The Thesis That Survived Everything

Most investing conversations I come across — on social media, in forums, and in person — share a pattern. Someone explains why they own a stock. The sector has a long runway. The business has a strong track record. Management has been consistent. All of it is true. All of it sounds like a thesis.

Then ask what would change their mind. The answer softens. Things would have to deteriorate. Results would have to disappoint consistently. Nothing specific — just a general sense that they would know it when they see it.

The thesis sounds complete. It just has no failure mode.

A thesis that nothing can break is not a strong thesis. More often, it is a theme.

Why does a theme feel like a thesis when it isn't?

My original thesis for CDSL was: a toll on India's rising financialization.

That felt complete. It had a mental model, a macro backdrop, and a structural reason for the business to benefit. Reading it back, it sounds like a proper thesis.

But try to break it. If financialization slows? India's financialization has moved in one direction for decades — that is not a real falsification condition. If something goes wrong at CDSL specifically? The thesis never said anything specific about CDSL. It said the financialization wave is rising, and CDSL sits in its path. That statement survives any quarter because it was never precise enough to be contradicted by one.

What I had was a theme, not a thesis. A theme is a backdrop. It justifies owning a dozen different businesses simultaneously — CDSL, CAMS, BSE, any AMC — and cannot be falsified because nothing was ever specified that would break it. A claim that applies to everyone tests no one.

The distinction matters because everything downstream depends on it: what you monitor, how you interpret a bad quarter, and whether the thesis is playing out or breaking down when the price moves against you.

What does a falsifiable thesis actually look like?

The thesis I eventually converted my CDSL position to was different in one specific way: it traced the growth mechanism.

Rising financialization means more Indians entering equity markets. More participants mean more demat accounts. More demat accounts mean more folios across issuers. CDSL charges issuers annually per folio and earns on every transaction. So if financialization continues, the earnings follow mechanically through that chain.

Once the mechanism is named, what to watch follows almost automatically: folio count growth per issuer, total issuer additions, and per-folio revenue realisations each quarter. And now a failure mode exists: if folio growth stalls for two or three consecutive quarters despite a healthy market environment, something in the mechanism has broken. That is a signal worth investigating. That is not noise.

The original thesis and the converted thesis point to the same business and the same macro tailwind. The difference is that one names how the macro story reaches the company's revenue, and the other does not. Naming the mechanism is what makes a thesis falsifiable — once you have specified how the tailwind reaches earnings, you have also specified what would interrupt it.

How do you make a thesis specific enough to work with?

Three steps, applied in order. These work whether you are writing a thesis before buying or testing one you have already held for a year.

Name the mechanism, not just the backdrop

The backdrop is the sector tailwind or the macro condition — the thing that is true for many businesses. The mechanism is the specific chain through which that condition reaches this company's revenue. "CDSL benefits from financialization" is a backdrop. "Rising folio counts per issuer drive CDSL's annual issuer charges, which is where the revenue actually comes from" is a mechanism. The mechanism is what makes the thesis testable, because once you have specified the chain, you have specified where it can break.

Derive the KPIs from the mechanism, not the sector

A precise mechanism generates its own metrics; you do not have to search for them. For CDSL, once the mechanism is clear, the KPIs follow directly: active folio count per issuer, total issuer count, and annual issuer fee realisations. Not total demat account additions — that number gets reported most prominently, but it is a step removed from how CDSL actually earns. If you find yourself writing KPIs that apply to any business in the sector — revenue growth, PAT margins, RoE — the mechanism is not specific enough. Go back to it before looking for better metrics.

Write the failure mode before you need it

Not "if results disappoint." A specific, observable event that would tell you the load-bearing assumption in the mechanism has failed. For the CDSL thesis: if folio growth stalls across the issuer base for three consecutive quarters despite a stable market environment, the mechanism is not working. If regulatory intervention introduces fee caps that compress per-folio realisations materially, the economic assumption needs revision. These conditions can be monitored and triggered. "Things deteriorate" cannot.

Writing the failure mode is the most skipped step, and the discomfort of writing it is diagnostic. An investor who genuinely understands the mechanism finds the failure mode relatively easy to write, because the load-bearing assumption is already named. If you cannot write it, either the mechanism is not named yet, or you do not want to name it because doing so makes the thesis feel more contingent. That contingency is not a weakness in the thesis. It is what makes it honest.

What happens when a thesis has no failure mode?

You would expect an unfalsifiable thesis to produce confident holding. In practice, it tends to produce the opposite.

Because the thesis can neither be confirmed nor denied, every results season becomes its own fresh judgment call, made under whatever emotional conditions happen to be present that week. When the stock is up, every data point looks like confirmation. When it is down, the same data points feel ominous. The thesis is compatible with both readings — which means it is doing no work at all.

Investors who stay rational through volatility are not always the ones who feel the most conviction. They are the ones who wrote a thesis specific enough that the first question after a bad quarter is not "how bad does this feel" but "has the folio growth assumption actually changed." That question has an answer. The answer either settles the matter or tells you something real has shifted.

Rules protect you from taking actions. They do not eliminate the emotions underneath. The only thing that does — or at least reduces it — is a thesis specific enough that the emotions have something concrete to push against.

Why I built this into Stockport

When I see investors pause at the thesis step, it is not usually because they do not understand the business. It is because they have been holding a theme and calling it a thesis — and writing it down makes that visible in a way that keeping it in your head does not.

The goal is not a thesis that reads well. It is a thesis that a bad quarter can actually test.