How to Write a Dated Investment Thesis: Step-by-Step Guide

What you need

• A stock you already know well enough to explain in two sentences.
• The latest annual and quarterly reports, and any recent 8-Ks, from SEC EDGAR.
• A simple valuation. For special situations this is often net cash per share or a sum of the parts.
• A document or spreadsheet that keeps a version history, so the original text is preserved.
• About an hour for the first draft, then 15 minutes for each review.

Step 1: Date it and state the claim

Write the date at the top. Then write the claim in one or two sentences. A good claim names the outcome, the timing and the gap between that outcome and the current price.

A weak claim: "Company A is undervalued and management is good."

A strong claim: "Company A trades at $8 against about $11 of net cash and saleable assets per share. The asset sale scheduled for a vote in the first quarter of 2027 should turn most of that value into a cash distribution."

The strong version can be wrong, which is the point. If you cannot write a claim that could fail, you do not yet have a thesis.

Step 2: Name the catalyst and grade its date

A thesis needs an event that forces the market to reprice the stock. Our explainer on what a stock catalyst is covers the main types. Write down the event, its expected date and where that date came from.

Then grade the date:

• Fixed: set by a filing or a rule, such as a scheduled shareholder meeting.
• Scheduled: announced by the company but easy to move, such as an expected closing "in the fourth quarter".
• Estimated: your own forecast. Write a range, not a single day.

If you are unsure whether the event is firm or hoped for, read hard vs soft catalysts first. A thesis that rests on a soft catalyst needs a wider date range and more patience.

Rules also set dates you can count on. A 10-Q is due 40 days after the quarter ends for large accelerated and accelerated filers and 45 days for all other registrants. Most material events must be reported on Form 8-K within four business days. These bound when new information can arrive, even when the company is silent. Put every date on your catalyst calendar so the thesis and the schedule stay linked.

Step 3: List the evidence, with sources

Under the claim, list the three to five facts that support it. Each fact gets a source: the filing, the page or note, and the filing date. This makes the thesis checkable by someone else and by you six months later.

Separate facts from inferences. "The company holds $140 million of cash" is a fact. "Management will return it" is an inference, and it should be labeled as one. Our guide to reading a 10-Q for special situations shows where the useful numbers sit.

Step 4: Put a value on it, as a range

Estimate what the stock is worth if the thesis plays out, and give a low, base and high case. A single number hides how much the outcome depends on your assumptions.

Write the main assumption behind each case in one line. The low case should be a plausible bad outcome, not a disaster you do not believe in. If even the low case sits above the current price, say so; if it sits well below, the thesis depends heavily on the catalyst happening as planned.

For companies with debt or large cash balances, it helps to work from enterprise value rather than market capitalization, so the balance sheet is counted once.

Step 5: Write down what the market believes

A thesis is a disagreement with the price. State, in a sentence or two, what the current price seems to assume. Perhaps the market expects the cash to be spent on a poor acquisition, or doubts the vote will pass, or is simply not paying attention to a small stock.

If you cannot say what the other side believes, you may be missing the reason the stock is cheap. This is also where you note any obvious overhang, such as a large holder selling or an index deletion.

Step 6: Set kill criteria before you act

Kill criteria are the facts that would show the thesis is wrong. Write them now, while you are calm, not after the stock has fallen. Each one should be specific and checkable in a filing or announcement.

• If cash falls below $110 million before the vote, the thesis is broken.
• If the asset sale vote fails or is adjourned twice, the thesis is broken.
• If the company announces an acquisition using the sale proceeds, the thesis is broken.
• If the vote has not been held by June 30, 2027, review the case from scratch.

Step 7: Set the review dates

Pick review dates tied to events: the next 10-Q, the proxy filing, the vote, the closing. Add one review on a fixed date in case nothing happens. Between reviews, read every new 8-K for the name.

Step 8: Add dated updates and record the outcome

At each review, add a short dated entry under the original. Say what changed, which kill criteria were checked, and whether the thesis still holds. Never edit the original claim. If the case changes enough that you would write a different claim today, close this thesis and open a new one with a new date.

When the catalyst resolves, record the outcome against the claim: did the event happen, on time, and did the value arrive roughly where you estimated? Note what you got right for the wrong reasons, too. Over time, these outcome notes show which kinds of situations you judge well.

Where this goes wrong

A dated thesis can still be a bad thesis. Writing it down does not make the reasoning sound; it only makes it visible.

The most common failure is a claim that is too vague to test, such as "the market will recognize the value". Another is kill criteria set so loosely that nothing short of bankruptcy triggers them. A third is quietly moving the deadline each time it passes. If you extend the time limit, write a dated update that says so and why, and count the original deadline as missed.

The method also says nothing about position size or price. A thesis can be right and still lose money if the stock was bought at the wrong price or sized too large. Dombey's published research writes each catalyst case in this dated form; see the research archive for how the updates read over time.

Common mistakes

• Editing the original. Once the claim is rewritten, you can no longer tell what you believed on the date. Add updates instead.
• No date on the catalyst. If the event has no date or range, the thesis cannot fail, so it cannot be judged.
• Kill criteria written after the fact. Criteria set after bad news tend to fit whatever you still want to believe.
• Mixing facts and inferences. Label which lines come from filings and which are your reading of them.
• Ignoring the other side. If you cannot say why the stock is cheap, you have not finished the work.
• Scoring by price alone. Record whether the event happened as claimed, not only whether the stock went up.

The bottom line

A dated investment thesis is a short written claim with a date, a catalyst, sourced evidence, a value range, kill criteria and review dates. The date at the top and the rule against editing the original are what make it useful, because they let you judge your past reasoning honestly. Keep the claim to two sentences, give every thesis a time limit, and add dated updates as filings arrive. Over a few years, the outcome notes become the most valuable research record you own.