The Return of the Bond Vigilantes or the Start of the 'Roaring 20s'?

There has been enormous focus on bond supply as a determinant of long-bond yields, but macro fundamentals and Fed policy still dominate

May 11, 2024

We have seen an enormous amount of attention from investors on the subject of US sovereign issuance and bond supply and its impact on long-term bond yields and term premium. Treasury auctions and quarterly refunding announcements (QRA) have become important events for macro investors to watch.

While the US deficit and consequent UST issuance are historically (and, importantly, pro-cyclically) elevated, the evidence for a persistent, significant and independent, causal impact of duration supply is challenging to detect.

The academic literature and our own empirical findings suggest a few key lessons:

Monetary policy and the expected direction of short-term interest rates still matter a lot;
Macro economic fundamentals (inflation and growth) are critical drivers of moves in long-term bond yields;
Total coupon supply matters more than the composition of coupon issuance. The volatility in debt/deficits is much greater than the volatility in the WAM of issuance;
The consolidated public sector (Fed and US Treasury) balance sheet is what matters most.

I am a "flows guy" so on a deep level I feel like bond supply really should matter. But at ExanteData we like to look at academic work for broad lessons and inspiration, while also doing our own empirical work as well. We try and take a balanced approach.

For clients, we have been focused on the topic of bond supply for many months.

Below we share some of what we have found (or not found).

Monetary Policy Still Matters

Short-rates and short rate expectations still matter. A lot. The simplest way to see this is a scatter of the daily changes in 3 month fed funds futures and US 10y yields. The former should capture day to day changes in short-rate/policy-rate expectations fairly cleanly, while the latter should in theory reflect the expected average short-term interest rate over the next 10 years plus a term premium (TP).

As we can see, roughly half the day to day variation in long-term bond yields is simply a function of changes in near-term policy rate expectations (i.e., the Fed).

Another way to see this is a simple model that tries to proxy the term premium. Below we show the cumulative residual term—the unexplained portion of moves in 10y UST yields—from a simple model of yields as a function of short-rate expectations (fed funds futures) and moves in oil prices. This is plotted against a fancier academic estimate of term premium such as the ACM model (link).

There definitely was a TP shock in Q3/Q4 last year, but the cumulative model residual and the fancier ACM estimate of the TP are little changed/negative in recent months.

If bond supply was driving changes in TP—that is driving changes in long-term yields over and above changes in the expected path of short-term rates—we might expect changes in the residual term from the model above to be explained by proxies of bond supply.

But the event study below, which uses intraday price action around QRAs as a proxy for unexpected supply shocks, does not suggest that announcements about Treasury borrowing are having a meaningful impact on our TP proxy. The chart below shows the cumulative residual from our TP proxy since October 2022 around QRA announcements and outside them.

Finally, taking inspiration from this paper (link), we looked at the longer-term of yields around Fed meetings, Treasury auctions and QRA announcements. Over the last 30 years a very significant share of the cumulative decline in 10y yields has occurred only on FOMC meeting days (yellow line) while a much smaller share of the cumulative decline has occurred on days with UST auctions.

But yields also declined on auction days on net despite US debt/GDP rising by from 50% to 120% over this period.

The chart below does the same comparison of FOMC days and QRA days (since 1997) and the result is broadly similar.

In short, Fed policy and policy expectations still matter a lot—both over the short-term as shown in the scatter above and over the long-term as shown by the cumulative decline in yields around FOMC meetings relative to proxies of bond supply (auctions and QRAs). However we acknowledge there are some interesting signs of this relationship breaking down in this cycle: Since the Fed began hiking in 2022, yields are up more than 250bps and yet they have declined on net on FOMC days since then. Nevertheless we still expect that if/when the Fed embarks on a cutting cycle long bond yields will feel the gravitational pull (down) from short-term rate expectations.

However if they do not—that is, if long bond yields do not fall despite a shift down in the expected path of policy—then we would expect to see estimates of the term premium rise and we would have to reevaluate our findings.

It’s The Data: Macroeconomic Fundamentals Matter

Another key lesson from our empirical work is that fundamentals—i.e., economic data—really matter.

A very large share of the increase in long-term bond yields (10y and 30y in particular) over the last 18 months have occurred just over the 30 min around major economic data releases such as NFP/CPI/PCE (blue bars).

But again, we find that the cumulative move in yields in the 30min around supply shocks—proxied by UST auctions and QRA announcements—has been negative. Nor has this patterns shown much sign of variation across tenors i.e., steepening/changes in TP on concerns about Treasury duration extension.

In fact if we look at moves in yields just on QRA announcement days, yields often move down around the announcements themselves, there is not clear pattern of yield increases and the moves across tenors are generally symmetrical.

Some proponents of the thesis that bond/duration supply is driving yields point to the dramatic increase in fiscal deficits and net issuance since COVID as being orders of magnitude larger than the past. But duration- and GDP-adjusted private marketable coupon supply (grey line) has only been marginally above (green line) the post-GFC peaks in 2009 and 2019 (red line) in recent months. In fact the post-COVID peak in private coupon supply was in 2021—well before concerns about duration supply or moves in long-bond yields, steepening and TP came to the fore.

So it is not duration supply itself that is dramatically different this time. What is different is the economic fundamentals. Inflation is much higher than in the post-GFC decade—and growth is stronger and unemployment lower.

Hence, to the extent that bond supply is a derivative of large deficits, and that both strong growth and inflation pressure and the Fed’s policy response to these fundamentals reflect the role of pro-cyclical fiscal policy, then yes, higher bond yields do reflect the impact of bond supply. But this is not the argument the new ‘bond vigilantes’ are typically heard making.

The question is whether bond/duration supply itself has an independent causal impact on long-term yields (and the TP) over and above the impact of fiscal policy on macroeconomic fundamentals. On that count, evidence from the empirical findings above and the academic literature are decidedly mixed.

What Do the ‘Doctors’ Say

The impact of bond supply on yields and term premia has received a lot of academic focus by those with more time and fancier degrees than we possess (see the appendix for some selected further reading). And indeed a recent Fed paper (link) surveying the literature on this topic (table below) finds that the average impact of a 1 percentage point increase in debt/GDP is about 3-6bps (the authors propose a new model and their own estimates are on the lower side of this range).

If we take the 3-6bps literally and compare the CBO baseline for US debt to GDP in 2030 before the pandemic and their latest projection, debt to GDP is expected to be about 10-12 percentage points higher than before. This suggests that long-bond yields or TP 2s10s might be expected to be about about 50bps (midpoint of 30-72bps) higher than would otherwise have been the case.

If we take the average TP over the last hiking cycle (2015-2019) as a baseline (-30bps) this suggests TP could be in the 0-40bps range going forward while 2s10s could steepen from an average of about 75bps over the prior cycle to perhaps 125bps over this cycle. If we were to take the Fed’s SEP projections on the longer-run fed funds rate of 2.5% —though rising recently—that would suggest a 10 yield perhaps in the 4-4.5% range.

We are sympathetic to the findings of Greenwood & Vayanos (2023) for instance, suggesting that duration supply leads to curve steepening but that the composition of coupon supply is not a big driver of the curve shape (see chart from the paper below).

Reading the academic literature there were really a few key/consistent findings we think are interesting and worth bearing in mind (we plan to touch on 4 and 5 in particular in future posts):

The impact of bond supply is non-linear and time varying
Monetary policy and monetary policy expectations dominate in equilibrium
The total amount of issuance and the split between bills and coupons matters much more than the composition of coupon supply
The consolidated public sector balance sheet (Fed + Treasury) is the key
Long-term yields in the US are impacted by developments abroad (eg global safe asset supply). Post-covid bond supply and duration extensions are not only “made in the USA.”

https://www.hbs.edu/ris/Publication%20Files/Debt%20Management%20Conflicts%20between%20the%20US%20Treasurey%20and%20the%20Federal%20Reserve_623991cb-69a8-47f2-95e7-17df71edd18e.pdf

https://www.brookings.edu/wp-content/uploads/2016/06/16-monetary-policy-zero-lower-bound-williams.pdf

https://www.newyorkfed.org/medialibrary/media/research/advisory_panel/far/nyfed_far_dec2020_greenwood.pdf?la=en

https://www.nber.org/system/files/working_papers/w31879/w31879.pdf

https://www.federalreserve.gov/econres/feds/government-debt-limited-foresight-and-longer-term-interest-rates.htm

The content in this piece is partly based on proprietary analysis that Exante Data does for institutional clients as part of its full macro strategy and flow analytics services. The content offered here differs significantly from Exante Data’s full service and is less technical as it aims to provide a more medium-term policy relevant perspective. The opinions and analytics expressed in this piece are those of the author alone and may not be those of Exante Data Inc. or Exante Advisors LLC. The content of this piece and the opinions expressed herein are independent of any work Exante Data Inc. or Exante Advisors LLC does and communicates to its clients.

Exante Advisors, LLC & Exante Data, Inc. Disclaimer

Exante Data delivers proprietary data and innovative analytics to investors globally. The vision of exante data is to improve markets strategy via new technologies. We provide reasoned answers to the most difficult markets questions, before the consensus.

This communication is provided for your informational purposes only. In making any investment decision, you must rely on your own examination of the securities and the terms of the offering. The contents of this communication does not constitute legal, tax, investment or other advice, or a recommendation to purchase or sell any particular security. Exante Advisors, LLC, Exante Data, Inc. and their affiliates (together, "Exante") do not warrant that information provided herein is correct, accurate, timely, error-free, or otherwise reliable. EXANTE HEREBY DISCLAIMS ANY WARRANTIES, EXPRESS OR IMPLIED.

Money: Inside and Out

The Return of the Bond Vigilantes or the Start of the 'Roaring 20s'?

There has been enormous focus on bond supply as a determinant of long-bond yields, but macro fundamentals and Fed policy still dominate

Monetary Policy Still Matters

It’s The Data: Macroeconomic Fundamentals Matter

What Do the ‘Doctors’ Say

Further Reading