Investment thesis

The one-page thesis is the deliverable of the entire deep dive. Its discipline is brutal and deliberate: one page. If your reasoning does not fit on one page, you do not yet understand the business. You have notes, not a thesis (the Feynman test, and Buffett's "if you can't explain it simply"). The one page is also the artifact your future self reads when the stock is down 30% and you need to know whether the thesis is broken or merely unpopular. It has a fixed six-part structure, and all five spine questions must be answerable from it.

What goes into a one-page investment thesis?

The structure:

1. The business (Q1). One tight paragraph: what it does, how it makes money, and the real profit engine (often not the biggest revenue line).
2. Why it's a good business (Q2 + Q3). The moat (source + the ROIC-persistence proof), its direction, and the management/incentive verdict. This is where reading the incentives and earnings call analysis land in one or two sentences.
3. The variant perception (the heart). What the market's price implies (from the reverse DCF), what you believe instead, and why you might be right when the market is wrong. One or two sentences of genuine edge. If you cannot articulate a variant perception, you do not have a thesis. You have a quality company at a fair price, which is a fine thing to own but not an edge.
4. The three things that must go right (Q4, from the steelman). The short list of falsifiable conditions the thesis depends on. Three, not ten. These are the break-variables from the bear case, stated as the positive conditions.
5. Valuation & expected return (Q5). Base/bull/bear values, a rough expected value and margin of safety, and what you'd pay. A range, never a false-precision point (M3.05, the valuation workflow).
6. Sell criteria (feeds Phase 9). The pre-committed conditions under which you exit: thesis broken (a break-variable trips), price far above value, a better opportunity, or size beyond tolerance. Written now, before you own it, so the future you cannot rationalize.

Investment thesis example: Titan

Here is the structure filled in: a fully-written model thesis on Titan (illustrative; a teaching specimen, not investment advice; every figure "modeled on ~FY24/FY25 scale — pull the filings"):

TITAN COMPANY LTD (NSE: TITAN) · one-page thesis · [date] · price [₹x] · data through FY25 AR + Q[x] concall · sources: annual reports FY16–FY25, screener.in, 8-quarter concall transcripts, store channel checks (3 cities), Glassdoor/AmbitionBox, RPT annexures 1. The business. Titan is India's largest branded-jewellery retailer; jewellery (Tanishq, Mia, Zoya, CaratLane) is ~85%+ of revenue and the overwhelming majority of profit, with watches and eyewear as smaller, steady, brand-and-distribution-building segments. It makes money on the making charges and studded (diamond) margin on jewellery, not on the gold itself (gold is largely a pass-through, financed and hedged via gold-metal loans). The real profit engine is Tanishq's trust franchise. In a market historically plagued by under-caratage, "Tanishq = guaranteed purity" is a brand people pay a premium for on the most emotionally and financially charged purchase of their lives. 2. Why it's a good business. Durable moat from intangibles/brand + scale + trust in a trust-scarce category; proven in high, persistent ROCE (well above the ~12% INR hurdle) and self-funded growth (broadly flat share count, Fisher 13). Capital-light by retail standards (gold-metal-loan financing, customer advances → low invested capital per rupee of sales). Management is professional, Tata-governed, minority-aligned: the incentive read shows no promoter pledge, a modest and stable Tata brand fee, a defensible CaratLane buyout, and market-level (not cap-hitting) remuneration; the 8-call audit shows strong candor: specific in bad quarters, promises broadly kept, consistent core-metric narration (Fisher 14–15 pass). 3. The variant perception. The price implies mid-teens growth and >25% ROCE for a long CAP: demanding, and the consensus reads the runway as finite. The variant view: India's jewellery market is still ~half-to-two-thirds unorganized, and Tanishq's national share is still modest, so the formalization runway is longer and higher-return than the implied fade allows. The reinvestment opportunity (new towns, studded penetration, CaratLane's digital-native cohort) is the rare case where a great business also has decades of high-ROIC redeployment. The market is right that it's a great business; the variant claim is about runway length and reinvestment, not quality. 4. The three things that must go right. (i) Studded mix and margin hold despite lab-grown-diamond disruption (the one thesis-level risk); (ii) Titan keeps its share of the formalizing market without margin-destroying discounting against Reliance/Kalyan; (iii) the reinvestment runway stays high-return: new stores and studded penetration keep incremental ROIC above the hurdle rather than fading as the base scales. 5. Valuation & expected return. Base case: worth [₹range] on mid-teens growth with margins held; bull: longer CAP + studded premiumization; bear: LGD-driven margin compression toward ~10–11% EBIT + a valuation de-rate. Expected value and margin of safety: computed in your own model and stated for the date of its figures (the numeric example in the worked examples runs the arithmetic on assumed inputs), with the evidence that would move each case named beside it, not a price to act at. 6. Sell criteria. Exit if: (a) studded ratio and studded margin decline for two consecutive years (thesis break: the LGD risk realized); (b) incremental ROIC falls below the ~12% hurdle on the reinvestment (runway exhausted / bad growth); (c) an integrity break: brand-fee creep, a value-destructive RPT/acquisition, or a governance shift (Fisher-15 veto); (d) price rises so far above intrinsic value that expected return no longer compensates (a valuation sell: trim, don't necessarily exit a compounder, per M9.03); (e) a better risk-adjusted opportunity forces the capital elsewhere.

What does a good one-page thesis do?

Notice what the one page does. Every spine question is answered on it: what it does (1), whether it's a good business (2), whether the people are good (2), what can kill it (4 + 6), and whether the price is beatable (3 + 5). It states a falsifiable variant perception rather than a mood. Its sell criteria are written before ownership, so they cannot be rationalized after. And it fits on a page. That is the proof that the deep dive produced understanding, not a pile of tabs. This page, dated and filed, is what M8.05 certifies and what Phase 9 sizes.