The Daily Spark

While 91% of Japanese and 88% of American households have air conditioning, just 4% of UK homes do, see chart below.

The 60/40 portfolio is broken because equity returns are driven by AI concentration rather than the business cycle, while bond returns are now driven by fiscal constraints rather than cycle dynamics.

With the AI trade slowing down and government debt projected to reach 175% of GDP (see chart below), neither the 60 nor the 40 responds to what made it work in the first place.

The bottom line is that the 60/40 portfolio has lost its diversification benefit, with fundamental implications for asset allocation.

The real risk emerges if the AI trade reverses or markets become more worried about government deficits. In either scenario, both stocks and bonds would face pressure simultaneously, leaving investors with no hedge.

For decades, the carry trade dominated USD/JPY, as investors borrowed cheaply in yen to buy higher-yielding dollar assets, and the currency moved in lockstep with the US-Japan interest rate differential, see the first chart below.

That link broke down after Liberation Day in April 2025, when trade wars unleashed the kind of volatility that makes carry trades dangerous, since the strategy earns a slow, steady yield that a single sharp move in the yen can wipe out, prompting investors to unwind their positions regardless of the still-wide yield gap, see the second chart below.

With the carry trade's pull now diminished, the currency has taken its cue not from the yield math but from Japan's deteriorating fiscal outlook.

Alongside this currency shift, a deeper transformation is underway in Japanese equities, where corporate governance reform has driven a record rise in shareholder activism and pulled foreign ownership to around a third of the market, see the third and fourth charts below.

The bottom line is that the yen carry trade has broken down, and the yen is no longer a rates story. Until volatility subsides, it will trade on Japan's fiscal outlook rather than the interest rate gap.

For more discussion, see our chart book available here.

The chart below looks at what Fed chair Kevin Warsh said during the press conference minute-by-minute:

1. There was no specific moment during the press conference where long rates jumped higher.
2. There were several instances when Warsh said something hawkish where rates began to go down.
3. Rates went up more after the press conference was over than during the press conference.

This pattern reveals something important about what's driving the market reaction to Fedspeak, including the steepening of the yield curve.

One interpretation of what happened is that markets understand the Fed’s commitment to 2% inflation, but with no forward guidance, the market does not understand how the Fed will get to 2% inflation.

Is the way to 2% inflation through higher rates, a smaller balance sheet or tighter financial conditions? The answer to this has significant implications for the yield curve and how and when we will achieve 2% inflation. The lack of clarity about how to get there is what's pushing yields higher because there is now a risk that it may take longer or involve a policy mistake.

Imagine saying: “I will take you from New York to Los Angeles, but I'm not telling you how quickly, what it will cost or how you'll get there." By not explaining these important elements, you may question if we are actually getting to Los Angeles.

The risk with abandoning forward guidance is a steeper yield curve with investors asking more questions about the journey ahead, which is what we have seen since the statement came out.

The bottom line is that an important part of the Fed’s credibility is not just to say that it has certain goals but also to explain how it will achieve those goals.

Why are long rates at the highest level in 20 years?

1) Inflation
2) Fiscal problems
3) Hyperscaler issuance

The bottom line is that we're not going back to the 2010s, and this is good news for everyone cutting coupons in high-quality fixed income.

Analysis of actual Claude usage data shows workers in AI-exposed occupations are experiencing slower wage growth, while employment levels in these occupations remain unchanged, suggesting companies are capturing AI productivity gains through wage compression rather than workforce reduction.

This paper was written by Sania Edlich and me using a difference-in-differences methodology with occupation and year fixed effects across 321 matched occupations from 2015 to today. The paper is available here.

CDS spreads have started to widen out for names in AI, and the market is asking three fundamental questions:

1) Will the AI capex pay off, and how quickly? With trillions committed to data centers, chips and power up front, the question is whether AI monetization ramps fast enough to clear the cost of capital before the assets depreciate, or whether it's an overbuild whose ROIC never catches up to its WACC on a massive, front-loaded outlay. For more, see also here.

2) How is all of this being financed, and at what spread? Hyperscaler spreads are widening as the buildout is increasingly funded with debt rather than organic free cash flow, and as issuance surges, the question is whether the all-in yield climbs to a level where the marginal data-center dollar no longer clears its return hurdle, forcing the capex cycle to self-throttle.

3) Will there be unlimited demand for compute, or will compute demand peak? The bull case assumes demand is effectively insatiable as inference workloads, agentic systems and new model generations compound, but the risk is that efficiency gains, model commoditization or slower-than-expected enterprise adoption cause demand to plateau well below the capacity now being built, leaving the industry with a glut of expensive, rapidly depreciating infrastructure.

Over the past month, the consensus has revised up operating cash flow expectations for hyperscalers over the next five years, see the first chart.

However, capex expectations have been revised up even more sharply, see the second chart.

As a result, the consensus now expects free cash flow to be materially lower over the same period, see the third chart.

The bottom line is that the consensus has over the past month meaningfully increased its near-term hyperscaler capex forecast, reducing expected free cash flow despite stronger operating cash flow forecasts.

For investors, the key question remains the speed with which these capex investments will translate into proportional revenue or profit growth, see also here.

US enterprises survive longer than their global peers due to a combination of easier access to capital, market size, flexible labor markets that let companies adjust costs quickly and lighter regulation that enables faster growth and faster pivots, see chart below.

The recent performance of US software indices reveals a stark divergence in market expectations regarding the disruptive threat of AI, see chart below.

High-level groupings seen as AI-immune, such as security software and infrastructure software, have significantly outperformed the broad benchmark because they provide the critical data backbones, compute pipelines and complex protection systems that generative models actively expand rather than displace.

Conversely, categories deemed AI-vulnerable, specifically application software and vertical software, are severely lagging behind as investors price in potential disintermediation, lower switching costs and reduced seat-based licensing demand driven by automated AI workflows.

The bottom line is that the market is clearly favoring foundational security and infrastructure layers over legacy applications as artificial intelligence reshapes software business models.