In October 2013, we provided an analysis of “risk parity” strategies. The specifics of this particular investment fad are less interesting, but the core issue of bond-equity correlation is central to asset allocation. We concluded at that time, that the rise in the term premium implied that long-dated bonds offered attractive diversifying properties:
“as many are now rethinking the merits of “risk-parity”, the central correlation that created its track record, is now attractively priced. The long end of the yield curve is now paying elevated levels of term premium – which is precisely when it is cheap as a diversifier against cyclical growth risks.” (October 2013)
This year 30-year treasuries have provided considerable diversification at weekly and monthly frequencies, while delivering strong cumulative positive returns, as have US equities. The chart below shows the weekly correlation between the S&P500 and the 30-year bond future. In all periods bar one week in September and one in August, 30-year treasuries rallied whenever the S&P500 declined significantly.
The term premium is compressed
There are many ways to estimate the “term premium” (or how much premium over expected interest rates a bond investor is being paid). In our previous blog we simply showed the spread between US 2-year treasuries and US 30s, which in October last year was around 340bps. This spread is now 215bps. The vast amount of commentary around bond markets this year usually fails to mention that US 2-year yields have in fact risen in the past 12 months (and 5-year yields are barely changed). It follows that the decline in 10- and 30-year yields is largely about the term premium or medium-term interest rate beliefs, not about US growth or short-run inflation trends.
A number of factors drive the term premium in the treasury bond market. The most plausible candidates are declining global interest rates, a change in “equilibrium” rate expectations, and a reassessment of the risk properties of bonds. The latter may be the most important, and the most fickle.
Most important, though, is that the odds on backing bonds as a diversifier have deteriorated very significantly. The 2s-30s spread is 130bps flatter. The 5y5y implied by the US yield curve, and a reasonable proxy for “equilibrium” interest rates, is now 2.8% – twelve months ago it was 4.5% (see figure 2).
The impact of global rate expectations is now likely to be muted: not only have Euro rate expectations declined substantially, but the zero bound is becoming relevant to most of the major developed government bond markets outside the US. Cyclical facts may also provide a challenge to prevailing beliefs: the US economy appears to be strengthening, and has just been given a shot in the arm from a collapse in the oil price.
Now is the time to view duration as an addition to risk in multi-asset portfolios, not a diversifier. Indeed, the probability that bonds and equities could simultaneously lose money is now elevated – particularly those equities which have benefitted from the sustained decline in yields.
Changing risk properties
Endogenous risk is an important concept, often neglected or explained confusingly by “positioning”. It refers to the fact that asset prices are not always determined by a straightforward relationship with fundamental newsflow. Investor perceptions of risk properties, correlation and volatility, can themselves drive asset prices, and often amplify moves.
Perceived risk properties are important determinants of demand for assets. We have just been through a phase of significant equity market volatility where bonds have shown their value as diversifiers. This itself has caused the term premium to decline. If the bond equity correlation changes, beliefs could shift violently in the opposite direction: bonds will become risky again.