Markets Revisited

Three months ago, I wrote an article suggesting we were close to a major market top.  I made two points:  the historical experience of low or negative market returns from such an extreme level of valuations, and expectations of a different and less market-friendly monetary policy under Chairman Warsh.

With hindsight my timing was clearly wrong, as markets promptly rose by 10% to new highs.  The first-quarter earnings season was much stronger than expected, which helped to justify the valuation levels.  I could say that the facts changed, but that doesn’t alter the fact that my call was early.

SpaceX and why investors bought it

The launch of SpaceX has brought into focus another reason for why investors are willing to embrace such high valuations.  This article examines it, and updates what I wrote in March.  Spoiler alert: I haven’t changed my view that we are close to a market top.

SpaceX, despite its name, is really about AI.  Make no mistake, AI is for real, will change our lives, and will create huge opportunities for some firms.  But it is equally true that the aggregate financial investment going into this area is of such scale that it cannot realistically all be recouped.  In other words, there will be winners and losers – big time.

Retail investors are prepared to pay over 100x earnings for SpaceX because they hope it will be a winner.  The scale of growth for AI-related companies that succeed is such that the price paid to obtain exposure to the opportunity today becomes a secondary consideration. 

I am not suggesting that investors are right to back SpaceX – far from it, as we don’t know today whether or not it will be a winner in the future.  But that is why they countenance valuations which look ludicrous to value-oriented professionals like me.

Paying up for the AI winners

If you expand this thinking across the mega-tech universe, it helps to explain the divergence between it and the rest of the market in recent years.  Investors of all sorts have paid up to obtain exposure to the companies they hope will be winners without paying too much attention to the price.  So, in some ways the argument I made earlier on about the first-quarter earnings season isn’t really relevant either, though it may add to shareholders’ confidence in the bets they have placed.

(Here I would note how much any quarterly earnings figure relies on accounting accruals and assumptions.  Financial training has taught me to focus on audited balance sheets and the change from year to year, as any interim profit and loss report may be significantly restated). 

What about the AI losers?

But we also need to look at the other side of the coin: there will be losers too, and they should bring the equity indices down.  I say ‘should’ because delving into the past offers examples where companies which have imploded have not been removed from the index at zero.  How many people recall Enron* or Polly Peck?

Increasing levels of bankruptcies in the US

I came across an interesting statistic this week: the number of large company bankruptcy filings in the U.S. in 2025 was 778 (source: S&P Global Market Intelligence) – the largest since the aftermath of the Global Financial Crisis in 2010.  2026 is shaping up to be the second largest.

These failed companies are largely spread across sectors which borrowed aggressively when financing costs were low, such as industrial, consumer discretionary, and healthcare.  Refinancing at much higher rates has proved to be a step too far for them.

I draw several inferences from this.  It is further evidence of the fragility of the U.S. financial system today, something I regularly draw readers’ attention to.  Higher bond yields leading to increasing numbers of companies running into refinancing or cashflow problems should come as no surprise to anyone. 

It is how markets used to work before the cost of money fell to zero in the wake of the Global Financial Crisis.  It is how they will work in the future.  But it may be less familiar to readers who weren’t involved before about 2003.  The danger is the ‘domino’ effect that one company’s failure can have on its creditors and suppliers.

AI companies are not immune either

Secondly, and coming back to my first theme, the companies investing in AI are borrowing too.  For example, SpaceX has launched a USD20bn bond issue.  They are doing so at higher rates, so the refinancing risk is less.  But the risk of cashflow problems for those who don’t end up as winners is greater because of the higher debt service costs. 

How will Warsh and Bessent keep the cost of financing down?

The third inference I draw from this is that the authorities will want to find ways to keep the cost of finance for corporates (and voters) down.  They have some well used weapons to deploy.  The Federal Reserve can reduce interest rates, the traditional response to a downturn prior to about 2003.  It can expand its balance sheet and increase the supply of money and credit in the system as it did between 2009 and 2020 i.e., Quantitative Easing.  Or the Treasury can tilt its financing towards more short-term instruments, restricting the supply of Treasuries and thereby keeping the cost of longer-term financing down.  This has been the lever of choice since 2023.

What will Warsh do?  The first option is looking difficult optically with inflation so high; for example, his first meeting as Chair of the Federal Reserve last week removed the ‘easing’ bias.  He has also publicly stated his commitment to rein in inflation.  The second is probably also off the table, as he has publicly stated his desire to reduce the size of the Federal Reserve’s balance sheet.  I therefore expect the focus to be mainly on the Treasury using short term instruments to finance its borrowing i.e., a continuation of current policy.

The problem is inflation

The big problem for Warsh and Bessent here is inflation and its implications for bond yields: the latest figure is 4.1%, double the 2% target.  Bond markets may take the view that higher redemption yields are needed to compensate for the increased risk here.  And that would in turn increase the refinancing problems for stretched companies – and indeed the U.S. Government itself.  He may find that whatever policy he chooses fails to lead to the desired outcome.

Coming back to the equity markets, this is why I still expect a market top soon.  The combination of higher long-term bond yields, at least in the short term, growing numbers of companies running into refinancing difficulties and an increasing realisation that the U.S. Federal Reserve is facing an unsolvable conundrum will finally find its way to equity markets.

See, I’ve hardly mentioned valuations in this article!  But they are still high.

*Fun fact:  Enron’s replacement in the S&P index in 2001 was………NVIDIA.