Q2 2025: “Terminal Phase Whipsaw”
The period began with acute market weakness following April 2nd “Liberation Day” tariff pronouncements. The S&P500 suffered a two-session 10.5% plunge, the worst sell-off since March 2020. The Banks sank 15.8%, and the Broker/Dealers lost 13.3% in two days. It was virtual credit market panic. Indicative of a major tightening of conditions, risk premiums spiked. High-yield spreads to Treasuries widened the most since March 2020, with high-yield CDS prices spiking the most since June 2020. The junk bond market seized up, as Bloomberg referred to “one of the darkest weeks for US leveraged loans this decade.”
At the time, I thought it likely that the bubble had been pierced - that an aggressive tariff regime, prospects for destabilizing trade wars, along with other administration policies, were a likely catalyst to end a protracted period of “terminal phase excess.” For a few tumultuous days, markets feared faith in the “Trump put” had been misplaced. Responding to intensifying market instability, the President announced a 90-day tariff pause. The S&P500 posted a stunning 9.5% one-day advance, while the Semiconductors spiked 18.7%, and the Broker/Dealers jumped 10.3%. The Goldman Sachs Short Index surged 12.5%.
Fear abruptly returned to greed. The “Trump put” was real and could be counted on as a predictable compliment to the “Fed put.” T.A.C.O. – “Trump Always Chickens Out” – took markets by storm. Bloomberg ran the headline, “Retail Investors Who’ve Only Known Bull Markets Are Buying the Dip.” Reporting on the retail dip buying phenomenon, the Financial Times quoted a trading firm executive: “Pops and drops will occur… but the dip-buying belief has become the new religion.”
The past few months have been telling: In the throes of terminal phase excess, it’s a thin line between bubble deflation and invigorated bubble inflation.
In a sign of the times, my search for short-only hedge fund performance came up empty. The strategy has vanished from industry performance tabulations. Here’s a data point underscoring the recent challenge on the short side: From the April 9th intraday low to last Thursday’s intraday high, the Goldman Short Index rallied 71%. That which does not destroy a bubble only makes it stronger. Yet another brutal short squeeze – another example of why we have avoided shorting individual company stocks.
Last week, the meme stock phenomenon returned with a vengeance. Krispy Kreme, Opendoor, GoPro, Rocket Mortgage, Kohl’s – and a bunch of other stocks with suspect fundamentals enjoyed ridiculous gains. And there’s nothing like a big short squeeze to feed speculative impulses, with the major indices running to all-time highs. Renewed meme stock attention elicited interesting analysis. Professor Peter Atwater, who studies retail investors, made an important point. He said, “We’ve normalized memeing. There’s a yawn to it now.” “The most aggressive traders have already moved on to riskier frontiers – digital tokens, leveraged ETFs, prediction markets.”
Additional insight from Bloomberg: “Customary warnings about speculative excess fell on deaf ears. What once felt seismic now feels like a normal part of daily trading — another episode in a US financial system where bursts of retail speculation are routine, expected, and largely unremarkable.”
We are witnessing important late-cycle dynamics. Bubble “terminal phase excess” is characterized by a wholesale disregard for risk. After all, over a long cycle, it is the aggressive risk takers – in the markets, throughout finance, within corporate America, and all about the economy – that have the most to show for their efforts. They have come to confidently dominate positions of power and influence. If you’re not a risk-taker, you’re a loser. These days, risk aversion has been virtually eradicated.
I have delved into this analysis in previous calls. There is nothing comparable to today’s “terminal phase excess” other than the late-1920s bubble blow-off. And it seems rather obvious, at least to me, that today’s excesses greatly exceed anything from the fateful “Roaring Twenties” culmination.
When analyzing bubbles, there’s always underlying sources of credit fuel to identify. Today, key sources of exorbitant credit expansion are readily apparent. In a defining feature of the current “global government finance bubble” cycle, U.S. system credit growth is dominated by an ongoing historic expansion of federal debt – perceived money-like credit instruments that essentially enjoy insatiable demand. The federal deficit is expected to again approach $2.0 TN this year, or 6.5% of GDP. Importantly, ongoing massive deficit spending supports system incomes, business earnings, asset prices, and economic activity generally. It’s also worth noting that state and local bond issuance is on pace to surpass last year’s record $500 billion.
Corporate debt issuance is on near-record pace, with forecasts for $1.4 TN of investment-grade bond sales – double-digits ahead of 2024 growth. Despite the April shutdown, high yield issuance is approaching $200 billion, ahead of last year’s pace.
Meanwhile, there are consequential sources of system credit that fly under the radar of conventional analysis. It’s worth underscoring the ongoing historic expansion of speculative credit – the primary source of destabilizing market liquidity overabundance.
“Repos” – repurchase agreements – are the key source of financing for levered holdings of Treasuries, agency debt, and MBS securities. It is the principal mechanism for financing today’s massive Wall Street and hedge fund securities portfolios. Estimates have the repo market funding over $1 TN of highly levered Treasury “basis trades” – one of many levered so-called “carry trades” that now permeate the entire credit system.
Bear with me as I highlight some important detail. From Q1’s Z.1 report, we know Broker/Dealer Assets expanded an incredible $490 billion, or 37% annualized, to $5.76 TN - the strongest quarterly growth since Q1 2007.
How did Wall Street finance this remarkable balance sheet expansion? Repo Liabilities ballooned $361 billion, or 62% annualized, during Q1 to $2.7 TN – the high since Q3 2008. Repo Liabilities inflated about $1.1 TN, or 67%, over 10 quarters. Total system Repo Assets surged $717 billion, or 41% annualized, during Q1 to a record $7.78 TN. One-year growth of $1.1 TN compares to peak mortgage finance Bubble 2007’s $655 billion expansion. System Repo Assets ballooned $2.97 TN, or 62%, over 21 quarters.
Q2 Z.1 data won’t be available until September, but we do have recently reported Q2 bank and broker data that suggest an unrelenting historic expansion of speculative finance.
At JPMorgan, Total Assets surged $195 billion, or 18% annualized, to a record $4.55 TN during the quarter. The asset “Fed Funds Sold & Repos” jumped $41 billion, or 38% annualized – with first-half growth of $176 billion, or 120% annualized. On the liability side, “Short-Term Borrowings and Repos” surged $86 billion, or 46% annualized.
Bank of America posted Total Asset growth of $92 billion, or 11% annualized, to a record $3.44 TN. The asset “Fed Funds Sold & Repos” posted growth of $24 billion, or 29% annualized. “Short-Term Borrowings & Repos” surged $32 billion, or 24% annualized.
Citigroup Total Assets expanded $51 billion, or 8% annualized. On the liability side, “Short-Term Borrowings & Repos” expanded $28 billion, or 18% annualized, during the quarter. Balance sheet detail is not yet available, but exceptionally strong earnings reports from Goldman Sachs and Morgan Stanley were indicative of overheated securities finance.
It goes unrecognized that the expansion of “repo” borrowings generates additional marketplace liquidity. As a proxy for this new age monetary onslaught, look to money market funds – the holders of trillions of repos. Money funds expanded $921 billion, or 15%, y-o-y – having recently surpassed $7 TN ($7.075 TN) for the first time. It is one of history’s most consequential - yet somehow disregarded - monetary inflations: money fund assets have inflated $2.5 TN, over 50%, in just the past 32 months.
I’ll highlight another key source of system credit, one that suffers from a notable lack of transparency. It’s a sector that has become an instrumental funding source, while demonstrating distinct bubble dynamics. I’m referring to the “private credit” boom. It makes no appearance in quarterly Z.1 reports, and it’s not even clear the Fed includes private credit in system credit growth and financial flow data.
I have issues with “private credit” – and serious concerns as this credit expands exponentially in “blow-off” fashion. For one, it is chiefly subprime corporate credit, focusing on lower-rated entities willing to borrow at high yields. As I’ve discussed in the past, no lending business enjoys near-term accounting profit potential like lending aggressively to high-risk borrowers.
This space is inherently prone to bubble dynamics. So long as financial conditions remain loose and credit readily available, most negative cash-flow enterprises will remain current on their obligations. They simply borrow to satisfy ongoing cash-flow and debt service requirements. But this structure suffers inherent fragilities and credit cycle vulnerabilities. When finance inevitably tightens and borrowers lose access to new borrowings, credit problems pile up quickly. The prospect of losses scares lenders, who back away from the marketplace – leading to the painful credit cycle downside. The mortgage finance bubble illuminated subprime lending’s propensity for boom-and-bust dynamics.
Private credit was at the cusp in April. Leveraged loan prices were under major pressure, as high-risk lending markets seized up. It’s no coincidence that AI and technology stocks were in freefall. The MAG7 Index was clobbered 12% in just two sessions. From the April 2nd close to April 7th intraday lows, NVIDIA collapsed almost 19%, and the Semiconductor Index sank 17%. Private credit has become an indispensable funding source for the historic AI investment boom.
Again, we can’t overstate private credit’s significance. I strongly argue that this high-risk lending boom has become fundamental to terminal phase excess. And there are critical idiosyncratic characteristics to note. Most of the high-risk loans in this sector are neither priced in the marketplace nor traded. Risk-embracing speculators and investors are attracted to high yielding products that operate outside the vagaries of volatile markets. In April, with junk bond and leveraged loan prices under intense pressure, private credit investors did not face mark-to-market portfolio losses.
These days, private credit risk expands exponentially, with overheated late-cycle growth of increasingly risky loans. Insurance companies have leaned on private credit for the high yielding instruments underpinning booming annuities businesses. Now, there’s a push to open private credit access directly to retail investors – through closed-end funds, ETFs, and 401k investment vehicles.
In my nomenclature, private credit demonstrates powerful “inflationary biases”. This bubble is inflating rapidly, in a self-reinforcing speculative dynamic. Recovering swiftly after Trump’s tariff pause, leveraged lending and private credit booms have only gained momentum. Here’s a crazy number. The past week’s $120 billion pushed July leverage loans launches to a monthly record $218 billion. The majority were refinancings, but it’s still indicative of overheated demand for high-risk credit. It’s worth adding that a booming junk bond market powered the strongest July issuance since 2021.
Again not coincidently, as leveraged loans, junk bonds and private credit booms power ahead stronger than ever, Nvidia, Microsoft, and Meta Systems, along with the major technology indices, recovered to new all-time highs. An already historic AI arms race has further intensified.
Case in point is Mark Zuckerberg and Meta. Meta is currently seeking $29 billion of private funding for AI data centers – with plans to spend over $70 billion on data center buildouts this year. Zuckerberg enticed a leading AI executive from Apple with a $200 million compensation package. A Wall Street analyst stated that Meta could spend $20 billion on total compensation for the quarter to build out its AI team. The company took at 49% stake in startup Scale AI for almost $15 billion.
Microsoft, Meta, Google, and Amazon are projected to spend a staggering $320 billion this year on cap-ex. Incredible AI-related growth is expected to continue for years. Goldman Sachs forecasts AI investment will surpass $1 TN over the next few years. And that excludes the massive investment expenditures necessary to ensure sufficient electricity to power hundreds of colossal data centers. Training large language models consumes enormous amounts of power, while the standard ChatGPT query is said to consume more than 10-times that of a Google search.
According to Goldman, roughly $720 billion will need to be spent on grid upgrades over the next five years. JPMorgan refers to a $7 TN race to scale AI data centers, with power infrastructure requiring up to $500 billion. The International Energy Agency (IEA) projects electricity demand from data centers worldwide is set to more than double by 2030. Just yesterday, I read of a planned data center in Cheyenne that will consume more power than all the households in Wyoming.
A deeply ingrained perception of endless liquidity and credit availability is a defining feature of “terminal phases.” Speculative credit is self-reinforcing. When a hedge fund borrows in the “repo” market to purchase Treasuries or other government securities, this additional liquidity creation is unleashed into the marketplace – liquidity that fuels inflating asset prices and only more credit-expanding levered speculation.
The current terminal phase is uniquely dangerous. On the one hand, unprecedented leverage in perceived money-like Treasuries has created trillions of marketplace liquidity – feeding myriad speculative bubbles. On the other hand, the AI arms race is a historic borrowing and spending blackhole. Thus far, the two have dovetailed with phenomenal results. So long as growth in leveraged speculation is uninterrupted, bullish perceptions of endless money to finance the AI buildout will remain ironclad. But this brotherhood between the bubble in leveraged speculation and the AI mania is replete with latent fragilities. Disregarded of course, but April’s market convulsion forewarned that deleveraging and bursting speculative bubbles come with dire consequences for the AI arms race.
April provided another critical warning. Extraordinary vulnerability lurks below the Treasury market’s surface – fragility that rapidly reverberated across markets, especially in the high-flying and over-owned technology sector.
Treasury yields spiked 50 bps during the week of April 11th - an extraordinary development. Global markets were hit with intense instability, yet the standard safe haven demand for Treasuries and the dollar was completely missing in action. Market talk was of hedge fund “basis trade” unwinds, along with fears of China Treasury liquidations.
Stocks were faltering, but it was the “yippy” bond market that forced the President’s tariffs retreat. Stocks, Treasuries, corporate bonds, and derivatives reversed sharply. There are important “terminal phase” dynamics that don’t get the attention they deserve. “Terminal phases” are characterized by an exponential rise in systemic risk. On the stock side, inflating prices detach from deteriorating fundamental prospects – while surging market prices stoke industry investment booms that further inflate corporate earnings. In credit, there’s a surge in the quantity of increasingly unsound loans, as I mentioned with “private credit.” Here you can also think Treasuries, speculative leverage, and subprime leveraged loans.
A historic surge in risk hedging is one manifestation of this massive increase in systemic risk. In the stock market, players are ready to buy put options and derivative hedges that offer protection against a market downturn, while hedge fund operators and others will boost short selling. Bond speculators and investors are ready to use derivatives to protect against Treasury losses, while shorting liquid Treasuries to hedge corporate credit, mortgage securities, munis, and myriad structured products.
This creates latent instability. If many within the marketplace move to offload risk, as they began to do in April, markets quickly turn illiquid and prone to dislocation. However, and this is especially pertinent, these outsized bearish hedges and derivatives create rocket fuel for market reversals and rallies.
The President’s retreat triggered precisely this dynamic – a big whipsaw. Derivative operators that were one day selling stocks and Treasuries - to hedge market protection they had sold - were the next day aggressively covering shorts and buying securities to rebalance their derivatives exposures.
Markets that had turned illiquid on the downside abruptly reversed course and turned liquidity-challenged to the upside. And there’s nothing like a big short squeeze to really get the markets’ speculative impulses humming. And when markets start to run, the great FOMO takes on a life of its own. And this fear of missing out is such a powerful “terminal phase” dynamic. Paraphrasing the great economist Charles Kindleberger, nothing causes more angst than watching your neighbor get rich. Especially late in the cycle, this dynamic applies to the crowds of retail traders, options speculators, hedge fund operators, mutual fund managers, and institutional investors. I’ve discussed the important role short squeezes play. Also critical in speculative dynamics these days is the so-called “gamma squeeze,” where those that have sold call options are forced during rallies to aggressively purchase stocks and ETFs to hedge contracts that have gone “in the money.”
And with virtually everyone having grown so comfortable with risk-taking, leverage and derivatives, FOMO has never toted today’s level of firepower. Market structure has evolved to the point of ensuring wild downside and upside volatility – and we saw this dynamic play out spectacularly with Q2’s “terminal phase whipsaw.”
I haven’t been surprised by market instability and volatility. The analytical case for Treasury and dollar vulnerability has been validated. In general, I subscribe to the view of a powerful late-cycle propensity for crazy markets. Expect the unexpected. I certainly appreciated that market structure created the potential for a market recovery after April’s “risk off.” Where I’ve been wrong is the belief that a vulnerable Treasury market would remain calm in the face of such dramatic financial conditions loosening and resurgent “risk on” speculative bubbles.
Treasuries – bonds globally – are extraordinarily vulnerable. Risks are multifaceted. Let’s start with supply. It’s now massive fiscal deficits as far as the eye can see. The latest projections from the CBO have the big, beautiful bill adding $3.4 TN over the next decade to already egregious deficits. It is now likely that deficits have entered the perilous out-of-control phase. It’s worth noting that 10-year Treasury yields are today about 80 bps higher than when the Fed began its 100 bps rate reduction last September. We saw in April the potential for Treasury yields to rise even as risk markets buckled and economic prospects dimmed. It is today reasonable to contemplate scenarios where annual deficits approach 10% of GDP.
Meanwhile, the AI mania ensures massive and unending borrowing requirements for years to come. There is also the unappreciated dynamic with ever-increasing numbers of negative cash-flow enterprises - and a bubble economy structure defined by enormous high-risk borrowing requirements.
Treasuries are also vulnerable to what I believe are now deeply rooted inflationary pressures throughout the economy. I know Federal Reserve and Wall Street forecasts call for inflation to retreat nicely to the Fed’s 2% target. But it has been above target for over four years. Once having established a foothold, inflation has historically proven uncooperative.
Important factors significantly increase the likelihood of upside inflationary surprises. There are the massive government deficits already mentioned, which are the ballast for ongoing elevated total system credit growth. I also believe that accelerating climate change presents greater inflationary risks than policymakers and analysts are willing to admit. Rising food prices are only the most obvious. Expensive recovery and rebuilding efforts – and I’m thinking here of Hurricane Helene and Milton, along with the devastating LA fires. More recently, there were the catastrophic floods in Texas and elsewhere. According to the National Center for Environmental Information, there were 27 confirmed weather disaster events last year with losses exceeding $1 billion. The 1980–2024 inflation-adjusted annual average is nine. LA fire damage and economic loss estimates exceed $250 billion.
I expect climate-related damages and rebuilding costs to continue to inflate, underpinning prices for materials and many things, including labor. And while conditions are notably uneven, labor markets remain generally tight. With a 4.1% unemployment rate and inklings of tightening, I’m assuming wage pressures will surprise to the upside.
From my analytical perspective, the economy was demonstrating notable inflationary impulses even before the new administration’s policies. There is the push to deport illegal immigrants, which is already causing labor issues in agriculture, construction, and myriad industries. The administration now has the financial resources and is gearing up for millions of deportations. Inflationary consequences will not be insignificant.
Tariffs are in the process of raising prices for so many things. And I don’t buy the wishful thinking that tariffs only create a one-time price impact - and thus are noninflationary. For one, the new tariff regime is hitting in a backdrop of already elevated prices and inflation expectations. Companies will likely space price increases over time, rather than to slam consumers with full tariff increases all at once. Similarly, international exporters will initially eat some of the tariffs. What’s more, the dollar’s 9% y-t-d devaluation will translate into even higher prices for many imports.
The University of Michigan Consumer Survey has one-year inflation expectations at 4.4%. It’s worth noting that the market’s gauge of inflation expectations – the five-year “breakeven rate” – recently traded to 2.54%, the high since April 3rd.
Loosened financial conditions over recent months can be expected to underpin both economic activity and price pressures. I question how long the bond market will disregard these dynamics. I think in terms of the “bond vigilantes” regrouping and plotting their next attack. Three key markets seem to be in the vigilantes’ crosshairs.
It was the UK gilts market back in the fall of 2022 that seemed to mark the end of the vigilantes’ lengthy hiatus. Gilt yields last week traded to 4.67%, the high since May. Japanese 10-year JGB yields closed Friday trading at 1.60%, the high back to pre-Great Financial Crisis July 2008 – as long-bond yields surpass 3%. Japan’s CPI has been running above 3% since last November. Year-over-year food inflation surpassed 7% last month. And results from Sunday’s parliamentary election confirm that the scourge of inflation is delivering harmful effects to Japan’s traditionally cohesive and stable society. The ruling LDP party lost its majority in the upper house, placing it in a minority position in both houses for the first time since 1955. Japan’s rising right-wing nationalist party boosted its seats in the upper house from one to 15. I’ll quote Bloomberg: “Since World War II, Japan has built a reputation as a global safe haven for investors in part due to a consensus-driven political system helmed by one of the world’s most successful big-tent parties. Now the center is unraveling, testing much of what has held true about the nation for decades.”
Approaching 240% of GDP, the sustainability of Japan’s government debt is a serious – likely pressing - issue. Political pressure is mounting to assist those most hurt by inflation. And with a policy rate of only 0.5%, pressure is building on the Bank of Japan to accelerate the pace of policy normalization to counter rising prices. The BOJ faces a critical predicament: rate increases necessary to contain inflation risk triggering a bond market crisis. Literally decades of misguided policies, including near-zero rates and reckless debt monetization, are coming home to roost.
In last week’s CBB, I used the quip, “What happens in Tokyo doesn’t stay in Tokyo.” For many years, Japan has operated as a critical epicenter for global speculative finance. Trillions abandoned zero Japanese rates to partake in higher-yielding instruments – including Treasuries, U.S. corporate credit and structured finance, European periphery bonds, emerging market debt, and beyond. For thirty years, the Japanese financial system has provided virtually costless borrowings to the global leveraged speculating community for the financing of so-called “yen carry trades” all over the world. It’s been one year since the “yen carry” deleveraging near blowup – another market warning too easily dismissed.
The timing of these kinds of things is impossible to predict. But elements are falling into place that raise the odds of global bond market instability. When JGB yields surged in January to a then 14-year high of 1.24%, 10-year Treasury yields spiked to 4.80%. After retreating, yields surged together in May, and then pulled back in unison into early July. Treasury and JGB yields are both higher this month, though Treasury yields have yet to follow JGBs to multi-year highs. I’ll also note recent tight correlations between Treasury bond and UK gilt yields.
There’s a natural ebb and flow – a greed and fear dynamic that cycles through markets. Unparalleled speculation and derivatives trading, along with other market structure issues, have the recent “risk on/risk off” dynamic operating on steroids. The next deleveraging episode is the proverbial “when and not if.” All the craziness materializing since April only exacerbates system fragility.
Financial conditions are too loose – and loose is especially dangerous during this most extended “terminal phase.” Upside inflation surprises are a serious risk. The administration’s focus on the 10-year Treasury yield is by design. Treasury Secretary Scott Bessent operated in the hedge fund universe for over three decades. The deep-pocket hedge fund industry got their man to head the Treasury Department. And Bessent is keenly aware of the huge “basis trade” and speculative leverage more generally. These days, especially post April market instability, the “Trump put” is grounded in faith that Secretary Bessent will sprint to the Oval Office when necessary – and the President will listen. It’s the Secretary that explained ramifications for April’s “yippy” bond market. It is Bessent who is on top of the administration’s arsenal of measures to promote lower longer-term Treasury yields. More recently, it is the Treasury Secretary explaining the market realities of forcing Chair Powell’s early exit.
Curiously, Treasury futures positions did not indicate a meaningful reduction in “basis trade” leverage back in April. And I’ve been surprised how willing the speculator community has been in holding firm with heavy leverage in the face of such elevated uncertainty. I suspect “basis trade” and other levered positions have essentially become too massive to liquidate – key hedge funds too big to fail. This dynamic worked to contain April’s leverage unwind. It also heightens risks of highly disorderly future deleveragings. According to the Financial Times, hedge funds ended 2024 with U.S. government bond holdings at $3.4 TN, having roughly doubled since the beginning of 2023.
Thus far, the bond market – or, more specifically, the hedge fund community – has accommodated the administration’s pro-growth, pro-Bubble policy regime. I see a tenuous codependency – accommodation that sow the seeds for a contentious break-up. From my analytical framework, today’s mix of loose financial conditions, liquidity overabundance, bubbling stocks, booming credit, massive deficit spending, deregulation, and the administration’s investment push should underpin growth. So long as “risk on” and loose conditions hold, inflation risks remain elevated.
In recent quarterly calls, I’ve repeated that I hope my analysis is wrong. At this point, I would welcome developments that refute my view of an ongoing worst-case-scenario. “Terminal phases” are precarious affairs. This is undoubtedly one for the history books. The confluence of late-super cycle speculative excess, unprecedented leveraged speculation, runaway deficits, a high-risk “private credit” lending bubble, a historic AI mania and arms race, a rapidly inflating crypto bubble, and the Trump administration ensures this “terminal phase” ends badly.
So long as conditions remain so loose, I question whether the bond vigilantes will remain in hiding. The hedge funds should be nervous. But for now, we’re in the feel-good summer doldrums. Be prepared for the return of tumultuous markets this fall, if not sooner.