Daily Commentary

MarketMinder’s View: As this article references several companies, please note MarketMinder doesn’t make individual security recommendations—any firms mentioned here are coincident to a broader theme we wish to highlight. Here is another fear surrounding the booming AI industry, this one regarding accounting treatment of graphics processing units (GPUs), chips used for advanced large language models—and one of AI firms’ biggest costs. Companies buying loads of pricey GPUs are required to estimate how fast they will depreciate for accounting purposes. As the fear goes, “Choosing a longer period spreads out the cost over more time, allowing companies to report higher profits now. But there’s a risk in claiming that equipment will be good for six years if, in reality, it will have lost almost all of its value in four. A company might have to buy new, even more expensive chips sooner than expected. Any loans it took out using the now-useless chips as collateral, meanwhile, could get complicated. At the very least, a company might have to write down the value of obsolete equipment, creating a one-time hit to profits.” First, as the headline inadvertently demonstrates, this isn’t a “big unproven assumption,” but an issue investors are aware of. Capital depreciation isn’t a hidden risk, but one of many factors companies report—and analysts question. It is just receiving more attention now amid all the AI hype (and pushback), which also shows potential accounting discrepancies aren’t flying under the radar. Suprises move stocks most, and this isn’t among them. Second, contrary to concerns, the Big Tech firms buying all those GPUs have increased the useful life of their servers and network equipment over the last five years through software efficiency gains, extending their assets’ life. We don’t dismiss the potential for this to become a risk, but for now the fear seems overstated. For more on mushrooming AI fears, please see yesterday’s commentary “What to Make of AI Bubble Talk.”

MarketMinder’s View: The examples highlighted here show that while tariffs aren’t without costs—they are a tax on importers—they hurt some (small US retailers) more than others (large multinationals with global supply chains). We think this is one reason why small businesses are shedding jobs more than larger ones—a headwind and hardship for some, though not necessarily a big negative for the broader economy or markets. (Since the article names some, please note MarketMinder doesn’t make individual security recommendations; the references serve only to underscore the broader theme.) Consider the plight of one small-sized retailer: “When [President Donald] Trump threatened tariffs as high as 180% on Chinese imports in mid-April, [sleep-wellness brand founder Matt] Hassett explored shifting production to Thailand, where duties were lower. But when the rate on China was later cut to 20%, the alternative factories with 20% higher production costs proved to be costlier than the tariffs. In the end, Hassett stuck with his Chinese manufacturer. But the scramble delayed orders, leaving him dangerously short of stock ahead of the year’s busiest shopping season. November and December typically account for a third of U.S. retailers’ annual profits.” Whereas much larger outlets “can soak in the supply jitters by leveraging scale more easily than smaller firms. Operating margins for small retailers with total assets less than $50 million have plunged to negative 20.7%, according to business analytics provider RapidRatings, leaving 36% of them at a high risk of bankruptcy compared to 12% of large retailers.” With bigger firms making up the bulk of the market, stocks can mostly look through them. That doesn’t mean they are ignoring tariff costs, but sensibly scaling them—and moving on.

MarketMinder’s View: After all the handwringing over the UK Budget: “In a release first reported by Reuters, the Office for Budget Responsibility (OBR) said the headroom—the amount of extra spending or tax cuts possible for the government while staying within its budget rules—now stood at almost 22 billion pounds ($28.9 billion) in five years’ time. A Reuters poll had expected Reeves to leave herself just under 17 billion pounds of headroom, up from just under 10 billion. Thirty-year gilt yields were down 11 basis points in late trading at 5.215%, set for their largest one-day drop since mid-April, as prices surged. The rally in bond prices gathered pace over the afternoon, as the amount of new longer-dated debt the government intends to sell was expected to decline, after the Debt Management Office cancelled several planned auctions, analysts said.” While we caution against reading too much into one-day moves, the market reaction to the UK’s much-anticipated Budget indicates investors weren’t surprised (at least negatively). When it comes to potentially big government policy changes, getting clarity allows uncertainty to fade and markets to get on with pricing the next 3 – 30 months. Whatever the actual policy shifts and their effects, clarity typically soothes sentiment as markets see reality more clearly. Like one analyst put it, “It could have been a lot worse and that’s what the market was fearing.” Businesses and households can move forward knowing the lay of the land. For more on what is in the UK’s Budget—and why markets appear none too bothered by it—please see today’s commentary, “Few Surprises: Leaks and Trial Balloons Mute the Market Effects of Britain’s Tax Shifts.”