Here's a question that breaks the word "competition."
When a pension fund, an insurer, and two rival asset managers all sit down to decide what to buy and sell — and all four of them are reading their risk numbers off the same software, licensed from a single vendor — in what sense are they competing?
That software is Aladdin. The vendor is BlackRock. And the answer is: they're not. They're four terminals on one machine.
Forget the "Aladdin is the brain of capital" framing for a minute. That's the philosophy. This is the plumbing — colder, more concrete, and in some ways scarier. Aladdin isn't only powerful because it's smart. It's powerful because everyone rents it, including the people who are supposed to be BlackRock's rivals. That's not a brain. That's a platform monopoly wearing a risk-management badge.
What "renting the same engine" actually means
Aladdin is licensed software. Firms pay BlackRock for access — and by BlackRock's own account it sits underneath assets on the order of $20+ trillion, far more than BlackRock itself manages. That gap is the whole story. The majority of the money watched through Aladdin isn't BlackRock's. It belongs to clients — many of them BlackRock's direct competitors, insurers, pension systems, and sovereign wealth funds.
Think about what a firm gives up when it plugs in. It stops building its own risk view. Its portfolio managers, its compliance, its stress tests, its overnight reports — all shaped by one vendor's models, one vendor's data pipes, one vendor's assumptions about how the world moves. Traders learn to think in Aladdin's categories. Compliance learns to file in Aladdin's format. Regulators start asking for Aladdin's outputs.
That's vendor lock-in, and any engineer knows the shape of it. It's the same trap as building your whole company on one proprietary cloud, one closed database, one API you don't control. At first it's convenience — why reinvent risk modeling? A year later it's dependency. Five years later, ripping it out would mean rebuilding your nervous system from scratch, retraining every hand, re-certifying every process. So nobody rips it out. The switching cost is the moat. BlackRock didn't have to beat its competitors on risk modeling. It just had to become the thing they can no longer function without.
Why there's no "competition" left to speak of
We're told the market is a contest of many independent judgments. That diversity of opinion is the immune system — when I bet the world tilts left and you bet it tilts right, the market stays balanced, resilient, alive.
Now collapse that. If the pension fund, the insurer, and the two "rival" managers all model risk on the same engine, they don't hold four independent judgments. They hold one judgment, licensed four times. They see the same risks, flag the same dangers, get nudged toward the same trades. The appearance of competition, the reality of a monoculture.
Our record: competition is supposed to be the market's immune system — many independent judgments, so no single error can infect the whole body. Rent the immune system from one vendor and you don't have four defenders. You have one defender, cloned. When it's healthy, the body looks fine. When it's wrong, nothing in the system is holding a different view to catch the mistake. This is Isfet's most elegant trick: sell everyone the same "safety," and quietly dissolve the very diversity that safety depended on. The engine that promises to measure risk becomes, at scale, the thing that manufactures it — one shared blind spot for the entire herd.
Look at agriculture for the word: monoculture. Plant one genetically identical crop across every field, and the yield is spectacular — right up until one blight arrives that the single genome can't resist, and then it isn't one farm that fails. It's all of them, the same day. Financial risk built on one shared engine is a monoculture of judgment. Efficient. Uniform. And catastrophically exposed to the one thing the model gets wrong.
The single point of failure nobody stress-tested
Here's the irony that should keep regulators up at night: Aladdin's entire job is to find risk — and its ubiquity has quietly created a systemic risk it cannot see, because it is the risk.
Run the failure modes an engineer would run on any critical system:
- A model bug. One flawed assumption in Aladdin's code doesn't stay in one firm. It propagates through every institution running that build, all at once.
- A shared blind spot. If the model underweights some danger, it underweights it everywhere it's installed. There's no rival model in the room to say "wait."
- A herd exit. When enough capital reads "reduce risk here" off the same screen on the same morning, they all reach for the same exit simultaneously — and the model's warning becomes the crash it warned about.
- A vendor problem. Outage, breach, capture, or simple concentration of power in the firm that owns the engine — and a fifth of the world's investable capital is looking through a lens controlled by one company.
In software we treat this as gospel: never run your whole fleet on one un-audited proprietary binary, because the day it fails, everything fails together. Global finance quietly did exactly that — routed a huge share of the world's capital through one black box whose internals a handful of people can read. The efficiency is real. So is the fragility. It's one fact seen from two sides.
Where the lever is
No doom. Lock-in has an exit — it's just up a staircase most institutions are too comfortable to climb.
Name the dependency out loud. The first defense against lock-in is refusing to call it neutral infrastructure. It isn't. It's one company's proprietary product sitting under a fifth of the world's capital. Say that sentence to anyone who calls it "just risk software."
Demand plurality. A healthy system runs many models, ideally including open, auditable ones. If your fund, your pension, your regulator can't say what risk engine they depend on and who controls it, that's the audit to demand first.
Back the open alternative. The counter-move to one proprietary black box is transparent, forkable logic. On-chain finance already runs risk, lending, and settlement on open code anyone can read, audit, and fork. No single Aladdin behind the curtain deciding what the numbers mean — the logic is on the table, and ten thousand eyes find the bug that one room misses.
Refuse to be a terminal. At every scale, from a global fund to your own portfolio, the same principle holds: don't outsource your entire judgment to a lens you can't inspect and can't control. Keep at least one independent view alive. In a monoculture, the seed that's genetically different is the one that survives the blight.
The competition isn't competing. It's four windows into the same machine. Once you see that, "beating the market" and "reducing risk" both need a rewrite — because right now the market and the risk are, increasingly, the same rented engine.
Diversity of judgment is the thing worth defending. Start by refusing to be cloned.