Building conviction in a thesis: a checklist for self-directed investors

Most retail investing mistakes aren't analytical. They're failures of discipline. A thesis you can't write down in 5 sentences is one you can't defend in the next drawdown, and a position you can't defend is the one you sell at the worst possible time.

Below is a 7-point checklist for building conviction you can stand behind. It works on its own, and inside Finapolis. The discipline is the point, not the tool.

A thesis you can't write down in five sentences is one you can't defend in the next drawdown — and a position you can't defend is the one you sell at the worst possible time.

What conviction actually means

Conviction isn't certainty. It isn't stubbornness either. It's the discipline to articulate what you think, why you think it, and what would prove you wrong, all before you put money on the line.

The best test of a thesis isn't how it sounds in your head. It's whether you can write it down in 5 sentences and hand it to someone who has never heard of the company. If they can read it, ask 1 question, and understand the bet, you have a thesis. If you can't finish the 5 sentences, you don't have one yet.

The 7-point checklist

Each step is a paragraph or 2, not an essay. You should be able to do this in 1 sitting, not 1 week.

1. Write the thesis in 5 sentences

The 5 sentences are: what the business is, why it's undervalued (or attractive) right now, what catalyst closes the gap, the time horizon you're giving the thesis, and the single biggest risk. No jargon, no hedging.

If you can't do this, the work isn't done. Open the 10-K, the latest earnings call transcript, and a peer comparison, and try again.

2. Name the catalyst

A catalyst is a specific thing that will happen by a specific time. "Earnings grow" isn't a catalyst. "The next 2 quarters of guidance reset the multiple from 12x to 15x" is.

If your catalyst is "the market will eventually figure it out," your timing assumption is also "eventually." That's fine for a long-duration thesis, but it has to be sized accordingly (see step 5).

3. Define the downside case in dollars

If the thesis is wrong, what do you lose? Not "I'll get out." Not a stop-loss. The actual dollar loss as a percent of the position, in the realistic bear case where the catalyst doesn't happen and the multiple drifts to the trough of the cycle.

For most quality businesses, this number is 25% to 40%. For thesis-on-thesis names (small caps, early-cycle, leveraged balance sheets) it's often 50% or more. If you can't stomach the downside, don't take the position.

4. List what would change your mind

This is the most uncomfortable step, and the most important. Before you buy, write down 2 or 3 specific things you might see in the next year that would invalidate the thesis. Segment growth below a stated number. Margin compression to a stated level. A named competitor making a specific named move.

These aren't stop-losses. They're pre-commitments to revisit. The point is to take "I have an open mind" out of the realm of self-flattery and put it on paper.

5. Size the position to the conviction

Position size is the most underused tool retail investors have. Conviction and size should move together. A useful default ladder:

• 1% to 3%. A name you're watching, building a position in, or hold with low conviction.
• 3% to 6%. A standard conviction position with a clear thesis and downside case.
• 6% to 10%. A high-conviction position. Reserved for the small number of names you'd still buy if they fell 20% tomorrow.
• 10% or more. Rare. Should require a written defense and a downside case you're comfortable showing a peer.

Most retail books are too concentrated in low-conviction names and too thin in high-conviction ones. The checklist is meant to help you see which is which.

6. Pre-commit to a review date

Pick a date 90 or 180 days out and put it on the calendar. On that date you reread the thesis and ask: is it intact, has it played out, or has something changed? You revisit regardless of price.

Revisiting on price is anchor bias. Revisiting on time is discipline.

7. Write down why you would sell

Two bullets, on the same page as the thesis:

• Thesis met. The catalyst has played out and the position should be trimmed or exited.
• Thesis broken. One of the items in step 4 has actually happened.

You don't get a 3rd reason to sell. If the price falls and neither bullet is true, you hold or add. If a bullet is true, you act, regardless of where the price is.

What this looks like inside the workbench

A few of the steps above are clerical. A few are analytical. The Analyzer module in Finapolis is built for the analytical ones: editable DCF and sensitivity for step 3, peer benchmarks for step 5, scenario tables for step 4. The Reporter module surfaces the sourced numbers that go into each step.

None of this thinks for you. The math runs faster, the sources are auditable, and the work product is the same 5-sentence thesis on the page in front of you.

Where the checklist saves you

Three situations where the discipline is worth more than the analysis.

• Drawdowns. When a position is down 20% and the news is loud, you reread the thesis and step 4. If nothing in step 4 has happened, you hold.
• News cycles. A single quarter, a single headline, or a single downgrade doesn't invalidate a thesis written for a 2-year horizon. The checklist makes this concrete.
• Anchor bias. Once you've written down what would change your mind, "I've always liked this company" stops being a reason to add.

FAQ

How long should writing a thesis take?

2 to 4 hours for the first pass, including a focused read of the 10-K and the latest call. 30 minutes to revisit on a scheduled review date.

What if I can't get to 5 sentences?

Don't take the position. The 5 sentences are the floor, not the ceiling. If you can't articulate the bet, the bet isn't ready.

How do I size to conviction if I have only a few names?

The same ladder applies. A 6-name portfolio with 1 high-conviction position at 10% and 5 mid-conviction positions at 18% each is a legitimate book. A 6-name portfolio with 6 equal positions is a default, not a decision.

When do I sell?

Only from step 7. If a thesis is met, you trim or exit. If a thesis is broken, you exit. Everything else, including a falling price, is noise relative to your written reasons.

Does this work for ETFs and index positions?

The catalyst and the 5-sentence thesis still apply. A sector ETF position should have a written reason. The downside case and step 4 matter less for diversified ETFs because the basket dampens the specific risks the checklist is built to surface.