About that Treasury ‘tsunami’ . . . 

Receive free Sovereign bonds updates

We’ll send you a myFT Daily Digest email rounding up the latest Sovereign bonds news every morning.

As widely expected (and trailed on Monday), the US Treasury has just announced that it will be jacking up the size of its debt auctions to rebuild reserves and finance its growing budget deficit.

The Treasury expects it will have to sell more than $1tn of bills and bonds in the coming three months. Here is a table showing the auction sizes from the May-to-July 2023 quarter (in billions of dollars), and the expected sizes in the coming three months.

The post-debt-ceiling deluge of Treasury bills failed to spark the financial carnage that some were predicting/hoping for, and this increase in auction sizes will in effect push out that borrowing into longer-dated tenors of notes and bonds.

A couple of interesting points in the table above: The rise in long-term debt sales will be front-loaded, with the biggest jump in 10-, 20- and 30-year debt coming this month. Issuance of shorter-maturity notes (2s, 3s and 5s) will rise in a steadier fashion.

This makes sense if the yield curve is on its way to un-inverting itself at some point in the next year; it’s better to sell more long-dated bonds when long-term rates are relatively low. Strategists at CreditSights expect the yield curve to right itself in the next year. Other banks, such as Deutsche Bank, frame this as “increasing term premium”, which is sort of a catchall term for a steeper (or less inverted) yield curve.

In any event, the chunkiness of the coming US government coupon-bearing debt sales — at a time when the Fed is still shrinking its balance sheet — is again worrying some people.

And to be fair, the numbers are big! Barclays caused a minor stir yesterday with a note on what they termed the coming “Treasury tsunami”, which they expect to be a major force beyond the current quarter (their emphasis below):

Given that borrowing needs are likely to remain persistently high (as indicated by the Treasury’s Q4 borrowing need of $850bn), we believe this process of raising coupon auction sizes will continue for a few quarters. Even as the Fed ends QT in the first half of next year, wide budget deficits will ensure that the Treasury would still need to issue more notes/bonds to keep the share of T-bills within the desired range. We believe auction sizes will eventually rise to levels beyond the COVID peak (with the exception of 7y and 20y, for idiosyncratic reasons). Hence, we expect the Treasury to signal that further increases are needed at the upcoming meeting. We would not be surprised if net issuance of notes/bonds to investors were to be close to $2trn in CY 24, vs. $1tn in CY 23.

As roughly eight quadrillion pearl-clutching headlines have asked over the past 10-15 years, who on earth is going to buy all those Treasuries?

Well, as you might have noticed at some point over the past few decades, the answer is pretty much everyone. A lot of demand exists for the dominant global reserve asset, especially now that it yields roughly what junk bonds did a few years ago.

A recent BIS paper on the effect of “quantitative tightening” gives an indication of what kind of impact increased net supply might have on Treasury yields. With Alphaville’s emphasis:

We estimate that a 1 percentage point increase in long-term yields leads to an 11% increase in the demand by non-central bank players for US Treasury securities. Based on estimates of the yield elasticity of different sectors, we infer the market-clearing yields under different quantitative tightening scenarios. For example, we find that, all else constant, a hypothetical reduction in the central bank balance sheet of around $215 billion leads to an increase in long-term bond yields by 10 basis points in the United States. Our estimates are quantitatively close to those for the euro area and they are also in more or less the same ballpark as estimates of the impact of quantitative easing obtained via other methods.

This is admittedly a bit rough, but holding everything else constant, that means Barclays is forecasting a ~$1tn increase in US government debt supply, but only a 46bp increase in the 10-year Treasury yield.

It’s possible that the price impact will be different when it’s the Treasury Department that is increasing supply, rather than the Federal Reserve letting its bonds mature. But it’s hard to see why, because the immediate effect of the Fed’s balance-sheet shrinkage was that it cut back on its purchases of Treasuries at auctions, leaving more supply for governments, banks, investment funds and other buyers.

Anyway, “all else constant” is the actual giveaway. What will actually move the bond market is the shifting outlook for economic growth, inflation and interest rates. If the US economy suddenly starts teetering, Treasury yields will plummet no matter what the supply dynamics are. And if the economy accelerates then yields will rise, even if tax income increases and borrowing needs fall.

And no, the Fitch downgrade won’t matter either. C’mon.