Sovereign Bonds May Not Save Your Portfolio

My working life has included pretty much the highest and lowest government bond yields in history. In the early 1980s U.S. 10-year yields touched 16%. Earlier this year they fell to 50 basis points. That’s generous compared with yields in Europe, mind, where any borrower approaching creditworthiness (and lots that aren’t) offer negative yields.

So it seems a little odd that anyone is pondering whether government bonds should still have a place in protecting investors’ portfolios. At best, they can’t provide anything like as much protection as before and, if history is any guide, they will amplify losses.

Central banks have been mainly responsible for bonds’ vertiginous performance in recent years. Since February this year, according to the International Monetary Fund, up to 75% of government bonds have been hoovered up by countries’ own central banks. This excludes the buying of other nations’ sovereign debt by foreign reserve managers, not least those of China, Japan and Switzerland, anxious to stop their currencies rising. The combined domestic and foreign assets of just the central banks of the U.S., China, Japan and the euro zone reached $26 trillion at last count, according to Yardeni Research. It was $5 trillion in 2007.

No wonder bonds have become breathtakingly expensive. As well as vanishingly small yields, you can measure this by something called the term premium. This is the extra return investors should receive compared with expected short-term rates, because the world can change over the life of longer-dated bonds.

Calculating a forward-looking term premium is tricky; different models come up with different answers. All have fallen precipitously. In the U.S., the expected 10-year term premium over cash peaked at roughly five percentage points a year in the early 1980s. On UBS estimates, the U.S. 10-year term premium is now some 29bps. The Federal Reserve’s calculations put the U.S. term premium at about minus 50bps. These numbers have never been lower.

This matters to the overall risk of portfolios. Those that contain negatively correlated assets — that is, some go up when others go down — are less volatile. How much less volatile depends on how much they both move. Unfortunately, the kick that portfolios receive from government bonds when equities have dropped sharply has been falling in recent years, because the starting yield in each big risk sell-off has been lower than the one before and investors have balked. By definition this means that portfolios — and equities — are riskier.

And from today’s starting point there is, at best, very little potential offset because there’s a limit to how far yields can fall. At some level investors might just as well buy a safe. Ask yourself the following: How far would yields on, say, German bunds (currently about minus 60bps) need to fall to offset a slump in equity markets? The answer you end up with makes no sense. If other central banks take the Bank of Japan approach and lock yields at a certain level, prices wouldn’t move at all, so there would be no offsets.

The biggest question is whether government bonds themselves become risk assets and go in the same direction as equities. From their current levels, yields will have to go up at some stage, and bond prices fall. Will the price of equities? Correlations between bonds and equities only turned negative in the late 1990s. Before that — and in the U.K. for at least the previous 250 years, according to the Bank of England — the two markets moved in the same direction. If that were to happen again, then rising bond yields would mean falling equities, too.

While you can’t say for certain this will happen, you can say that the risk of correlations never going positive again is massively mispriced and the dangers are huge. First, all financial assets — bonds, equities, credit, you name it — are expensive. Second, investors in government bonds are essentially buying zero-coupon bonds. These have a higher duration than those with an interest rate (for a given change in yield, the price will mathematically move more) so any rise in yields will mean greater losses than many expected. Third, institutional investors have often leveraged bond portfolios to boost returns.

Worst of all, pretty much every institutional investor uses a risk-management system that assumes a negative correlation between bonds and equities. If this flips around then they would have to cut all risk willy-nilly. All these things mean that a correlation turnaround, were it to occur, would be fast and savage.

A change in correlations would be dreadful for pretty much all financial assets. Some of the most popular investments would be hammered. Corporate bonds would suffer from rising government yields, wider credit spreads and vanishing liquidity. At some stage bank shares would start to outperform, since they would at least have yields and credit spreads. But you’d first have to get through the carnage of bad investments made when they didn’t.

The biggest beneficiary would be volatility. Many institutional investors sell options (and thus volatility, the main variable in an option’s price) to boost returns. In a positive correlation world, you should be buying volatility. And leaving a large slug of your money in cash: Earning nothing is better than losing a lot.  

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Richard Cookson was head of research and fund manager at Rubicon Fund Management. He was previously chief investment officer at Citi Private Bank and head of asset-allocation research at HSBC.

©2020 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.