Oh, What a Tangled Web We Weave…

I have been spending a good portion of yesterday afternoon and this morning fielding questions about the effect that CoreWeave (CRWV) and its bonds are having on broader markets.  In more normal times, credit concerns at a single corporate borrower should not resonate dramatically throughout global equity and fixed income markets, but these are not normal times.  The artificial intelligence buildout has become such a key factor not only in markets, but in the global economy writ large, that tremors in a relatively marginal player can resonate wildly.

The basic scenario is this: the yield on corporate bonds issued by CRWV, a key AI cloud stock, jumped as high as 15% yesterday, indicating market concerns about the company’s creditworthiness.  To be clear, I am not a corporate bond analyst, nor do I have an independent view on CRWV’s ability to repay its creditors in a timely manner.  But as a long-time market participant and observer, it is quite fair to say that a bond yielding 15% while Treasuries of similar duration are yielding around 5% reflects some genuine nervousness among its owners.

To be precise: according to TRACE data, the CRWV 9.625% notes maturing on July 15^th, 2032 are yielding 15.568% as I type this, while the 5-year Treasury is yielding 4.85%.  Moody’s rates these bonds B1, though it appears that the market views that as a bit sanguine now.

I believe that the CRWV situation has rejiggered institutional investors’ priorities (though not so much for retail, where IBKR customer activity skewed heavily toward dip-buying in tech).   As I summed it up to someone yesterday:

Bad CPI -> big bond selloff -> CRWV troubles -> tech selloff (along with everything else) -> bond rally

This morning, there is no obvious reason for bond prices to be screaming higher (yields sharply lower) after a worse-than-expected PPI report.  If inflation was the only driver of yields, jumps of 0.7% and 0.5% in September headline and Core PPI, respectively (when 0.3% was expected for both), we would not see bond yields fall by 8-10 basis points.  Therefore, there must be something else afoot.  A good portion of the rally in Treasuries is a big bounce from massively oversold conditions and the rest is spurred by a flight to safety.

The change in mindset that began yesterday remains evident this morning, even if it is ebbing in stocks as I type. 

My thesis has been that stocks have been able to shrug off higher interest rates because the AI stocks that dominate the large indices are willing to pay any price to borrow because the market perceives the ROI on AI to be essentially infinite.  That creates a feedback loop because the marginal borrowers don’t really care about what rate they pay and because big government borrowers have no choice.  Yesterday was a crack in that thesis, because CRWV lenders are now realizing that they might have difficulty getting paid back.  This is like the rising cost of credit default swaps that we’ve previously seen at Oracle (ORCL), though with greater magnitude.  If enough large borrowers start to encounter concerns about their creditworthiness, this could become a big issue that we’re just seeing the start of.  (Hopefully not!)

One example from very early in my career stands out this morning.  In August 1987, shortly after I finished the Salomon Brothers training class, the firm was exceptionally proud of a massive asset allocation trade it had done on behalf of a major pension fund (I think it was the State of Maryland).  They took advantage of the huge run-ups in bond yields and stock prices and moved billions from stocks to bonds.  That may be one of the greatest trades of all time, and it presaged what was to come.  I can’t say the outcome will be remotely the same as 1987 since we’re nowhere near as extreme, but these types of trades have secular consequences if they persist in size.  That said, even then the consequences didn’t manifest themselves until several weeks later.

I noted above that the picture for stocks was changing as I typed, and around 11:30 ET we now find the S&P 500 (SPX) slightly in positive territory and the Nasdaq 100 (NDX) well off its lows, though still down more than 0.5%.  Even in the pre-market, most non-tech SPX sector ETFs were higher, and we now see 7 of 11 sectors on the plus side.  Tech, of course, is not one of them.  Furthermore, NYSE advancers solidly outpace decliners, though the margin is less than 2:1. Thus we have switched away from outright selling to rotation, but considering how overweight tech is in SPX, it takes an awful lot of rotation to overcome tech weakness.

[By the way, the quote that we paraphrased is by Sir Walter Scott, not Shakespeare, as is commonly believed (by me, among others).  The full sentence, from the 1808 poem Marmion: A Tale of Flodden Field, “Oh, what a tangled web we weave / When first we practise [sic] to deceive!”  Hopefully there is no outright deception occurring here, other than perhaps our deceiving ourselves that this new technology might offer more boundless returns than any prior life-changing development.]