Companion files
A Practitioner’s Guide to Sourcing, Diligence, Deals and Portfolio Work
These are the companion files of The Venture Capital Associate by Julian R. Sterling.The book on Amazon →
Two Excel workbooks, now corrected. One runs the reserve arithmetic the book says you should be able to run in your head — and then never runs. The other computes the unit economics of chapter 6, a chapter about whether a business works that contains two numbers. Around them: a blank set for your own fund and companies, twelve working documents to print, forty questions that mark themselves, and three cases the book names and never takes to a number. All free. Nothing is gated behind a sign-up, and no email address is asked for.
Two workbooks, the blank set, the three cases, the question book, the printable documents and the read-me. Each group below can also be downloaded on its own. Last revised 24 September 2026.
Reading the paperback or the Kindle edition? The same book is also sold as a DRM-free PDF: searchable, printable, yours on any device. $19.99, instant download.Buy the PDF edition → Free sample
Finished the book? The Private Equity Career Pack: five PDF editions, from buyouts to venture capital, this book included, $59.99. See the pack →
The working models behind the books, built out for live deals: the Pro models →
Optional. When a workbook on this page is corrected or a new worked case goes up, readers on the list get one email, plus a short note when a new book comes out. The downloads do not need it.
Join the reader listRun by MailerLite, off this site. How your address is used.
These carry the book’s own figures. Both were corrected on 24 September 2026; what changed is at the end of the read-me inside and at the foot of this page.
Appendix E · Chapter 12
Chapter 12 is the longest in the book. It says that “underneath every reserve decision sits arithmetic that you should be able to run in your head” — and prints figures — forty companies, a stake of twelve per cent that ends near five — and runs none of the arithmetic beneath them. This file runs it.
Start with the number that reframes the chapter. Holding a ten per cent stake through Series A to D costs 14.76 million in one company. A 40 million reserve pool therefore defends 2.71 companies out of forty. The book describes firms with “a policy of taking their full pro-rata in nearly every up round”; at this fund shape that policy costs 14.8 times the pool. It is not a policy. It is an impossibility.
Which settles what the follow-on framework is really for. Ranking companies against each other is not a counsel of rigour — it is a binding constraint. Recommend a follow-on and you are arithmetically recommending that two other companies get nothing. And note where the money goes: sixty-one per cent of the cost falls in the Series D alone, which is why reserve pools run dry late, and why pro-rata taken cheaply at the A can be the reason there is nothing left for the round that mattered.
The book’s own claims hold up, with one condition. Ten per cent at seed becomes 3.70 per cent at exit if you never follow on — its “double-digit to low single digits”, measured. Eight per cent of a two-billion outcome is worth 1.60× the fund and two per cent 0.40×: its “difference between returning capital and returning several times capital” needs a landmark of at least fifty times the fund.
But one convention does not survive being computed. The book says funds reserve “a roughly even split, about half”. On this fund shape half returns 2.47× and a quarter returns 3.14× — a gap of 67 million on a 100 million fund. Turn the selection-skill dial down and the optimum stays at a quarter until skill falls below about 0.37, then drops to nothing at all. It never rises to a half, and it never lands in between: the curve has one corner, the ratio at which the pool defends every fund-returner. Corrected: the first version said a fifth at half skill and a twentieth at a quarter; the gap against the half runs from 59 to 71 million, not 63 to 82. The direction is the opposite of intuition: the worse you are at picking follow-ons, the less you should reserve, because a fund that reserves half and chooses badly has arranged to make its mistakes with the larger pile of money.
One more correction, and it changes an argument. The average reserve dollar, concentrated, returns 3.35. But the dollar a fund holds back “just in case” is the last one, and on this fund it returns 1.08: the first 25.2 million can be held back at eight cents a dollar, less than the fee. The 26th million comes out of the fund-returner’s defence, at 6.22. The column once labelled TVPI is now “Multiple of the fund”, since an unspent reserve is never paid in. Thirty-six checks.
Download the workbook31 KBAppendix F · Chapter 6
Chapter 6 asks whether “a single unit of it, typically a single customer, generates more value than it costs to acquire and serve”. It runs to some two thousand four hundred words and contains two numbers: the chapter number, and a stray ten. This file supplies the arithmetic, using only the chapter’s own example and its own claims.
The chapter describes two companies, both doubling revenue, one whose cohorts expand and one whose cohorts decay, and stops one step short of computing them. Computed: the declining company must buy 1.53 times as much new revenue every year to print the identical growth rate. Nothing in the top line shows it.
Then the question the chapter cannot answer as posed. It says a customer’s value depends on how long they stay, and never names a horizon — but for a cohort that expands the sum has no limit. The same customer is worth 2.78 over three years and 81.95 over twenty. The decaying company’s converges on 2.67 and goes no further. Choose three years and one is worth 1.59 times the other; choose twenty and it is 30.76. Both are defensible, and they are nineteen times apart.
Which makes the threshold exact. Growth adds value below an acquisition cost of 10.98 for one company and 2.51 for the other: a margin for error 4.37 times narrower, on a business that must also buy 1.53 times more. Its exposure is 6.68 times larger at an identical headline growth rate.
And the chapter’s warning that broken economics hide behind fast growth “for a surprisingly long time” gets an answer that is not a duration. At a common acquisition cost both companies run a cumulative deficit every single year — −3,020 against −5,720 — because anything doubling spends faster than margin arrives. The accounts do separate them — by size: the second company burns 1.89 times as much for the same revenue. What they never show is which one is broken. Larkspur’s accounts turn positive only below an acquisition cost of 1.80, the other’s below 1.21; across the whole band where one destroys value and the other does not, both are in deficit. Corrected: the first version said the accounts never separate them.
So the test the chapter proposes and never performs is the one that works. Cut acquisition to zero and the healthy company still grows, reaching 1.75 times its revenue in four years without winning a single customer. The other falls to 0.24 times: it loses 76 per cent of its revenue, having done nothing wrong except stop buying. Forty-two checks, now including every figure Appendix F prints.
Download the workbook20 KBOpen sheet 5 and set selection skill to 0.50 — a fund whose follow-ons land on the right companies half the time and at random otherwise. The optimal reserve ratio is still a quarter, and the conventional half returns 1.86×. Set it to 0.30 and the optimum drops to nothing. Put skill back to 1 and push the fund-returner’s exit value up: the optimum does not move. Change the number of fund-returners instead, two in forty rather than one (take it from the zeros), and it jumps to 40 per cent. The right reserve is the one that defends every fund-returner: it is set by how many there are and what defending one costs, not by how large they become. That is the whole of reserve strategy on this model, and it is two inputs away. Corrected: the first version of this paragraph said a fifth, 1.81×, and an optimum that climbs with the exit value.
It cannot tell you which companies are the winners, which is the entire difficulty and the reason the skill dial exists. It cannot price signalling — the chapter is right that declining a pro-rata can damage a company you still own, and no model here captures that. And it assumes reserves can be deployed when you want them, which recycling and fund-life limits complicate. What it does settle is the size of the constraint you are working inside, which is the thing most easily left vague and most expensive to leave vague.
Two workbooks with every input emptied. The checks sheets count what is still missing and test what holds for any inputs.
Chapters 6 and 12 · Appendices E and F
The two workbooks with their inputs emptied: your fund, its fee, life and reserve ratio, your rounds and pool top-ups, your outcome set and your selection skill; and two companies of your own, with their retention, growth, margin, acquisition cost and the horizon you allow. The grids (reserve ratios, skill levels, horizons) are steps and are kept.
The book’s figures are listed in a text file inside the archive. Type them back in and both files return the book’s results, every check at OK.
Download the blank set44 KBTwelve pages to print and use. A4, with margins wide enough for US Letter.
Chapters 2 to 13 · Appendices B and D
The fund you work for, with the exit that returns it. The sourcing engine and its monthly measures. Founders and the team, evidence against impression, with the bias check. Market and product, sized from the bottom up. The business model and unit economics, with the horizon named and the cut test. Reading the term sheet, term by term and across three exits. The cap table and the conversion, headline against reality. The return model with the fund beside it. The memo and the pre-mortem. The diligence checklist of Appendix B. The ranked book and the follow-on. The first ninety days and the decision journal.
Download the documents34 KBA workbook that marks itself. Answer in the yellow cells; the marking fills in as you go.
All parts of the book
Seven on the fund and the role, five on sourcing and diligence, seven on terms, cap tables and returns, eight on the memo, the portfolio and reserves, seven on unit economics, and six on the three cases below. Numbers are marked to a tolerance, multiple choice to the letter, and the score sheet counts your errors part by part: that is your revision list.
Nothing is hidden. The marking sheet holds every answer, every tolerance, and the reason for each answer.
Download the question book17 KBNew material, not a restatement. Three situations the chapters name and never take to a number. Each case is one workbook: the note is the first sheet, the model is the rest, and a checks sheet closes it. Type your own figures in and every finding recomputes.
Chapters 7 and 8 · Case one
Chapter 8 says a founder who owns forty per cent “may receive far less than forty percent of a modest sale”. On a stack of three 1x non-participating rounds holding 42 of preferences, the founder receives 12.31 of a sale of 60, 20.5 per cent, and nothing at 30. “Modest” ends at 200, 4.76 times the capital invested, where the last series converts. Make the Series B participating and the gap never closes: 88 instead of 100 at a sale of 250.
Chapters 7 and 8 · Case two
Three SAFEs, a pool shuffle and a Series A sold as “20 per cent at 24 million”. The founders end at 48.96 per cent, not the 72 the headline suggests, and the shares that existed were priced at an effective pre-money of 16.32 million. The discounts only matter below 22.125 million, and any pool above 10.56 per cent of the post-money takes the founders below half of their own company.
Chapters 2 and 9 · Case three
Chapter 9’s ten-times success case is exactly 10.24× and returns a fund of 204.8 at most. In a fund of 200 (ADDED), the firm could have paid 2.4 per cent more and no more; each unit added to the entry price costs 2.05 of exit proceeds. And the chapter’s choice between a safe and a risky company holds once the risky one’s “real chance” of 100× passes 1.8 per cent.
Download the three cases41 KB| Amber fill | an input — you may edit these |
| Grey fill | a formula — do not overtype these |
| Blue on pale fill | an input in the reserve workbook |
| Checks sheet | thirty-six controls in the reserve workbook and forty-two in the unit-economics workbook, each stating its own verdict; every other workbook ends with its own |
| ADDED | a figure the book does not give, used in the cases and marked wherever it appears |
Every figure is illustrative, as everything in the book is. The outcome set — one company at two billion, two at 300 million, four at 75, eight at 20, twenty-five at nothing — is a power law of a particular steepness, and the optimal reserve ratio is sensitive to it. It is an input. What is not sensitive, across everything tested, is the direction of the answer.
The workbooks open in Microsoft Excel, LibreOffice Calc, Google Sheets and Numbers; the PDF prints on A4 or US Letter. They use no macros and no add-ins, so nothing needs to be enabled or trusted. If your spreadsheet asks to update links on opening, decline — there are none.
| 24 Sep 2026 | Both workbooks corrected. Reserve arithmetic: the selection-skill dial now does what its label says, and Appendix E’s reading of it was wrong — the best reserve stays at a quarter at half skill (not a fifth) and falls to nothing at a quarter (not a twentieth); the gap against the half is 59 to 71 million, not 63 to 82; the best ratio returns 3.14×, not the 3.16 the read-me printed; the last reserve dollar returns 1.08, not the 3.35 average; “TVPI” renamed “Multiple of the fund”; a typed-in budget is now a formula. Thirty-six checks. Unit economics: the accounts separate the two companies by size (1.89 times), not never; a dead input removed; every figure Appendix F prints is now checked, forty-two checks. Added: a blank set, twelve working documents, forty questions that mark themselves and three new cases. Each workbook can now also be downloaded on its own. |
9 of the calculations in this workbook, worked out in full. All articles.
More in private equity and funds, with free companion files. The full list of titles is on the author page.
Also available as a standalone free template: LTV and CAC unit economics model.
The Venture Capital Associate develops the complete framework behind these models, with the assumptions, trade-offs and worked examples tied together.
View the book on AmazonAlso as an ebook on Apple Books · Google Play · DRM-free PDF edition ($19.99).
These files accompany The Venture Capital Associate. The book is on Amazon.
If this book helped — or didn’t — a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.