Investors tend to see the sovereign bonds of developed countries as a kind of asset that offers modest but reasonably higher yields than cash at a low risk. In fact, unlike shares, sovereign bonds are usually labelled as «safe assets» that provide a «risk-free return». However, today’s particular situation invites us to review this trait.
On the one hand, the expected return on public bonds (measured via their internal rate of return or IRR) is at an all-time low. A basket of 1 to 10-year bonds from the G7 countries yields an IRR of barely 0.3%, almost the level of the official interest rates of the central banks. On the other hand there are serious reasons to doubt that the market is currently an efficient judge of economic prospects and risks: distortions resulting from the mass purchases made by central banks (quantitative easing or QE), the presence, in its wake, of investors with momentum strategies looking for short-term capital gains (hedge funds), the existence of forced buyers for regulatory or statutory reasons (pension funds), as well as the herd behaviour adopted by many traditional investors when a certain asset is in vogue. Consequently bonds can entail a higher risk than at first glance. To analyse this area, it is useful to break the IRR for nominal bonds into four elements that can be affected by corresponding risks.
First, «inflation compensation» to preserve purchasing power. This is made up of two parts: the average «expected inflation» rate during the lifetime of the bond and an «inflation risk premium» to offset the possibility of the effective rate being higher from the expected rate. At present investors expect inflation to remain between 1% and 2% for years, clearly below central bank targets. This expectation, however, ignores the slow but sustained rise recorded in US core inflation over several months now, already standing at 2.3% and inflation expectations could go up dramatically in the future. Moreover, if this occurred without a parallel increase in expectations of economic growth, the stock markets are very likely to suffer losses. In this respect we should note that, over the last few years, the inflation risk premium has been very low and even negative as investors have seen bonds as a hedge («insurance») against a scenario of economic crisis and deflation. This hinges on the expectation that bonds have a negative correlation to shares, which in turn respond to expectations of the pro-cyclical behaviour of inflation. If this comes about, there is likely to be a notable rise in the inflation risk premium.
Second, the «real return», which is also made up of two parts: the average «expected real short-term interest rate» during the lifetime of the bond and a «real term premium» to offset the risk of effective interest rates being higher than expected. Interest rate risk is also onsiderable at present. Investors seem convinced that real short-term interest rates will remain ultra-low for several years within a scenario of secular stagnation. Any small positive surprises in global economic growth could therefore be enough to trigger an upward shift in this expectation, leading to bond losses. Moreover the powerful downward effect of QE on the «real term premium» (pushing it into negative terrain) would disappear if this economic recovery convinced central banks to reverse their policy.
Third, a «default risk premium» to offset the risk of the issuer not paying back coupons and capital in full. All sovereign bonds entail credit risk although this may be very low. The ECB’s QE has undoubtedly contributed to pushing down Europe’s public debt risk premia, mainly that of the peripheral countries. But once QE comes to an end, the assessment of the sustainability of public accounts will become more important in establishing the risk premium.
And four, a «liquidity premium» against the risk of not being able to sell bonds before maturity, on the desired date and at the desired volume, due to no counterparty being willing to buy them. Liquidity risk has increased in the last few years due to the regulatory restrictions imposed on banks when operating as market creators, and due to QE itself, although the perception of this risk by investors is intermittent. We have witnessed a couple of storms (in 2013 due to the Fed’s tapering and in 2015 after the ECB’s measures) when the initial rise in the IRR triggered a sell-off due to a combination of poor liquidity conditions and the large number of bonds in the portfolios of many investors. Both episodes were short-lived but might well happen again and push up the «liquidity premium».
Given this horizon of risks a 0.3% IRR does not seem very attractive. Perhaps we should start paying more attention to those who have been warning for some time that sovereign bonds now offer «risk without return».