usdaed

Effective January 1st, a new series of directives has entered the operational phase. This is not a matter of simple adjustment, but rather a fundamental recalibration of expectations. The implications for those engaged in the development and deployment of AI – and, by extension, for those who allocate capital to such ventures – are becoming increasingly… defined. The year 2026 will not be marked by unbounded innovation, but by a rigorous accounting of processes and potential liabilities.

They increased their stake. That’s what filings are for, isn’t it? Signaling intent. A quarter-end bump of $7.26 million in value, factoring in price movement. It wasn’t a surge, but a steady hand reinforcing a position. Like adding another layer of felt to a worn poker table.

The prevailing wisdom, of course, clings to Nvidia as a moth to a flickering flame. They design the illusions, the shimmering mirages of artificial intelligence, but they do not conjure them from the earth itself. That labor, that quiet alchemy, falls to others. To Taiwan Semiconductor, a titan of understated power, a foundry where the very building blocks of the digital age are forged in the crucible of relentless innovation. They are the silent partners, the unseen hands that shape the future, and they ask for a modest recompense, a mere fraction of the spoils. It is said that the company’s engineers, generations removed from the sun-drenched rice paddies of their ancestors, dream in layers of silicon and gold, a language understood only by the machines themselves. The projections speak of a 60% compounded annual growth rate for AI chips, a figure that would make even the most jaded accountant raise an eyebrow. A mere 23.4 times forward earnings, they say. Almost…reasonable. A rarity in these times.

Apparently, Microsoft spent $37.5 billion on capital expenditures. Thirty-seven and a half billion! What are they doing with all that money? Building a solid gold data center? And then they have the nerve to admit demand for AI is exceeding capacity. Exceeding capacity! It’s like a restaurant bragging they’re too popular. Just prepare, people! That’s what planning is for.

If this arrangement materializes—and one should never underestimate the power of mutual self-deception—Intel’s foundry will experience a fleeting moment of relevance. A reprieve, if you will, from the inexorable march of TSMC. One imagines the celebrations will involve a single, flickering fluorescent bulb and a lukewarm cup of instant coffee.

The reports arrived, of course, filled with the usual cold pronouncements of Wall Street—record fourth-quarter earnings, a 13% surge in revenue, adjusted earnings per share leaping a startling 72%. But these numbers felt less like calculations and more like the inevitable unfolding of a long-foreseen plan, a reckoning of the sea’s bounty. They spoke of streamlining efficiencies, of course, but one could almost hear the shipwrights humming ancient songs as they shaped the company’s fate. Even the slight “miss” on Wall Street’s expectations seemed a mere formality, a fleeting shadow against the brilliance of the larger picture. The true revelation lay in the projections for 2026: a doubling of sales, an increase in capacity of nearly 7%, a promise of continued ascent on the waves of demand.

The SEC filing, dated January 23rd, 2026, is a document of stark simplicity. 61,388 shares extinguished, a position reduced to the ghostly zero. It is not the magnitude of the transaction that chills, but its finality. Elevation Capital didn’t merely trim its exposure; it severed it completely. This is not the cautious pruning of a gardener, but the decisive stroke of an executioner. One is left to ponder the motivations. Was it a matter of principle, a refusal to participate in a game rigged in favor of short-term gains? Or simply a cold assessment of risk, a realization that the promise of buybacks is often a siren song leading to ruin?

It’s fallen a long way since that post-IPO high last August. A lot of people got caught out then, I suspect. And now? Approaching all-time lows. It’s a harsh reminder that even the prettiest designs can’t protect you from the brutal realities of market sentiment. It’s enough to make you question all your life choices, frankly.

Lately, everyone’s been obsessed with memory, but not the kind I’m worried about. They want digital memory. Specifically, the kind made by Micron (MU 0.62%). The stock’s been on a tear, up nearly 280% in a year. It’s enough to make a cynical investor like myself suspect I’ve missed something. Or, more likely, that everyone else is just experiencing a collective delusion.

But here’s the peculiar thing. It isn’t that these REITs are bad. Oh no. They’ve simply been hampered by a rather nasty combination of events – a bit like being stuck in a bog with a grumpy badger. Three particularly troublesome reasons explain this sorry state of affairs.