Bailout

The myth that risk goes away over time

By Felix Salmon

Rodney Sullivan has a great column which is nominally about “risk parity” strategies, but which in fact applies much more broadly — to anybody, in fact, who buys in to the idea that if you invest in riskier assets, you’ll end up with higher returns.

Of course, high risk sometimes means that returns are much higher than expected. But it can also mean that returns are far lower, perhaps crippled beyond hope of redemption. Given the reality of fat tails, the likelihood of large negative events is also higher than normal. And poor results can persist for long periods. By now, the myth that risk ?goes away? over time should be well put away ? risk accumulates over time. In short, investors cannot expect to ?get? the expected return, but rather a draw from a very wide distribution. And in reaching for high returns by using leverage, investors dwell in the extremes of the return distribution.

The idea behind a “risk parity” strategy is simple. Let’s say you’re 100% invested in equities, because you’re happy with that level of risk. The problem is that you have no diversification: bonds can outperform equities for very long periods of time. So you rotate some of your holdings out of stocks and into bonds. But bonds don’t return as much as stocks, and that can hurt your total returns. So you leverage your bond-market investments, to bring them up to the same riskiness as the stocks.

Sullivan is not impressed. “Using leverage simply to increase the expected return is almost always a bad idea,” he writes, credibly enough. Which makes sense to me: the person lending you the money you’re borrowing expects to make a profit on the deal, which is only going to make it that much harder for you to make a profit as well.

What’s more, you could hardly pick a worse time to start levering up your bond investments than now:

That this leveraged-bond strategy is emerging after a long bond bull run and poor equity performance is curious. This likely hindsight-driven idea runs strikingly close to the frequent overweighting of equities during the late 1990s and the more recent affinity for uncorrelated assets during the early years of this century. Asset allocation decisions should never be made by simply extrapolating the future from the past.

This, of course, doesn’t just apply to bond-market rallies: it can be applied equally to the stocks-for-the-long-run crew. If you’ve got a coherent fundamental reason why stocks should outperform other asset classes, that’s great. But “they always did in the past” is much less convincing.

Full post as published by Macroprudential on July 09, 2010 (boomark / email).