Shouter
The third solution is to find another asset class whose risk is as loud as equities, thereby challenging its dominance. This is another approach we advocate and it's what we strive to do through our asset allocation. The good news is there are a number of investments in the world that will do this. For example, commodities are also very volatile but are not highly correlated to equities over long periods of time. As a result, commodity risk will push back equity risk. This is best illustrated through the following simple risk budgeting example:
(click to enlarge)
Figure 2: Dollar vs. Risk Budget, Equities and Commodities (June 2006 to Oct 2012)
We create a simple portfolio with a 60% dollar allocation to the S&P 500 but switch out bonds for a 40% allocation to DJ UBS Commodities Index (using ETF proxy DJP). The historical standard deviation of the portfolio over the last six years was 15%, of which commodities contributed 45%. Without making any comment about the returns (which have been similar over that period), we can say that these two asset classes are evenly matched from a risk perspective. We have indeed found another shouter, and one with some interesting and differentiated return characteristics, such as the potential to perform in an inflationary environment.
The only fly in this ointment is similar to the problem with applying financial leverage. Correlations are unstable and can change significantly during periods of stress. Our newly found loud friends may all start shouting for mercy at the same time, and that won't really help us.
To conclude, this analogy represents the conundrum most investors face today. Placing standard deviation central to our definition of risk, the traditional paradigm of stock and bond investing appears deeply flawed. It hides a large bet on equity risk which is in turn a bet on growth.
We would never make the statement that "leverage equals risk". This is simply not the case and when used prudently leverage can actually meaningfully reduce risk. However, we are not keen on giving any of our quietly spoken asset class friends a microphone. They may turn out to have the voice of an angel (as illustrated in Figure 1). However, it carries the left tail risk of them using the microphone to shout obscenities, resulting in the levered loss scenario.
Our favored direct solution is sedation; release the long only constraint and allow managers to short stocks too. Not only will this reduce equities' contribution to risk, it also reduces overall risk and allows investors many more years to harvest the important equity risk premium.
Our favored indirect solution is to find other investments with equally strong personalities. We used the example of commodities, but managed futures and publicly traded real estate are two others that jump to mind. This solution also has one very attractive characteristic that our analogy does not take into account. In finance, two hysterical asset classes can actually sum to a more rational whole. If the asset classes are uncorrelated, they actually sedate each other! Don't try this at home though, it has been proven to be a highly unsuccessful strategy in marriage (much better to be the bond to your partner's equity).
 Using HFRI Equity Hedge Index as a benchmark, it was down -26.6% in 2008 (source HFR hedgefundresearch.com), when global equity markets were down in the range of -40% to -50% depending on the region.