August 2026: Where Does the Warsh Fed Go from Here and Outlining an End of Cycle Scenario
Last month’s Treasury Talk column discussed:
- The compounding of cycle-end risks, with credit and US recession risks remaining underpriced amid escalating geopolitical and commodity shocks;
- The Strait of Hormuz’s continued closure as a potential “US Suez Canal” moment, and the parallels to the UK’s post-1956 retrenchment from global hegemon and reserve currency;
- A Fed that remains late on inflation;
- A US AI narrative looking increasingly stretched.
In July, I wrote that I did not expect the Warsh Fed to “surprise” markets at the July meeting with a 25 bps rate hike— old habits die hard — and that I would instead look for Warsh to outline his policy views at Jackson Hole in August, with a first rate hike coming at the Sept FOMC. While I continue to believe the widening geopolitical conflict keeps commodity prices high and inflation elevated, I revise my prior base case of a Sept Fed rate hike to no rate hike.
The July FOMC meeting has now come and gone, and it is worth taking stock of what happened, what it revealed about the reaction function of the Warsh Fed and concluding with some thoughts on an end-of-cycle scenario.
My view is that the rise in long-dated Treasury yields reflects growing concerns about
- Warsh Fed credibility,
- the prospect for a further rise in inflation as a result of the worsening geopolitical conflict and associated pressure on global commodity prices; and
- risk premia coming into long-dated TIPS associated with potential downward revisions to US inflation measurement and potentially even the Fed’s inflation target itself.
Long Dated Treasuries Didn’t Like July FOMC
As I expected, the Warsh Fed did not deliver a 25 bps rate increase in July. The FOMC chose to prioritize the front end of the Treasury curve rather than deliver a July hike that wasn’t fully priced in. Three voting Reserve Bank presidents dissented, preferring to raise rates by 25 bps, a notable signal of FOMC discomfort with standing pat against a backdrop of above-target inflation and an elevated commodity price outlook. We will know more when the July FOMC minutes are released tomorrow.
But holding the front-end steady came at a cost to long-dated Treasuries. The decision — or more precisely, the lack of messaging around it — was negative for long-dated Treasuries. And it was less the July FOMC decision itself and more Chair Warsh’s comments during the press conference that seemed to pressure long-dated Treasury yields the most (chart of long-dated Treasury yields on July FOMC intraday).
A Casual Tone on Getting Inflation Back to Target Where Markets Wanted More Substance
What was most striking about the July FOMC press conference was the lack of specificity and, indeed, somewhat casual tone of Fed Chair Warsh’s comments. Some observers have rallied to Warsh’s defense, decrying a bond market that wants to continue to be “spoon-fed” with forward guidance. But that misses the point. While I continue to wish Chair Warsh success, his main focus seemed to be less on explaining how he intends to restore price stability — which, recall, he framed inflation in recent Congressional testimony as a “choice,” stating “we are committed to 2% goal” — and more on his taskforces, which he referenced no fewer than eight times in the press conference.
For a Fed that already looks behind the curve, communicating with a light touch on the strategy to restore price stability was not reassuring. When asked about the upcoming August Jackson Hole speech (Aug 28), Warsh indicated he was still trying to decide on what tack to take. Again, it does not demonstrate a well-developed strategy.
Long-dated Treasury yields have risen notably since June with much of the rise in yields reflecting higher real interest rates on Treasury Inflation Protected Securities (TIPS).
Could Long-dated TIPs Be Pricing in a Risk Premia Related to Possible Changes/Downward Revisions to US CPI Methodology?
Since my last Treasury Talk column, I wrote an op-ed in American Banker and also was a guest on the WSJ’s “Take On the Week” podcast where I had the opportunity to talk a bit about my views on possible drivers behind the notable rise in long-dated TIPS real yields.
Typical explainers/drivers of long-term TIPS real yields are thought to be 1) economic growth expectations, 2) term premium, and 3) expectations of long-run neutral rate as opposed to near-term Fed meeting expectations. At the July FOMC press conference, Chair Warsh evoked US economic growth expectations as explaining the move higher in long-dated TIPS real yields which have also dragged long-dated nominal Treasury yields higher.
However, a quick look at Bloomberg consensus forecasts for US GDP growth does not point towards an anticipated surge in the US economy.
Cameron Crise who writes for Bloomberg also notes that historically 10-year TIPS yields are not associated with higher future US economic growth.
(Source, Cameron Crise Macro Man column on Bloomberg, 8/5/2026)
Changes in long-dated TIPS real yields do seem to correlate better with the rise in estimates of the 10-year nominal Treasury term premia (ACM model). So term premia may be a factor.
Of course, there is also the potential that the rise in TIPS real yields reflects shifts in savings/investment balance, or the neutral rate of interest, r*. Indeed, the US personal savings rate has declined sharply this year, weighed on by elevated inflation and dis-savings associated with aging demographics.
This decline in US savings comes even as both Treasuries and US hyperscaler debt issuance to fund AI data centers continue to grow rapidly.
Indeed, the significant growth in the size of the Treasury market requires pulling in ever more mark-to-market sensitive buyers vs buyers of US government bonds with more stable balance sheets such as foreign central banks, the Fed, pension funds and insurance companies.
Yet, whereas early on in the Iran conflict oil price moves (yellow) were having a larger impact on long-dated TIPS (white) and nominal Treasury (green) yields — that correlation has broken down since Warsh stepped in as Fed Chair at the June and July FOMC meetings (red lines).
So perhaps a combination of term premia and Chair Warsh’s shift in Fed communications which began at the June FOMC meeting may explain the rise in long-dated Treasury yields.
This brings us to debates about whether Chair Warsh’s desire to communicate less may result in higher interest rate volatility. That seems plausible but for now measures of option implied rate volatility (the MOVE index) don’t seem associated with recent movements in the 30-year TIP real yield.
Could there be any other factors that might contribute to the rise in long-dated TIPS yields? That question seems insufficiently explored — specifically, Boskin Commission 2.0 risk.
Fed Chair Warsh in his July 15th testimony before Congress referenced US CPI and “whether we can do better” (19:04):
“In the last couple of days, we’ve gotten data on the consumer price index — today on the producer price index. Any central bank would be happy to have the data going in the right direction. My view is these are all imperfect measures of the state of underlying inflation. So one of the task forces is going to see whether we can do better—have better data from external sources, and even better ideas as to how important organizations like the Bureau of Labor Statistics, the Bureau of Economic Affairs might think about how they could do a better job in an evolving economy.”
Even if Chair Warsh did not intend to throw his weight behind methodological revisions of US CPI, some in the bond market may remember Fed Chair Greenspan’s 1995 Congressional testimony in which he argued that US CPI overstated inflation. What followed from Fed Chair Greenspan’s statement was the Boskin Commission, which concluded in 1996 that US CPI overestimated inflation due to quality and substitution biases. In addition to kicking off changes to inflation measurement, Chair Greenspan also ended up changing the Fed’s inflation target from CPI to core PCE in 2000.
The Boskin Commission’s findings and subsequent revisions to US CPI methodology reduced measured CPI by ~1.1 percentage point per year going forward and impacted estimates of US economic growth, productivity, reduced Social Security cost-of-living adjustments and affected interest rate policy at the Federal Reserve. All the Commission’s methodological recommendations reduced estimated US CPI.
The Boskin Commission’s timing — exactly as Social Security pressures mounted in the mid-1990s — suggested to some that the downward revisions to CPI methodology may have also considered the US fiscal picture. The renewed approach of Social Security trust fund solvency pressures in the next several years is a notable parallel to the mid-1990s.
In looking over the history of major US CPI methodology revisions in my lifetime — Boskin Commission, shift to hedonic/quality adjustments, use of a geometric means vs arithmetic mean for CPI components to account for consumer substitution —there does not seem to be a meaningful example of a historical CPI methodological change that increased reported CPI. The point here is that history alone would suggest that it is rational for TIPS investors to expect any methodology changes to US CPI to continue to revise downwards measured/reported inflation that TIPS are linked to.
But recall that former Fed Governor Miran argued in a Dec 2025 speech that the Fed was maintaining an overly restrictive stance due to structurally flawed inflation metrics.
Recall also that after the Administration removed the BLS commissioner in 2025, EJ Antoni was initially nominated to head the BLS. BLS is responsible for computing US CPI. In a Feb 2026 blog post, Antoni advocated using a private sector measure of US CPI – “Truflation” — which is ~1 percentage point lower than BLS reported CPI — and also noted that such a change in US inflation measurement would allow a Warsh Fed to lower interest rates. Ultimately, Antoni’s BLS nomination did not move forward but it certainly raises questions about the trajectory for CPI revisions.
Now let’s return to TIPs’ real yields. Revisions to CPI methodology that lower measured inflation would act as a consistent reduction in the inflation accretion into TIPS principal.
If a change in CPI methodology occurred that lowered CPI that reduction in inflation accrual would act almost exactly like an increase in the bond’s real yield. Long-dated TIPS (by virtue of their higher duration, remember a downward revision to CPI methodology is like an upward shock to the bond’s real yield) would be more sensitive to a change that might reduce cumulative CPI compensation – so one would expect their real yields to rise more.
Observed inflation breakevens (Nominal Treasury yield minus TIPS real yield) could decline because TIPS real yields rise – and could overstate to the Fed and other market observers bond market confidence about future inflation coming back to the Fed’s target. You can see that long-dated breakevens declined even as the FOMC did not tighten policy at the June and July meetings (red lines) and Chair Warsh also did not explain how he planned to navigate above target inflation.
While no downward revision to US CPI methodology has happened, a “methodology revision” premium or compensation for “Boskin Commission 2.0 risk” may be starting to get priced in and could be a factor contributing to long-dated TIPS real yields rising even as breakeven inflation measures appear relatively contained/not worrisome.
Ok — your bank probably doesn’t own or trade TIPS — why do you care?
We all still care about developments in TIPS because they influence long-dated nominal Treasury yields. The yield on long-dated nominal Treasuries, in turn, impact US mortgage rates, influence municipal government borrowing costs and set the discount factor for all other assets in the US economy (think CRE, home prices, US equities, etc).
The rise in long-dated TIPS yields also creates some worry for the AI bubble when I look at this chart which compares the 30-year TIPS real yield (green) with the S&P-500 current earnings yield (white). It appears that the rise in long-dated real interest rates may have contributed the end of the dot.com expansion and that same mechanism could begin to weigh on the AI expansion very soon (green > white).
The birth of the TIPS market in the late 1990s followed on the heels of the Boskin Commission changing CPI. To simultaneously communicate that US inflation measurement may be reconsidered given the history of CPI recalibrations and that inflation will come back to target without a well-articulated plan for monetary policy may embed a risk premia in long-dated TIPS and drag nominal Treasury yields higher. I believe that a pricing in of this risk premia is happening and the associated rise in long-dated Treasury yields could catalyze the end of the cycle.
Dollar/Yen and Leveraged AI Bets Groaned Into July FOMC
Perhaps most telling is what followed almost immediately after no Fed rate hike post July FOMC but long-dated yields rose — the joint US-MOF intervention to support the yen and the near collapse of the $45 billion Situational Awareness hedge fund. Both were clearly brewing in the background and both are developments Warsh should have been aware of in the run-up to the meeting.
The Situational Awareness firesale of assets to Citadel and steep losses, together with news this weekend about Jane Street’s $15 billion trading loss, paints a picture of high leverage and concentration around AI credit that should give FOMC members serious pause about the comfortable notion that monetary policy is “modestly restrictive.”
If monetary policy were genuinely restrictive, one would not expect to see this degree of leverage, concentration and speculative excess in the US financial system. These developments demonstrate a point that I raised last month: US financial conditions remain overly accommodative/loose.
There is a second-order consequence for bank treasurers to consider here. Significant trading losses and deleveraging episodes like these do not pass without a response from the largest global banks that finance hedge funds. As we head into fall, it is plausible to anticipate an organic tightening of large banks’ prime brokerage and counterparty risk limits.
That tightening can arrive well ahead of, and independent of, any move by the Warsh Fed, and one that could recreate the deleveraging dynamics that made Situational Awareness a problem in the first place.
Jackson Hole: Don’t Expect Clarity on an Inflation Strategy
With regards to Jackson Hole on August 28th, as of now I look for Warsh to update on the five Fed taskforces and to speak about payments which is this year’s Jackson Hole conference theme — while leaving the bond market with little guidance about the Sept FOMC meeting. That is a meaningful shift from my July expectation of a Sept hike as the base case, and it reflects Warsh’s casual communication style on display in July.
Note that I am absolutely thrilled to be wrong and see Warsh recover momentum as Fed Chair through either significant progress on disinflation (there are both one more PCE and CPI prints before Sept FOMC) or a well-crafted speech that conveys how he intends to bring inflation down/his reaction function.
Conflicts Keeps Commodity Prices and Inflation Elevated
Another notable feature of the July FOMC related to Iran’s decision to launch unilateral strikes on the eve of the Fed decision, pushing up oil prices. For now, the US strategy appears to be to double down on sanctions and apply economic pressure on Iran given challenges in asymmetric warfare.
As the US mid-term elections approach, it seems plausible to anticipate that hardliners in Iran – who have gained power domestically due to their ability to project military power in the Strait and region — may aim to put upward pressure on commodity prices and inflation to pressure financial markets.
The conflict’s scope in the Strait appears poised to continue to expand. Multiple ships were attacked over the weekend in the Strait and Israel launched strikes in southern Lebanon. Combined with mounting hostilities in the Red Sea and Russia-Ukraine disruptions to energy and agricultural exports, the global commodity shock continues to build with no obvious relief on the horizon. I anticipate these trends increase price pressures on goods which had diminished over the summer.
With an ever-growing Treasury market, more mark-to-market-sensitive and leveraged investors like hedge funds make up an increasing share of Treasury holders. Such investors are far more sensitive to commodity price developments — which can be expected to continue to trend higher — and accordingly are active with curve steepener trades. This is a self-reinforcing dynamic: pushes up long-dated yields, mark-to-market investors respond with steepeners, and a Fed pinning the front end while offering no credible guidance leaves the back end exposed to exactly these flows.
During my recent trip to NYC, the topic of long-dated Treasury yields breaching new highs (6% on 10-year yield) in the months ahead came up. Yes, a steeper Treasury curve is good for banks – but only to a point. Once rates rise enough, embedded optionality and risks in banks’ loans and deposits begin to kick in which can make for more challenging bank asset liability management.
Outlining the Contours of an End-of-Cycle Scenario
Let me now do what the title of this column promises and sketch how an end-of-cycle scenario could plausibly unfold from here. Calling the timing of a credit cycle turn is notoriously difficult. But the value in mapping the contours is not precision; it is in thinking about channels. Here is how the pieces of such a scenario could fit together.
Stage one — widening geopolitical shocks and super El Nino result in further increases in commodity prices and reaccelerate goods inflation.
Stage two — other central banks tighten policy and the long end of the Treasury curve and tighter US financial conditions do the Fed’s work for it, but in a disorderly way. The rise in KOSPI-NASDAQ correlation, the Bank of Korea’s move to lean against speculative excess and likely tightening of prime brokerage in light of signs of excess leverage and poorly controlled trading risks are early warning signs. Leverage in the global financial system begins to unwind.
Stage three — the AI investment narrative cracks. Extended US equity valuations rest on low long-term interest rates; further repricing undermines them, echoing 1999-2000. Cheaper Chinese competition (Moonshot AI’s Kimi K3) and a crowded AI industry in need of consolidation mean the equity wealth effect that has sustained upper-income US consumer spending goes into reverse.
A macro signal that US AI credit growth is vastly underaccounted for is the Basel Committee’s credit to GDP gap. As shown below, the US metric indicates no worry of an AI corporate credit bubble resulting in excess credit growth even as this same indicator clearly flagged one ahead of the 1999-2000 dot.com bust and 2008 crisis.
By contrast, Stijn Van Nieuwerburgh at Columbia University estimates that US AI cap-ex easily tops cap-ex as a share of GDP compared both to the dot.com and the US railroad booms. That US AI cap-ex > the US railroad boom should definitely raise eyebrows. Recall that the US railroad bust in the early 1870s was the largest default cycle in US corporate bond history.
Indeed, how can anyone look at this convoluted circular picture of AI credit that Bloomberg staff put together with confidence that they understand the full liability profile of these firms, including guarantees, securitization, leases and off-take agreements?
Stage four – private credit’s origination of AI credit to distribute to affiliated life insurers and credit funds unravels. Holders of annuities begin to run on life insurers.
Stage five — recession risk, currently underpriced, reprices violently. The yen strengthens materially as Japanese investors in US AI return repatriate capital. According to Bank of International Settlements data, as of Q1 2026 there is $2.3 trillion USD equivalent in yen funding outstanding globally, most of which flows to non-bank financial institutions mostly domiciled in the US and Cayman Islands.
Each of the prior stages compounds the next: the commodity/inflation shock hits, disorderly rise in long-dated yields and financial-conditions tightening accelerate deleveraging, the AI unwind removes the wealth-effect cushion and private credit/insurance sector risks come into focus. This is the end of cycle risk flagged in July. Credit spreads, which have not been pricing this, could gap wider. A withdrawal of foreign capital from US markets could be anticipated to weaken the dollar.
This outline of cycle end is not a certainty. The Strait could see a Chinese-brokered de-escalation, more US fiscal stimulus to try to keep the party going, or the Fed could surprise with a credible reaction function/plan at Jackson Hole that stabilized long-dated Treasury yields. But these five stages are how we could get to an end-of-cycle setup rather than a mid-cycle wobble, and the window for banks to position for these risks may be short.
What This Means for Bank Treasury Teams
The overarching message for bank treasurers remains largely unchanged, and, if anything, the July FOMC reinforced it. A late Fed, a widening global commodity shock, a steepening Treasury curve, other central banks preparing to tighten and the leverage and concentration laid bare by the Situational Awareness and Jane Street episodes all point in the same direction: the risks are compounding.
My updated views on bank treasury are:
- Recall that the weakest loans are generally made during booms and often at the top of the cycle. By contrast, best loans often are made at the trough of the cycle.
- Continue to stress test for a steeper curve driven by long-end repricing, alongside your scenarios for a second phase of the Strait’s closure with curve flattening or inversion if recession risks rise. The point is to be ready for both — the near-term steepening impulse from oil and the potential flattening/inversion if the AI bubble pops and recession risk crystallize.
- Growing loans without commiserate deposit growth or stable term funding is risky in light of possible further shocks.
- Strengthen term on-balance-sheet liquidity. The joint US-MOF yen intervention is a reminder that further commodity price spikes can produce dollar shortfalls in the global financial system. The Fed also recently announced a pause on its reserve management purchases until mid-September (going skinny on the Fed’s balance sheet could be ill-timed).
- Banks with prime brokerage operations should proactively review their counterparty and prime brokerage limits rather than wait for a forced repricing in the wake of further hedge fund stress.
- Keep ALCO meetings monthly and maintain an integrated view across liquidity, funding, profitability, and credit.
- With recession risk still underpriced, continue to evaluate reducing higher-risk credit exposures through asset sales or securitizations, and assess the CECL implications of a rising recession probability.
My sincere thanks to Treasury Talk readers. I look forward to the start of the fall semester at Texas A&M and also speaking at the Concentrate CRE US banking conference September 14-16 – I hope to see a few readers there!
Jill Cetina, CFA is Executive Professor of Finance at the Mays Business School, Texas A&M University, BTRM North America director and a former Associate Managing Director at Moody’s and former Federal Reserve officer. The views expressed here are her own.
This piece was completed on August 16. Readers should treat market pricing referenced throughout as indicative of the environment at time of writing.
