A Practitioner’s Guide to Sourcing, Diligence, Deals and Portfolio Work
Julian R. Sterling
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.
Get an email if these files are corrected
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.
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
The reserve arithmetic
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.
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.
Open 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.
What the arithmetic does not settle
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.
The same models, for your own fund
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
A blank reserve model and a blank unit-economics model
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.
Twelve pages to print and use. A4, with margins wide enough for US Letter.
Chapters 2 to 13 · Appendices B and D
Twelve working documents
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.
A workbook that marks itself. Answer in the yellow cells; the marking fills in as you go.
All parts of the book
Forty questions on 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.
New 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
What a forty per cent founder receives
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
What the founders actually sold
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
Can this one return the fund?
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.
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.
Opening the files
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.
What changed, and when
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.
Articles on this book
9 of the calculations in this workbook, worked out in full.
All articles.
How do you value a startup with the VC method?Post-money = exit value x retention / target multiple. A 200m exit, 10x target and 32 per cent later dilution price a 5m Series A at 8.6m pre-money.
How do you calculate a startup's cash runway?Cash divided by net burn gives 18.0 months on 9,000,000 at 500,000 a month, but with costs growing faster in money than revenue the cash runs out at 15.7.
How do you calculate TAM, SAM and SOM bottom-up?Count buyers times annual contract value, restrict to who you can serve, build SOM from the sales plan: $172.8m TAM, $69.1m SAM and $9.33m SOM in a worked case.
The other books with companion files. The full list of titles is on the
author page.
How to Read a Commercial LeaseThe three refinements chapter 19 names and never performs, and the renewal rate below which the mark-to-market is worth nothing.
How to Read a Real Estate Loan AgreementThe cure ratio in closed form, the four-point window in which the cheap cure works, and the cure sized to the wrong threshold.
The CBAM Compliance HandbookWhat 2027 really costs, how big the buffer should be, and the term the surrender formula counts twice.
Private Markets PerformanceThirty-one of the thirty-three figures chapter 19 publishes reproduce exactly — and the two that do not are named rather than quietly adopted.
The Private Equity Fund Controller PlaybookThe book defines IRR, DPI, RVPI and TVPI, tells you to update them at the exit, and prints not one value. Computed: a 1.833× deal inside a fund at 0.892 TVPI.
The Private Credit InvestorThe two coverage ratios are not measured on the same thing: the erosion is 47.7 per cent, not the 28.7 the headline implies.
The Distressed Debt InvestorWhere the fulcrum security actually breaks, and what an 80-cent recovery is worth once composition and time are priced.
The Real Estate Debt InvestorThe margin against the return on capital, the floor priced, and what prepayment protection buys.
The Fund Finance ProfessionalChapter 8 builds the reported-to-eligible NAV bridge; chapter 9 computes every ratio without it. Two points at every state — and what a subscription line does to the IRR.
Office Real EstateA six per cent yield that returns 3.2 per cent once the re-letting cycle is paid for, and the headline-to-net-effective rent arithmetic.
Real Estate FinanceFour people look at one building and reach four numbers; the lender is whole only above 105,109,489, twelve per cent below today’s value rather than forty.
Retail Real EstateThe occupancy cost of every unit in a centre, the sixteen per cent of the rent roll no tenant can sustain, and the right-size-convert-or-hold decision priced.
Sale and LeasebackA €179.5 million transaction end to end, with rent cover measured on the entity that actually signs the lease.
Self-Storage Real EstateThe cohort engine behind a 590-unit store, and the rate increase on existing customers priced against the move-outs it causes.
Financial Risk ManagementA fund inside every limit that cannot meet a redemption — and the number that decides it is the one with no currency attached.
Business ValuationThree advisers land 26.8 per cent apart on one company, and the whole gap turns out to be 1.96 points of perpetual growth.
Quantitative FinanceThree models agree to a quarter of one per cent about a number that one unobservable input moves a hundred and three times as much.
Asset ManagementFour people quote four returns for one mandate, all correct and 2.7017 points apart — forty-eight times the manager’s net skill.
Alternative InvestmentsA manager reports 13.29 per cent and the endowment earns 6.26 — both correct, and only a third of the advertised advantage arrives.
Credit AnalysisFour defensible EBITDAs on one borrower give leverage from 3.19x to 6.47x — and the add-back argument is fifty times the covenant headroom.
Venture CapitalOne company out of twenty-eight returns 56.7 per cent of the fund, and half the capital goes in after the decision — at half the return.
Machine Learning for FinanceFive people quote the accuracy of one credit model, all five are right, and the number that decides how much money it makes is none of them.
Commercial Real Estate InvestingThe equity earned 8.6647 per cent and the investor received exactly 8.0000 — the preferred return, and nothing above it.
Mergers and AcquisitionsThe board paper says the deal creates 13,436,667 of value. The arithmetic says it destroys 17,530,855. Nobody is lying.
DerivativesThe treasury report says the hedge cost 1,233,698. That is the interest differential, not a cost.
Treasury ManagementFive cash balances for one company, all correct and 145,600,000 apart — and the revolver that is two-thirds of the liquidity leaves at a revenue fall of 8.4127 per cent.
Financial Planning and AnalysisRevenue 3.0190 per cent above budget and operating profit 16.3209 per cent below it, in the same quarter, with every figure correctly stated.
Energy TradingA position report that is 91.7031 per cent hedged and correctly computed, on a book that is short 2,542,000 MWh — and a margin call of 198,400,000 the next morning.
Construction Cost ControlA contract sum of 26,301,102 became a final account of 29,153,363 on the building that was drawn — and 85.8 per cent of what was lost was knowable on the day it was signed.
Trade FinanceSix routes to payment on one 4,200,000 export order cost between 178,040 and 223,268, a spread worth 14.8 per cent of the margin, and a day of buyer credit costs 1,031.76.
Cost AccountingOne factory costed twice on the same 13,440,000 of overhead, and 4,053,091 moves between four product families.
Pricing StrategyA list price of 148.00, a pocket price of 112.51, and the nine deductions in between — with what one point of price is actually worth.
Contract ManagementA bid 5.50 a unit cheaper, signed on paper worth 9.81 a unit more — and 94.68 per cent of it was knowable on the day of signature.
Capital MarketsFour defensible readings of one bond, the cost of an issue under each convention, and the placement discount computed at a constant discount.
Financial RegulationThree ratios pass and the fourth is in breach by 1,956.0 million: the output floor, the liquidity buffer after encumbrance, and the loan split.
Bank ManagementFour defensible readings of one balance sheet, the deposit beta measured two ways, and the liquidity that lasts 3.9 days.
The Real Estate WorkoutFour ways out of one broken loan, discounted to today, and the 58.7 cents at which a payoff beats enforcement.
Go beyond the companion files
The Venture Capital Associate develops the complete framework behind these models, with the assumptions, trade-offs and worked examples tied together.
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.