usdaed

The firm, you see, provides the infrastructure – a sort of digital estate, if you will – for these artificial intelligences to frolic and calculate. They’ve got these Nvidia graphics processing units, which sound like something out of a science fiction novel, powering the whole shebang. Clients can rent cloud capacity, build AI models, and generally engage in all sorts of clever technological wizardry. One imagines a great deal of blinking lights and humming machinery.

Oklo, you see, proposes to build reactors – miniature suns, if you will – to supply this insatiable demand. The stock, briefly inflamed by optimism, has since suffered a rather precipitous decline – a fall of sixty-five percent from its recent zenith. One is left to ponder: is this a moment for the prudent investor to ‘buy the dip,’ or merely to observe the spectacle with a knowing smile?

The filing, dated February 17th, 2026 – a date which, I confess, feels strangely distant even as I write it – reveals a shedding of weight, a trimming of the portfolio. Howard Hughes Holdings, a name that evokes images of aviators and grand ambitions, has seen its shares hover at $82.15, a price that, while respectable, is but a shadow of what was once promised. A mere 9.5% increase over the year – a performance that, shall we say, lacks the vigorous pulse of a healthy investment. It underperformed the S&P 500 by 2.3 percentage points, a difference that, while seemingly small, is enough to cause a seasoned trader to feel a slight tightening in the chest.

So, what’s the snag? The mildly unsettling fact that two customers account for roughly 80% of that revenue. This, naturally, causes a degree of consternation amongst portfolio managers. And with good reason. When your business model is predicated on the whims of a handful of hyperscalers – those vast, data-intensive entities that seem to power everything these days – a delayed AI build-out (or, worse, an inventory digestion quarter – a phrase that sounds vaguely medical) can send the stock plummeting faster than a poorly-aimed satellite.

The purchase, 52,000 shares to be precise, was a small tremor in the vast tectonic plates of Wall Street, barely registering on the instruments of the hurried traders. Yet, Mateo, who had learned to read the silences between the numbers, understood it as a declaration. 13D, it seemed, believed Workiva held something of value, something beyond the fleeting metrics of quarterly earnings. It represented 5.34% of the fund’s U.S. equity holdings, a deliberate weighting in a portfolio otherwise occupied by the familiar constellations of Twilio, MRCY, VSAT, ALV, and PSO – companies that, like brightly colored birds, flitted across the landscape of innovation.

Nvidia. They’re everywhere, aren’t they? Like particularly industrious gnomes, burrowing into every corner of the digital world. They don’t just make things; they seem to enable everything else to be made. And in the realm of quantum, they’re not trying to build the entire clockwork universe themselves – a common mistake amongst ambitious wizards². They’re providing the tools – the forges, the anvils, the exceptionally precise calipers – for others to do so.

There’s been research done, a whole field of people staring at stock charts during wartime, and the results are… inconclusive. The CORP-DEPO folks have charts, graphs, the whole nine yards. But trying to predict market behavior based on conflict is like trying to predict what your boss is going to ask for on the way out the door – it’s a crapshoot. World War II is often cited as a boom time, but that was then. Now, we’ve got algorithms and high-frequency trading, which means the market reacts to things before anyone fully understands what’s happening. It’s like a really caffeinated squirrel.

Two companies, Nvidia and Broadcom, stand as particularly well-positioned beneficiaries of this unfolding spectacle. The question, then, isn’t which represents the better investment, but rather which will prove slightly less exposed when the inevitable reckoning arrives. A distinction, perhaps, of degree rather than kind.

They’re pushing this new ES8 SUV, which, if the marketing materials are to be believed, is the automotive equivalent of a cashmere bathrobe. Bigger, more expensive, better margins. It’s the sort of vehicle my Uncle Barry would covet, then immediately complain about the cost of the floor mats. The revenue grew 76% year over year to almost $5 billion. It’s a big number, and I found myself staring at it, trying to remember what I had for breakfast. It felt… disconnected. Like looking at a spreadsheet from another planet.

NuScale, you see, proposes to construct power plants not of towering immensity, but of a modest, almost charming, scale. Their reactors, they claim, can be contained within vessels of but sixty-five feet in height and nine in width – a feat of engineering that suggests a certain… compactness. These are prefabricated, assembled on site, and intended to reduce the exorbitant costs typically associated with harnessing the atom. They have, indeed, secured the approval of the U.S. Nuclear Regulatory Commission – a laurel wreath, to be sure, but one that does not guarantee triumph.