By Mike Dolan
LONDON (Reuters) – If not now, when? Investors typically demand some added compensation for holding a security over many years to cover all the unknowables over long horizons – making the absence of such a premium in bond markets right now seem slightly bizarre.
Disappearance of the so-called “term premium” in 10-year U.S. Treasury bonds over the past 5 years has puzzled analysts and policymakers and been blamed variously on subdued inflation expectations or distortions related to central bank bond buying.
And yet it’s rarely, if ever, been more difficult to fathom the decade ahead – at least in terms of inflation, interest rates or indeed quantitative easing or tightening.
Inflation is running at a 40-year high after the pandemic forced wild swings in economic activity and supply bottlenecks and was then compounded by an energy price spike due to war in Ukraine that may redraw the geopolitical map.
The U.S. Federal Reserve and other central banks are scrambling to normalise super easy monetary policies to cope – not really knowing whether to focus on reining in runaway prices or tackle what Bank of England chief Andrew Bailey this week described as a “historic shock” to real household incomes.
Bond yields have surged, much like they did in the first quarter of last year. But this time bond funds have suffered one of their worst quarters in more than 20 years and some measures of Treasury price volatility are at their highest since banking crash of 2008.
But the most-followed estimates of term premia embedded in bond markets remain deeply negative. And this matters a lot to a whole host of critical bond market signals, not least the unfolding inversion of the U.S. Treasury curve between short and long-term yields that has presaged recessions in the past.
“The 10-year term premium has barely budged even as inflation spiked to 8%, suggesting that long-dated yields are probably still capped by the Fed’s record-high balance sheet,” said Franklin Templeton’s fixed income chief Sonal Desai. “Or maybe investors think the Fed will blink and ease policy again once asset prices start a meaningful correction.”
“In either case, I think markets are still underestimating the magnitude of the monetary policy tightening ahead,” said Desai, adding that expectations of another more than 2 percentage points of Fed hikes this year still likely leaves real policy rates deeply negative by December even if inflation eases to 5%.
Graphic: US ‘term premium’ stays negative-https://fingfx.thomsonreuters.com/gfx/mkt/byvrjbqamve/One.PNG
Graphic: Fed contrast between Yield Curve and Near Term Forward Spread-https://fingfx.thomsonreuters.com/gfx/mkt/lbvgnmlxwpq/Two.PNG
BUMP IN THE NIGHT
So what’s the beef with the term premium?
In effect, the Treasury term premium is meant to measure the additional yield demanded by investors for buying and holding a 10-year bond to maturity as opposed to buying a one year bond and rolling it over for 10 years with a new coupon.
In theory it covers all the things that might go bump in the night over a decade hence – including the outside chance of credit or even political risk – but it mostly reflects uncertainty about future Fed rates and inflation expectations.
At zero, you’d assume investors are indifferent to holding the 10-year today as opposed to rolling 10 one-year notes.
But the New York Fed’s measure of the 10-year term premium remains deeply negative to the tune of -32 basis points – ostensibly suggesting investors actually prefer holding the longer-duration asset.
Although the premium popped back positive in the first half of last year, it’s been stuck around zero or below since 2017 – oddly in the face of the Fed’s last attempt to unwind its balance sheet.
And the persistent and puzzling erosion of the term premium to zero and below brings it back to the 1960s, not the much-vaunted inflation-ravaged 1970s that everyone seems to think we’re back in.
It matters a lot now as the debate about the inversion of the 2-10 yield curve heats up and many argue that the signal sent by that inversion is less clear about a coming recession as it’s distorted by the disappearance of the term premium.
In the absence of a term premium, the long-term yield curve is just a reflection of long-run policy rate expectations that will inevitably see some retreat if the Fed is successful in taming inflation over the next two years.
Fed Board economists Eric Engstrom and Steven Sharpe late last week also dismissed the market’s obsession with a 2-10 year yield inversion signalling recession.
In a blog called ‘(Don’t Fear) The Yield Curve’ they said near term forward rate spreads out to 18 months were much more informative about the chance of a looming recession, just as accurate over time and – significantly – heading in the opposite direction right now.
The main reason they pushed back on the 2-10s was it contained a whole host of information about the world beyond two years that’s simply less reliable as an economic signal and “buffeted by other significant factors such as risk premiums on long-term bonds.”
But what could see the term premium return?
Presumably the Fed’s planned balance sheet rundown, or quantitative tightening (QT), would be a prime candidate if indeed its long-term bond buying has distorted term premia.
But the last Fed attempt at QT in 2017-19 didn’t do that and Morgan Stanley thinks it will be some time yet before just allowing short-term bonds on its balance sheet to roll off and mature gets replaced by outright sales of longer-term bonds.
“QT is not the opposite of QE; asset sales are.”
Of course, maybe the world just hasn’t changed that much – in terms of ageing demographics, excess savings and pension fund demand, falling potential growth and negative real interest rates. Once this current storm has passed, investors seem to think that will dominate once more.
Graphic:Fed balance sheet and maturities- https://fingfx.thomsonreuters.com/gfx/mkt/lgvdwqobapo/Three.PNG
The author is editor-at-large for finance and markets at Reuters News. Any views expressed here are his own
(by Mike Dolan, Twitter: @reutersMikeD. Editing by Jane Merriman)
