Earnings reports that I’m reading are signaling that, in the US at least, the post-pandemic splurge on big ticket items like vacations is slowing. Whether this is due to people running out of money or the Fed’s restrictive money policy probably makes no difference. Arguably, the collapse of the yen carry trade is not simply the result of rising interest rates in Japan but also the sense that rates in the US are on the cusp of decline. If we assume, as we should, that fixed income is the main investment substitute for equities, we can anticipate two effects on stock prices: the market PE should rise and consensus earnings estimates for companies heavily dependent on US consumer spending should contract.

The result of this will likely be a continuation of the market rotation away from the worst-affected consumer cyclicals–like hotels, restaurants, vacation destinations–and into more stable growth or non-consumer areas.

I’ve been driving around in eastern Pennsylvania doing my usual photography work. I stopped in a Wegmans and bought an old-fashioned Charleston Chew candy bar a couple of weeks ago. The price last year was $1.20. It’s now $2.00–and although the packaging is the same, the bar inside is maybe 20% smaller. Around the same time, I was in a Weis supermarket (a wise choice, that company says). I bought a box of Good n Plenty that I expected would contain candy-coated licorice. Lots of sugar and corn syrup, but less than 2% licorice. This pales in comparison with the $16 billion in US income tax Coca Cola is supposed to have avoided over the years through recognizing profits in low-tax jurisdictions. All signs of demand weakness, in my view.

Super Micro Computer (SMCI). Certainly one of the worst pieces of analysis I’ve done in a long time. A nephew pointed it out to me last year. I read the financials and went through the website. I wasn’t impressed. I thought of the competition among box makers during the early days of the PC and figured we’d see a repeat now. What I didn’t know is at that time the only way to get Nvidia AI chips was by buying SMCI boxes. The stock quadrupled within six months, far outpacing NVDA’s double over the same span.

Since then, SMCI has lost 60% of it’s value, however, and NVDA is ~+30%, meaning NVDA has now outpaced SMCI on a 12 month view. What happened? Remember, I’m the dumb money here, but as I see it, NVDA decided it needed a second source to distribute its AI systems. It announced a wide-ranging partnership with Dell in May, which removed the halo around SCMI. This doesn’t make me less wrong about SMCI last year. What it does underline, though, is that the overall chip-making space is very competitive, where competitive advantage can be a fleeting thing. So buy-and-hold isn’t a great strategy.