Wednesday, July 30th, 2014

Everything’s Expensive

Scarcely a day goes by without a newspaper article arguing that one asset or another is trading at its most expensive valuation in the last ten or twenty years. A recent favourite is the statement (which we have no way of verifying) that Spanish bond yields are at their lowest since the French revolution. Even Janet Yellen has felt obliged to comment on US biotech stocks and non-investment grade corporate bonds. However investors can be forgiven for not reacting, because when everything is expensive in absolute terms, nothing is expensive in relative terms.

We take a different view. We believe that most assets are actually cheap when compared with the metric which really matters, by which we mean volatility. US equities currently have an annualised volatility of 8.6% (trailing 26-week total returns) which puts them in the bottom 1% of all observations over the last 20 years. When consultants measure the effectiveness of fund managers they measure return per unit of risk, not price per unit of earnings, or spread per unit of capital. The cost in this equation is volatility. When this is very low everyone’s performance look excellent. Thus the Sharpe ratio of the S&P 500 over the last 52 weeks is 2.13.

This is not just a debating point, because it is changes in risk conditions which drive changes in price, not the other way around. High PE ratios and low spreads are a function of low volatility. This is often accompanied by low interest rates, but it’s the volatility which counts, not the cost of money. It is possible to have low valuations and low volatility at the same time (EM equities in late 2006 would be a good example), but high valuations and high volatility hardly ever happen. This applies to all asset classes, not just US equities.

Low volatility clearly brings with it the risk that it will rise, but this is a latent, not a current risk, which can be measured, by ranking each asset in terms of its historic volatility. On this basis US equities have the highest latent risk, with 99% of all prior observations of trailing volatility higher than the present. Second is non-investment grade debt in the US with 97% of all prior observations, third are Emerging Market equities with 93% and fourth is medium-dated US Treasuries, on 91%. The least unattractive (though still vulnerable) are Japanese equities with 76% and Emerging Market bonds with 82%.

To complete the picture, we also need to assess the likelihood of a change in the risk regime. Our view is that the primary purpose of QE has been to boost the risk-bearing capacity of intermediaries and investors in US capital markets. We think this objective will remain in place long after the taper comes to an end. The fact that volatility is very low does not mean that it has to rise immediately. In April 2004, the volatility of US equities hit 9.3%, which was a new multiple-year low at the time. It did not rise above 12% until July 2007. Once again, we live in an era of managed volatility. Of course it will end, but the key is to work out the sequencing between different asset classes. Unless the Fed makes a mistake, we doubt that US equities will be the first to crack even though they have the highest latent risk.

Leave a Reply

Full Blog Archive:


Other Recent Posts