The single highest-leverage section is the thesis paragraph. A memo whose thesis paragraph is precise ("this stock trades at 12x earnings against a peer group at 18x because the market is over-discounting the temporary margin compression from the 2024 raw-material cycle; as input costs normalize over the next 12-18 months, the company should re-rate to peer multiples; expected return roughly 50% over 18 months; the thesis is wrong if the margin compression proves structural rather than cyclical") forces every later section to be tested against a concrete prediction. A memo whose thesis paragraph is vague ("a high-quality compounder available at a reasonable price") gives the author -- and you, the reader -- nothing to falsify. The discipline of writing a 3-sentence thesis is the discipline of admitting whether you have an actual view. If you cannot write it, you do not have one.
The genre's most common failure mode is the kitchen-sink memo -- 30+ pages, every metric the author found interesting, no clear thesis, no quantified downside. The kitchen-sink memo is a defense mechanism: by including everything, the author transfers the analytical burden to the reader ("here is all the data; you decide"). Treat it as a red flag about the author's conviction. The opposite failure mode is the headline-only memo -- one page, no supporting work, a confident price target with no method shown. Both fail for the same underlying reason: they substitute volume or assertion for the work of committing to a specific, defensible view. A strong long-form memo is 8-15 pages, has a thesis you can quote in one paragraph, shows its method in valuation, and treats risks with the same seriousness as upside.
Going Deeper -- the four pathologies of poorly-written long-form memos. (1) Confirmation-bias structuring: the memo presents only data that supports the thesis, never the contradicting data the author had to reckon with; spot it by checking whether the risk section names specific empirical concerns or hand-waves through generic ones. (2) Backfilled valuation: the price target was decided first, then the DCF assumptions were tuned to arrive at it; spot it by checking whether the implied perpetual growth rate in the terminal value is internally consistent with the explicit-period growth assumptions, or whether they conveniently diverge to produce the desired number. (3) Vague catalyst language: "continued execution" or "steady improvement" are not catalysts; they are the absence of catalysts. A real catalyst is time-bound and specific. (4) Asymmetric coverage: 8 pages of upside, 1 page of downside, and the downside reads like a legal disclaimer rather than a serious quantification. The bull-bear asymmetry section in memo-2 fixes this. AI prompt for self-review: "Given this investment memo, identify the three weakest claims and the one piece of contradicting evidence the author chose not to engage with." The next module turns from structure to the catalysts-and-downside section that separates the genre's professional practitioners from its hobbyists.
Sit with the ideas.
You receive a 22-page sell-side investment memo on a small-cap industrial. You have 10 minutes before a meeting. What is the disciplined order in which to read it so you can form a defensible first opinion?