What an investment thesis is (and is not)
An investment thesis is a short written argument that answers one question: why should this stock be worth meaningfully more than its current price, and how will you know if you are wrong?
A thesis is not a description of a good company. “Great brand, strong management, big market” is a list of facts most investors already know and the price already reflects. A thesis has to contain a disagreement with the market, a reason that disagreement should resolve, and a way to measure it. Three tests separate a thesis from a story:
- It is falsifiable. A future quarterly report could prove it wrong.
- It has a time frame. It says roughly when the market should see what you see, usually within one to three years.
- It is variant. It names where you differ from consensus. If your view matches what analysts already forecast, you have agreement, not an edge.
Ideas often come from a stock screener. The thesis is what turns an idea into a decision you can later check.
How to write an investment thesis: the seven parts
Write the parts in this order, because each one constrains the next.
1. The business in two sentences
How the company makes money, who pays, and why they keep paying. If you cannot explain the revenue model plainly, you will not be able to tell whether a quarterly report is good or bad news. Take the segment breakdown from the latest 10-K, not from marketing material.
2. What the price already assumes
A share price is a bet on future cash flows, so it implies a rate of growth and a level of margins. Work out that implied path before forming a view; a reverse DCF solves for the growth the current price requires, and the stock valuation guide covers the other methods. Your thesis then becomes a statement about a number: “the price implies 6% annual growth; I expect 10% because…”. On VALYSTR, the Belief Gap compares the growth the price implies with analyst forecasts, and The Bar lists the revenue, earnings and margin hurdles the price requires.
3. Your variant view
One or two sentences stating where you disagree with those expectations and why. Good variant views rest on something checkable: a cost program not yet visible in the income statement, a product cycle the market is underweighting, or a loss-making segment hiding a profitable core.
4. Catalysts
Events that should make the market see what you see: a product launch, a price increase, an investor day with new targets, the end of a heavy investment phase. A thesis without a catalyst can still be right, but it may take much longer to pay off than you planned.
5. What must stay true
The three to five conditions your argument depends on, stated as numbers you can check each quarter: revenue growth above a floor, gross margin within a range, customer retention above a level, net debt below a multiple of earnings. These are not forecasts. They are the load-bearing walls of the thesis.
6. Kill criteria
The specific results that would prove you wrong, written before you buy, while you are still objective. “Two consecutive quarters of revenue growth below 5%” is a kill criterion. “If the story changes” is not.
7. Position size and review cadence
Size the position to your confidence and to how much the downside case would hurt, and set a maximum in advance so that averaging down on a falling price does not quietly turn a small position into your largest. Then fix a review schedule, covered below.
Investment thesis template
Copy this into your notes and fill in every row before you buy. A row you cannot complete marks research you still need to do.
| Field | What to write | Test of a good answer |
|---|---|---|
| Company and date | Ticker, date written, price that day | Lets you judge the decision later by what you knew then |
| Business | How it makes money, in two sentences | A non-investor could follow it |
| Priced-in expectations | Growth and margins the current price implies | Stated as numbers, with the method used |
| Variant view | Where you differ from consensus, and why | Names a metric you expect to come in above or below forecasts |
| Catalysts | Events that should close the gap, with rough dates | Each is observable when it happens |
| Must stay true | Three to five measurable conditions | Each can be checked in a 10-Q or 10-K |
| Kill criteria | Results that would prove the thesis wrong | Specific enough that you cannot argue with them later |
| Downside case | What happens if you are wrong | Includes a rough value, not just “it falls” |
| Position size | Starting size and maximum size | Decided before buying |
| Review cadence | When you will re-check | Tied to reporting dates, not price moves |
Investment thesis example (hypothetical)
Illustrative only. Tallow Creek Software is a fictional company, and every figure below is a round, made-up number chosen to show the method. It is not a real security or a recommendation.
Business. Tallow Creek sells payroll software to mid-sized employers on annual contracts. It is moving customers from an older installed product to a cloud version.
Today. The shares trade at $30 with 50 million shares outstanding, a market value of $1.5 billion; assume no debt and no cash. Revenue was $500 million last year, growing about 10% a year, at a 12% operating margin. At a 20% tax rate that is $48 million of net income, or $0.96 per share: a price-to-earnings ratio of about 31.
What the price assumes. If growth stays at 10% and the margin at 12%, revenue in three years is about $666 million and earnings about $1.28 per share. At 20 times earnings, a multiple more typical of a mature software company, that supports roughly $26 a share, below today’s $30. So the price already assumes something better than the status quo, most plausibly some margin improvement.
Variant view. Management targets a 20% operating margin once the migration ends, because running one product costs less than supporting two. The thesis assumes only 18% within three years. That gives earnings of about $1.92 per share, or roughly $38 a share at 20 times earnings (three years out, before discounting back to today).
Catalysts. Cloud migration passing half of customers, disclosed each quarter; retirement of the old product in year two; a price increase at contract renewal.
Must stay true.
- Revenue growth of at least 8% year over year, every quarter.
- Operating margin above the same quarter a year earlier, from the fourth quarter onward.
- Customer count stable or rising, so the migration is not pushing customers away.
- Diluted share count growing no more than 2% a year.
Kill criteria. Revenue growth below 5% for two consecutive quarters; the margin target withdrawn or pushed back by more than a year; a large acquisition paid for mostly in stock.
Size and review. A starter position with a pre-set maximum, checked after every quarterly report against the four conditions.
Common mistakes: thesis drift, confirmation bias and anchoring
A written thesis mainly protects you from yourself.
Thesis drift
The original reason stops being true and a new one quietly replaces it. You bought for margin expansion; margins shrank, so now you hold for “the new product” or “takeover potential.” Changing your mind is fine. Doing it without noticing is not. If the reason changes, write a new thesis and judge it as if you were buying today, at today’s price.
Confirmation bias
Once you own a stock, you read the bull case closely and skim the bear case. Counter it by writing the strongest argument against your position and naming the data that would settle the disagreement. If you cannot state the bear case, you do not yet understand the stock.
Anchoring
Your purchase price feels meaningful, but the market does not know it exists. “I’ll sell when I get back to even” ties a decision to a number unrelated to the business. The relevant question is whether today’s price is attractive given today’s evidence.
When to sell a stock: thesis breaks vs price moves
A thesis separates two events that feel alike: the price falling and the thesis failing. Price is information, but noisy; the results in the 10-Q are the evidence. Decide in advance which you will act on.
| What happened | Thesis status | What your rules might say |
|---|---|---|
| Price fell, every condition still met | Intact | Re-check the evidence; holding, or adding within your maximum, is consistent with the thesis |
| A must-stay-true condition slipped once | Under watch | Find the cause and set a deadline for recovery |
| A kill criterion was triggered | Broken | Exit or cut as your rules state, whatever the price has done |
| Price reached your estimate of value | Played out | Re-underwrite at the new price; the upside you bought may be gone |
| A clearly better use of the money appeared | Intact but outranked | Compare both theses on the same terms before switching |
Two rules keep this honest. First, act on a broken thesis even when the stock is up; gains earned after the reasoning failed are luck, and luck is not a reason to hold. Second, when the price moves sharply without obvious news, look for what the market may know, such as a filing, a guidance change or a sector event, before calling the move noise. The market news feed explains why individual stocks moved, and the earnings report guide covers reading the quarterly results that test your conditions.
Review cadence and a decision journal
A thesis decays if nobody checks it. Tie reviews to the reporting calendar, not to the price on your screen:
- After each quarterly report: check every condition and kill criterion against the new 10-Q.
- After the annual 10-K: rewrite the thesis from scratch at the current price.
- On material news: an acquisition, a CEO departure, a guidance change or new debt. Check whether it touches a condition; most news does not.
Checking conditions by hand gets harder with every holding. VALYSTR’s Thesis Guardian lets you write the conditions your thesis needs to stay true and monitors them, and a company page’s expectations tab shows what the price currently implies (Apple is only a lookup example).
Keep a decision journal
For every buy, add, trim or sell, record the date, price, action, a one-sentence reason, and which condition drove it. Record decisions not to act too, such as holding through a drop. Over a year or two the journal shows patterns no single trade reveals: whether you sell too early, whether your kill criteria actually fire, whether your estimates run optimistic. Judge each decision by the reasoning given what you knew then, not only by the outcome.
Articles in this guide
Put your thesis on watch
Write the conditions your thesis needs to stay true, and Thesis Guardian monitors them so a broken condition does not go unnoticed.
Frequently asked questions
What should an investment thesis include?
A complete thesis covers the business model, what the current price already assumes, your variant view, catalysts, the conditions that must stay true, kill criteria that would prove you wrong, a position size decided in advance, and a review schedule. Most of it should be measurable, so each quarterly report can confirm or weaken the case.
How long should an investment thesis be?
One page is usually enough. The core argument should fit in two or three sentences, with the conditions and kill criteria listed underneath. If you need many pages to explain why a stock is mispriced, the idea is probably not clear yet, and it will be hard to check against quarterly results.
When should you sell a stock?
A thesis-based approach sells when a kill criterion is triggered, when the price reaches your estimate of value so the upside is gone, or when a clearly better opportunity outranks the position. A falling price alone is a reason to re-check the evidence, not an automatic sell signal. This is general education, not personal advice.
What is the difference between an investment thesis and a stock pitch?
A pitch tries to persuade someone else and emphasizes the upside. A thesis is written for yourself, states what the market already expects, and spends as much effort on how you could be wrong as on why you could be right. Its main job is guiding later decisions, not winning an argument.
How often should you review an investment thesis?
Check it after every quarterly report against your conditions and kill criteria, rewrite it fully once a year after the annual report, and revisit it when material news touches one of your conditions. A schedule tied to results, rather than to daily price moves, keeps decisions grounded in evidence.