Worked sample · printable PDF
INV-02 · Who gets paid at $40M, $120M, $400M?
If it is sold for $40 million, $120 million, or $400 million, who gets paid what?
Specialty centers · Venture & Private Capital
Underwrite the company, then audit the model that helped you do it
Twelve free capstone briefs and twelve short labs on fictional case files. Students read a cap table, find every kink in an exit waterfall, model a fund, and build an AI memo drafter that fails the build if it invents a number. Every brief has its own PDF; the bank PDF includes the labs. No account. No student data leaves your school.
An investment study, not an investment
No real money, no brokerage or trading accounts, no buy or sell recommendations, and no soliciting investment from anyone. Every company, fund, and cap table in the case files is fictional; public companies appear only through their public filings, analyzed as cases. Outside contact is arranged by the school. Nothing here is legal, tax, or investment advice.
The ground-truth rule
The calculator is the source of truth. The model is a junior analyst. A number from a model is a claim until it is reconciled.
AI work is tagged USE, BUILD, or AUDIT. BUILD runs on a stub provider only: no API key, no live model, no network call.
Featured packets
Featured · printable PDF
INV-06 · The analyst that never invents a number
Build an investment-memo drafter that cannot invent a number. Every number it writes must cite its source, and every number must match the calculator, or the build fails. Can you prove it, on twenty test cases you wrote to break it?
Catalog (10 briefs on the page)
- INV-01 · Read the cap table before the pitch
Where do the data room, the pitch deck, and your reconstruction disagree, and why?
- INV-03 · The dilution surprise
Stack two SAFEs and a priced round on Scenario S. What number will the founders not expect, and which term caused it?
- INV-04 · Thesis from public evidence
What would have to be true for the thesis to be right, and what range do you put on each assumption?
- INV-05 · Build a screen, then try to break it
Build a scoring rubric that claims to pick better companies from public data. Backtest it, then try to break it: how much of its apparent skill is look-ahead or survivorship bias, and what is the honest confidence interval?
- INV-07 · Underwrite a VEN team
Run diligence on a venture team's packet and write an investment-committee memo with a recommendation. Would you back this team, and what would change your mind?
- INV-08 · Fund math with nothing hidden
Model FUND-ACME-I from its terms: fees, carried interest, capital calls, and the J-curve. Why does a careful investor trust DPI before TVPI before IRR?
- INV-09 · Buy, fix, sell, and who pays
Model a buyout of Main Street Mechanical: purchase price, debt, operating changes, and exit. Then answer the question the spreadsheet does not: who carries the risk, and who pays if the plan goes wrong?
- INV-10 · When two systems disagree
Find every gap, say what kind of gap it is, and resolve each one without inventing a number. Which system do you trust, and for which question?
- INV-11 · AI-washing audit
Find "AI-powered" claims in public fund or company marketing and test each one against what the organization's own public filings and documents support. What would an honest disclosure standard require?
- INV-12 · Investment committee
Present a deal to an investment committee, defend it, and take a recorded vote. What are your kill criteria, and what does the dissent say?
Labs (12, in the bank PDF)
Short handouts, 40–90 minutes each, that build the skills the briefs assume. The number labs come with teacher answer keys that match the calculators to the cent; the judgment labs are scored on the rubric.
- L-01 · Who owns this company? — 45 min
- L-02 · The option-pool shuffle — 45 min
- L-03 · When the SAFE converts — 50 min
- L-04 · Kinks in the waterfall — 50 min
- L-05 · Rule of 40 — 40 min
- L-06 · MOIC, IRR, DPI: which one lies? — 60 min
- L-07 · Assumptions are choices — 50 min · AUDIT
- L-08 · The junior analyst — 60 min · USE
- L-09 · A glossary bot that says "I don't know" — 75 min · BUILD
- L-10 · Make the model fill in a form — 90 min · BUILD
- L-11 · The graveyard you can't see — 60 min · AUDIT
- L-12 · Read the risk factors — 75 min · USE + AUDIT
Tools (free, browser-only, no login)
- CapStack waterfall calculator — Who gets paid at a given exit
- CapStack dilution calculator — Priced rounds and option-pool top-ups
- CapStack SAFE calculator — SAFE conversion at a priced round
- CapStack Rule of 40 calculator — Growth plus margin, year by year
Six-dimension rubric
- R1 — Is there one deal and one decision, narrow enough that evidence could settle it?
- R2 — Is every number re-derivable from the case file or a cited public filing?
- R3 — Is there an honest calculator or spreadsheet answer, and what did the AI add that it did not?
- R4 — Are the biases, the AI-specific failures, and the assumptions that would flip the decision named?
- R5 — Would the work survive a skeptical investment-committee member?
- R6 — Who bears the downside — employees, limited partners, common holders — and what did the student refuse to claim?
Feedback
Scope unclear? Numbers off? Tell us here — you do not have to leave an email.