Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
Why have Treasury bonds sold off lately? The obvious answer is that Kevin Warsh isn’t exactly convincing investors that he is willing to do what it takes to bring inflation down to the Federal Reserve’s target.
But there is another, subtler explanation: the Treasury buyer base has changed over the past decade, with the influence of more price-agnostic central banks ebbing and the importance of price-sensitive private investors increasing sharply.
Alphaville touched upon this in a big post last week, which focused on the swelling hedge fund involvement in the US government bond market. But they are just one part of this broader trend (and one could argue that they are merely covering Treasury-shaped bags for mutual funds).
Barclays’ rate analysts Demi Hu and Anshul Pradhan tackled this phenomenon in a report last week, and reckon it has major implications for how Treasuries trade. Their emphasis below:
Historically, a large share of Treasury demand came from investors with structural or policy-driven reasons to own Treasuries, including foreign official reserve managers and the Federal Reserve. In response to the COVID shock, the Fed expanded its Treasury holdings through large-scale asset purchases, with SOMA holdings ultimately reaching $6trn. Subsequently, since the start of balance-sheet runoff in 2022, Fed’s Treasury holdings have fallen to about $4trn. Foreign official holdings have also failed to keep pace with Treasury issuance, as they have moved broadly sideways at around $4trn over the past decade, despite the continued expansion of the Treasury market.
As a result, private buyers such as mutual funds, foreign private investors, banks, and households have increasingly filled the gap. Total marketable Treasury debt outstanding increased from roughly $4trn in 2006 to $29trn in 2026, an increase of nearly $25trn. While private investors absorbed approximately $19trn of that increase, much of the shift in ownership composition occurred over the past decade. In 2006 and 2016, official and private investors each held roughly half of the Treasury market. By 2026, however, the private sector’s share had risen to 73% while the official sector’s share had fallen to 27%.
As a result, demand for Treasuries “has become materially more valuation-sensitive”. But what does this look like in practice?
To figure this out, the Barclays analysts have constructed an index of the US government bond market’s “elasticity”. This elasticity gauge is cobbled together from measures of price sensitivity derived from Treasury ownership and yield data, with an index reading of 1 indicating that they care little what yields are, the holding is mostly for structural reasons.
The data on individual buyer bases — such as domestic banks, overseas mutual funds, international central banks, or US pension plans — actually reveals that most investors have become less yield-sensitive. But because of the shift in who buys the most Treasuries, the aggregate elasticity index has climbed sharply over the past decade:
It’s important to note that a higher elasticity doesn’t necessarily mean that demand is weakening and yields must therefore be higher. It merely indicates that demand is becoming more sensitive to the level of yields, and that probably implies that they must be high enough to entice investors.
By using “Autoregressive Distributed Lag” models (don’t even ask), some NY Fed survey data and fancy maths, Hu and Pradhan show that this mainly plays out in the term premium — basically the extra yield that investors want to hold, say, a 10-year Treasury bond for 10 years rather than reinvesting one-year T-bills every time they mature into new one-year T-bills for the next decade.
As the New York Fed’s Roberto Perli also told the central bank’s policymakers last month, according to the minutes released later:
The ownership composition of Treasury securities has shifted somewhat over the past several years from relatively price-insensitive official-sector holders to more price-sensitive private investors, which could have implications for the term premium component of yields.
Term premiums can be a bit of a hand-wavy factor (as beloved former Alphavillain Matt Klein was fond of pointing out). And sure, we’ve scratched our head as to whether we really even need term premia models.
But they do seem to have been creeping up, at least according to the NY Fed’s own model, and are now firmly back in positive territory once more.
But Barclays reckons this could merely be the start. Given the growing influence of price-sensitive Treasury investors, term premiums could creep back to their pre-global financial crisis levels.
This suggests that, should the other factors also reprice higher, there is room for term premium to rise even if it looks as if it has returned to pre-GFC levels. For instance, rate volatility is below pre-GFC levels and has room to rise in a changing Fed communication regime. Fed balance-sheet shortening could pose additional upside risk to the term premium if the Treasury does not fully offset the Fed actions, and inflation risk premia may rise if confidence in the inflation anchor weakens. At the same time, global duration supply pressures remain elevated from the worsening of persistent fiscal deficits and increasing debt issuance across fixed-income markets.
We therefore view the shift in Treasury ownership as reinforcing the broader case for a return toward a higher term-premium regime more like that of the pre-GFC period than what largely has prevailed since.
In case you’re curious, the NY Fed term premium averaged 1.3 percentage points in the decade leading up to the summer of 2007, almost twice its current level. And all things being equal, that implies higher Treasury yields — whatever tricks Scott Bessent might try to pull.
