Risk and Portfolio Construction: Rates of change
To a large extent today the question of what to do in portfolio construction is really a question of what to do about interest rates.
This is partly because one crucial market – bonds – looks like a one-way bet. I’ve often heard the joke that Japanese equities were a “fantastic diversifier” through the 1990s and 2000s: it was a joke, of course, because no matter how diversifying a market is, no-one wants to put money in something that only ever goes down. Moreover, losses may not even be compensated by diversification: after such a profound change, from a 30-year bond bull market into a new bear market, who is to say whether past correlation characteristics are any guide to the future?
As many investors have painfully learned, bond yields can go lower than you think – and stay there for longer than you think. That is why we begin by questioning bearish assumptions about bond markets. Sentiment turned decisively during 2013 as yields continued to rebound from the lows hit in July 2012, particularly at the first hints that the US Federal Reserve was getting ready to slow down its bond purchases – but we try to assign probabilities to the alternative scenarios of a re-test of those lows or a prolonged period of rates being stuck at their new levels.
Then we turn to the question of how the new environment might influence correlations between equity and bond markets. It turns out that negative correlation has been the exception rather than the rule, and that the pattern may have less to do with the direction of rates than with the volatility of inflation expectations.
This debate cuts to the heart of risk parity strategies, which are focused on collecting the premium for being diversified. Because they often try to do this by applying leverage to their bond holdings, many have questioned their appropriateness for a bear market in rates and can now point to some bruising return figures for 2013 to make their case.
Our survey of the risk parity strategy reveals that the old ‘leveraged duration’ label probably mischaracterises modern iterations of the strategy. However, it is also worth noting that our risk-parity coverage also sees the assumption that rising rates are necessarily bad for bonds seriously questioned. As some practitioners observe – and as we discuss in our article on the potential returns from carry in a rising rate environment – bond market returns are as much about expectations of how yields will move as they are about the simple direction of travel.
If carry can be put forward as one strategy for bond portfolios, a more aggressive one might be multi-asset fixed income and other absolute-return approaches. Things start to become complex at this point, as the range of instruments and tactics – and consequently of risk exposures - opens up for investors. Should investors allow negative duration, through shorting and floating rates? Leverage? How far along the credit spectrum should they go?
This discussion moves us naturally into our closing subject matter: once it is decided that something has to change to reflect a bearish view on duration, how far do we move from the traditional ‘barbell’ portfolio, split between risk-seeking and risk-free assets, towards portfolios filled with assets that mix duration with economic-risk exposures?
Our investigation suggests a surprising reluctance for institutional investors to take that road. This may reflect the strategic importance of risk-free duration assets for liability-matching, or simple governance constraints; but perhaps it also indicates that a lot more thinking needs to be done about the forces described in this supplement.