This is most true for government bonds. However, corporate bond spreads are also close to all-time lows, so they are vulnerable to significant spread widening in a risk-off move even though they are now effectively underwritten by central banks.
A major new risk-off move is now looking as likely, if not more likely, to come from worries of overheating, a marked rise in inflation and Fed tightening rather than too little growth or deflationary fears as has been the norm over the last decade.
If so, far from government bond yields falling in a risk-off move, government bond yields would be rising. In essence, government bonds would no longer be negatively correlated with equities but positively correlated, thus removing most of the protection benefits of holding government bonds.
Yields on government bonds are zero or negative in real terms almost everywhere. So, they no longer provide the protection against inflation they did before. And if inflation does pick up significantly, real yields could turn more negative with governments/central banks both keen to limit any rise in yields.
This is more of a problem for fixed income heavy lower risk portfolios as the larger equity weightings in higher risk portfolios should provide protection against inflation at least while it remains below 3% or so.
With yields so low and the duration of bonds correspondingly longer as a result, bond returns are considerably more volatile with swings in their prices swamping the underlying yield. As someone wittily put it “bonds now offer return-free risk rather than risk-free return”.
Rising inflation expectations
Rising inflation expectations are one of the key factors affecting the traditional 60 to 40 asset allocation. Within the bond allocation, they warrant a larger proportion in inflation-linked at the expense of conventional bonds.
However, particularly in the UK, for structural reasons relating to pension fund demand, index-linked bonds are very expensive with a significant negative real yield of -2.5%. Inflation-linked bonds are also long-duration and exposed to any rise in real yields. Therefore, increased inflation risk should only partially be accommodated by a higher allocation to inflation-linked bonds with additional inflation protection being also required from other sources.
The ideal investment allocation to bonds will continue to depend very much on what the risk profile/time horizon of the portfolio is. The outlook for inflation is also crucial in determining the optimal allocation going forward. This should be rather clearer in a few years’ time when it should be much more apparent whether we have moved into a new inflationary era or are back to the disinflation of the last decade.
What is clear, however, is that the ideal bond allocation for any portfolio is significantly lower than in the past. That said, while bonds generally are looking expensive and unattractive for all the reasons above, there are still pockets of relative value in areas such as emerging market bonds. These have a role to play due to their different risk/return characteristics to developed market bonds and emerging market equities.
Replacements for bonds will continue to be centred on areas such as alternatives particularly those which purport to have low correlation to equities and bonds – dynamic macro-driven alternatives funds being a prime example. ‘Real’ assets such as gold, infrastructure and property provide some protection against inflation and, in the case of gold, also some protection against big risk-off moves.
Cryptocurrencies fall into the ‘real’ asset camp and in the future a small allocation might be merited as a bond replacement. However, the violent swings currently being seen in cryptocurrencies, which are still in their infancy, mean for the moment they clearly do not fulfil the key role of bonds which is to reduce portfolio risk.
Instead, an allocation could only increase the risk and consequently, we do not believe they currently have any role in mainstream portfolios.
*Rupert Thompson is chief investment officer at Kingswood