There Is a Long-Term Danger Lurking in Financial Markets

There is much talk among investors of a “new regime.” Those words headline BlackRock’s recent midyear outlook, for example, which emphasized how the trends of the past decade—low interest rates and volatility, with bonds cushioning stock selloffs—have reversed in the past two years.

Weirdly, however, one indicator from the era of near-zero rates has remained unchanged: Markets don’t demand much extra return in exchange forlocking away their money for longer periods.

This can be inferred from data published daily by the Federal Reserve, which estimates the “term premium” needed for investors to buy 10-year government bonds, rather than stash their money in, say, a money-market fund.

In theory, the yield on risk-free government paper should be equal to the interest rate investors expect on average until the debt is paid back. In practice, this only holds true for short-term bonds. Unless an investor or a market maker is 100% sure that a 10-year bond will be held to maturity, there is a risk that it will need to be sold at an unfavorable time. Without a discount for 10-year paper, giving a yield premium, it would make more sense for investors to roll over 10 one-year bonds.

This term premium has added an average 1.5 percentage points to 10-year yields since 1961. After 2016, though, ultralow rates, quantitative easing and low inflation turned it negative. Even now that inflation has shot up and rates have risen, this anomaly persists: On Friday, the term premium was a negative 0.75 percentage point. This is deeply odd: If anything should make investors wary of long-term assets, it is hawkish and unpredictable central bankers.

Should term premiums return to the historical average, the consequences would be dire for investors. Yields on 10-year Treasurys would jump to 6.1%, and there might be similar effects with global equivalents such as German bunds.

Also, government paper serves as a benchmark for other assets, particularly the debt of blue-chip companies. At constant credit spreads, U.S. investment-grade yields would rise from 5.5% to 7.8%. The S&P 500, which using 12-month earnings forecasts offers a yield premium of 1.4 percentage points over Treasurys, would instead yield 1 percentage point less, undermining an argument for buying stocks at today’s prices.

Is it time to abandon ship? It isn’t so straightforward.

Investors’ guesses about future interest rates aren’t recorded anywhere, so the Fed can’t directly observe the term premium. It estimates it based on regressions, following a 2008 paper by Tobias Adrian, Richard K. Crump and Emanuel Moench. The results could simply be wrong.

Steven Major, global head of fixed income research at HSBC, points out that the premium is actually positive if we use a rule of thumb and trust officials’ latest “dot plot,” which shows expectations that U.S. rates will average 3.2% over the next decade. With 10-year Treasurys yielding 3.8%, the implied term premium is 0.6 percentage point.

“The term premium was something invented by bond analysts to bamboozle their clients,” Major said.

Indeed, the Fed’s calculations of the term premium may mostly track the Treasury yield curve, which has been inverted for a year. The two have historically moved in lockstep, and the premium could normalize now that the curve is steepening again. But because this is happening for benign reasons—U.S. inflation cooled to 3% in June—it need not mean scary high yields.

Even if the Fed’s estimates are right, it may be that the financial ecosystem has been forever transformed by big institutional investors with an insatiable hunger for long-term assets, such as pension funds, life insurers and foreign-exchange reserve managers.

Or perhaps it is the fact that since 2008 rate setters have provided careful guidance on future policy that has reduced the perceived uncertainty of holding long-term assets. At times when economists’ long-term rate forecasts, as collected by the Federal Reserve Bank of Philadelphia, are more in agreement, the term premium seems to be lower.

Whatever the explanation, it is hard to deny that 10-year bonds offer scant compensation when 2-year Treasurys yield 4.7% and central bankers retain a hawkish bias.

Though flawed, term-premium estimates capture something real: Markets don’t require much of an incentive to favor long-term investments. This has been so for a decade and remains the case today, as the speed with which money has flowed back to technology firms this year based on hypothetical artificial-intelligence gains has again highlighted.

Whether it happens in a sudden rout or a slow creep, investors can’t dismiss the possibility that, one day, this trend will reverse too.