Evaluating Stocks Like Products: The Problem

May 7, 2026

Since mid-2024 I've worked as a lead engineer at Summit Management, a hedge fund. My job is the fund's system: data pipelines, portfolio returns and performance, the infrastructure behind investment decisions.

Along the way I picked up the basics: P/E ratios, ROIC, debt-to-equity, revenue growth. Enough to follow the conversation, not enough to call myself an analyst. At some point I asked myself a simple question: if I were to invest my own money, what would I actually look for?

I'm a product engineer. I've spent years on both sides, product and tech, and that shapes how I think. I build software, but I also ask whether something should exist at all. Is the problem real? Must-have or nice-to-have? What happens if nobody solves it? What are the alternatives? Those are the questions that build conviction for me.

The natural place to apply them was the company itself. That instinct was half right. The questions were the right questions. The subject was not.

Companies don't have problems. They solve problems that the world has. The world has the problem. The company is just who happens to capture it.

It splits in two, and the order matters:

The thesis. The permanent need itself, scored on its own without any company attached. If the need is not real or not durable, no company matters.
The player. A candidate that might capture the thesis. Scored on the business itself, on its defensibility, and on how much of it actually points at the thesis. If the candidate doesn't capture the need, the financials don't matter.

Both have to hold: a thesis that's alive and a player that's strong enough. One doesn't save the other.

This post is about the thesis. The player comes next.

The unit isn't the company

The natural starting point is the company, because the company is what you can actually buy. Pick one, run the product questions on it, decide. The trouble is that by the moment you've picked, you've already chosen the problem to evaluate. "Is this problem real" turns into confirmation, not exploration.

Starting from the world reverses the order. Articulate the problem first, then ask who could capture it. Companies become candidates, not the unit. So instead of asking whether MSCI is a good business, I ask whether benchmarking institutional money is a permanent need. If yes, MSCI is one of the names worth a look. If no, MSCI isn't interesting at any score.

What a thesis is

So what counts as a real thesis? Not a vague belief. "AI will change the world" isn't a thesis. But "pension funds, ETFs, and insurers are required by law to measure performance against externally administered benchmarks, and that requirement won't go away" is, because I can check it and I know what would make me wrong. That sentence is the start of a real thesis I keep in my registry, called Financial Index. I'll use it as the example throughout.

Each thesis I keep has five pieces:

Problem Score. How severe and durable the problem is.
Base rate. The reference class for this kind of need and how often it survives.
Trajectory. The direction it's moving.
Signals. What would make me walk away (falsifiers) or sustain my conviction (confirmers).
Status. Where the thesis stands today, derived from the signals.

These are below. The roster gets its own section after them.

Problem Score

One to five, how severe and durable the problem is. Four sub-questions feed in: whether anyone solves it without this kind of provider, whether it's must-have, whether regulation mandates it, and whether it's getting worse. A five is critical infrastructure, mandated or close to it. A one is a perceived problem with no urgency.

Financial Index scores a 5 on all four. ERISA, MiFID II, and Solvency II all enforce some version of the mandate. There is no parallel solver structure outside externally administered benchmarks.

Base rate

Problem Score is an inside view: how severe and durable the need looks on its own merits. Base rate is the outside-view check: how often have needs in the same class actually survived over the long run?

The reference class for Financial Index is independent third-party financial-data infrastructure providers operating under explicit regulatory recognition: the Big-3 credit ratings (post-1975), the Big-3 index providers, CUSIP (single-vendor since 1968), CRSP. Across the last forty years, seven incumbents preserved top-3 dominance in their primary asset class, and around ten entrants tried without displacing a single one. The base rate of preserved dominance sits at 85%.

The inside-view adjustment is that index providers carry deeper lock-in than the rest of the class. Switching a benchmark breaks a public fund prospectus (SEC Form N-1A, Rule 6c-11) and triggers an ERISA fiduciary review. That's a stronger constraint than the internal contracts behind credit ratings or identifiers. Index revenue also rides global AUM growth instead of issuance cycles, which compounds the durability the 85% already captures. The expected players sit in the upper tail of the reference class.

The danger the base rate catches is conflating critical infrastructure with the survival of any one incumbent. The need outliving most incumbents is the typical case for infrastructure rails. That's why the thesis is the unit, not the ticker.

Trajectory

Improving, stable, or weakening. The score is a snapshot. The trajectory is where it's going.

Financial Index is intensifying. Five primary-source signals from 2024 to 2026 all push the same way:

Index funds rose from 36% of US equity fund assets in 2016 to 57% in 2025. Passive crossed active in late 2024 and kept widening through 2025.
EU Regulation 2024/3005 extends the regulated-administrator template from financial benchmarks into ESG ratings, with ESMA registration and separate methodology.
Solvency II Review Directive 2025/2 amends Article 132 to require macroeconomic and macroprudential considerations in investment decisions, effective 2027.
The US Department of Labor's Proposed Rule on DIA selection operationalizes "meaningful benchmark" inside ERISA fiduciary process.
Direct indexing AUM, projected to reach roughly $1.1T by 2028, consumes standardized index methodology as a starting universe rather than displacing it.

Signals

Signals come in two directions.

Falsifiers. Concrete thresholds that, if crossed, kill the thesis.
Confirmers. Concrete thresholds that, when met, register evidence the need is still there.

Falsifiers come first. The discipline they impose is what makes everything else work. If I can't write down what would make me wrong, I don't have a thesis. I have a feeling.

Both come from adversarial research, queries designed to break the thesis rather than confirm it. If I can't find a primary source for a claim, the claim doesn't make it in.

Open falsifiers under Financial Index:

The UK FCA publishes a policy statement imposing FRAND or open-access remedies on benchmark licensing before the end of 2027.
The European Commission formally proposes extending the 2024/3005 template (separation of business, ESMA authorization, transparency on methodology) to non-critical financial benchmarks before the end of 2028.
ESMA's register of financial-benchmark administrators drops more than 50% by end-2026, alongside primary-source evidence of buy-side migration to unregistered boutique benchmarks.
Four consecutive quarters of equity-segment passive net outflows alongside active equity net inflows.

Confirmers, all currently firing:

ICI 2025 Investment Company Fact Book: $39.2T total US-registered investment company AUM at year-end 2024. ETF AUM crossed $10T for the first time. Long-term index funds took $109B in February 2026 net inflows versus $35B for active.
EU BMR Regulation 2025/914 (effective January 2026) preserves the authorization-plus-supervision regime for "critical" and "significant" benchmarks. That is the tier where dominant index incumbents operate.
Solvency II Review Directive 2025/2 preserves Article 132's prudent-person principle through the 2027 transposition deadline.
ESMA's recognition of CME Group Benchmark Administration in April 2025 confirms continued operational use of the third-country recognition regime. The EU regulatory perimeter remains the gating mechanism for benchmark use.

Each is concrete enough that a quarterly check tells me whether the thesis is still standing.

Status

Where the thesis stands today. Five possible states.

Alive. Falsifiers haven't fired and at least one confirmer is.
Searching. Registered without a vehicle yet, monitored until a candidate qualifies. No forcing.
Weakening. Evidence is starting to push the other way and I owe myself an explanation.
Uncapturable. The need is real, but the capture universe is closed. No public vehicle exists.
Falsified. Evidence has crossed a threshold I committed to in advance. The thesis is dead.

The state I underestimated at first is Searching. The temptation is always to attach a ticker to a thesis as soon as you have one. The discipline is to wait.

Financial Index is Alive.

Filters before the roster

Not every alive thesis becomes investable. Some sit in sectors I can't evaluate from primary sources, like biotech mechanism-of-action, deep semiconductor process economics, or opaque commodity chains. Those go to a "too hard" list, neither alive nor falsified, just outside the lens. Financial infrastructure is inside.

The roster

Public companies, private incumbents, regulators, ecosystem participants. I monitor the roster, not any single name on it. At this stage every entry is a candidate. Whether any becomes a position depends on how it evaluates.

The Financial Index roster:

Role	Examples
Index providers	MSCI, S&P Global, Bloomberg Index Services, FTSE Russell, Nasdaq, Intercontinental Exchange
Ecosystem	Bank for International Settlements, IMF, IOSCO, CFA Institute
Regulators	SEC, US Department of Labor, ESMA, UK Financial Conduct Authority

What does a quarterly signal look like? A new SEC rule on ETF index licensing, a competitor cutting prices on a thematic index family, a new IOSCO principle on benchmark administration. None of these moves quarterly numbers immediately, but each is a signal about the thesis. The roster is what makes the thesis monitorable.

The cascade also runs the other way. If the thesis is Falsified, every position under it exits with it, no matter how strong the player looked on its own. Player strength does not save a dead thesis.

What this changes

The questions I started with are still the questions. Is the problem real, must-have, getting worse, what happens without it. What I changed is where I point them: not at the company, but at the world's need. The company shows up as one of the candidates the need lets in.

What I haven't covered is how I pick a candidate once the need has a roster. That's the next post.

The lens didn't change. The unit did.