Companion files
A Practitioner’s Guide to Measuring What Actually Breaks a Fund
These are the companion files of Financial Risk Management by Julian R. Sterling.The book on Amazon →
The four Excel workbooks that go with the book, corrected where they were wrong. Every figure the book prints is reproduced in them by a live formula rather than a typed constant, and each workbook ends with a Checks sheet setting the printed figure beside the computed one: 96 controls in all, every one green. Around them: the same four models with the inputs emptied, thirteen working documents to print, forty questions that mark themselves, and three cases the book names and never takes to a number. Nothing is locked, protected or watermarked.
Free to download. No sign-up, no email address, nothing to fill in.
Eleven workbooks, 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? Two or more PDF editions bought together are 15% off. Browse the PDF editions →
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, so that every number it prints can be traced to a live formula, and each closes on its checks sheet. Download these four82 KB
Chapters 4, 9, 12, 13 and 14
The four numbers the book opens with, computed on one set of inputs so the gaps between them move when you change an assumption rather than staying where the book left them. A one-day value at risk of 7,242,456, which is 0.862 per cent of net assets. A stress loss of 93,277,800, which is 11.10 per cent. A cost of getting out of 5,101,785 or 2,016,532, depending on a decision nobody has written down. And a transfer of 0.8098 per unit to the investors who stay.
Then the fifth number, which is not denominated in money and therefore appears in no risk report: what the fund can still sell inside a week. It reads 31.2 per cent before the redemption and 8.3 per cent after the cheap way of meeting it. The impact coefficients sit on the Inputs sheet with a switch that halves, doubles and triples them — being wrong by a factor of three moves the answer 44 per cent, and the choice of selling policy moves it 153. The parameter is the one everybody argues about, because it is the one with a formula. Revised: the switch this page described is now in the workbook, a scale beside the coefficients with the four settings side by side; tripled, the vertical slice costs 7,360,306 against 5,101,785.
Download the workbook26 KBChapters 4, 6 and 7
The parametric construction worked to 7,242,456 one visible step at a time: sleeve volatilities, the correlation grid, the quadratic form, the daily scaling and the confidence multiplier, each on its own row. Expected shortfall sits beside it. The correlation matrix has a switch between the reported values and the crisis values, and the diversification benefit — 0.767 of a point of volatility, 11.5 per cent — falls to 27.6 per cent of itself when you throw it.
The horizon block scales one day to twenty by the square root of time, to 32,389,246, and puts beside it the objection of Chapter 7: the scaling assumes you can trade, and the book that produced this number takes eleven days on average to sell. The last sheet is the exception counter, with the green, amber and red zones drawn at their conventional thresholds, so a year of breaches can be typed in and read rather than argued about. Revised: the correlation switch is now in the workbook, with both matrices computed side by side; and the zones are drawn by the binomial rather than typed, so on 500 observations the green zone ends at 8 exceptions, not 9. On 250 the same formulas return the published 4 and 9.
Download the workbook18 KBChapters 9 and 16
Spread duration by sleeve against a shock you set in basis points. On the book’s calibration — 120, 450, 300 and 600 — the four sleeves lose 18,748,800, 50,274,000, 1,575,000 and 22,680,000, a total of 93,277,800, and the net asset value per unit falls from 100.00 to 88.8955. On your calibration it totals to whatever your calibration says.
The scenario does not stop at the mark-to-market, which is the argument of the chapter. The financing sheet carries the repo, the haircut moving from 5 to 15 per cent, and the collateral call of 14,860,681 that arrives on the same morning as the redemption and competes with it for the same cash. And the reverse stress runs the engine backwards: rather than asking what a scenario costs, it solves for the redemption the fund could not have met. Revised: the financing sheet was only in workbook 1; it is now in this one, and met from the same sleeve the call lowers the largest redemption the fund can meet inside a week from 31.2 per cent of net assets to 29.2.
Download the workbook15 KBChapters 12, 13, 14 and 15
The liquidation engine, position by position, with the half-spread and the market impact priced separately and the impact rising with the fraction of the sleeve you sell. The two selling policies run side by side on the same redemption: the vertical slice costs 5,101,785 and leaves one-week liquidity untouched at 31.2 per cent; liquid-first costs 2,016,532 and leaves 8.3 per cent.
Six successive rounds are laid out beneath, with the cumulative cost of each policy and the liquidity remaining after each. The cheap policy stays cheap throughout — 4,753,970 against 8,292,048 over the six — and that is the finding the book had to be rewritten around: there is no crossover. The expensive policy never catches up in money, because what the cheap one spends is not money. The last sheets carry the queue, the first-mover advantage of 0.2611 per unit, and the swing factor each policy implies — 2.733 per cent against 1.080 — which is the same choice seen from the board paper. Revised: the first-mover advantage is 0.2611 per unit, not 0.0804. The queue carried the second and third redemptions forward from the portfolio before costs, so each earlier exit’s cost vanished from the next net asset value. The selling-policy choice of 0.4897 is still the larger, but 1.88 times the queue, not six.
Download the workbook25 KBThe four workbooks with their 151 inputs emptied. The checks sheets count what is still missing and test what holds for any inputs.
All chapters
The four numbers, value at risk and the backtest, the stress book and the liquidation engine, with the same formulas and every input emptied: the fund, the four sleeves, both correlation matrices, the spread shocks, the impact coefficients, the financing and the redemptions. The impact scale, the correlation switch and the horizon rows stay, because they are choices and the axes of tables, not data.
The book’s 151 inputs are listed in a text file inside the archive. Type them back in and the book’s figures come back with them: 7,242,456 of value at risk, 93,277,800 of stress loss, 4,753,970 against 8,292,048 over six redemptions.
Download the blank set76 KBThirteen pages to print and use. A4, with margins wide enough for US Letter.
Chapters 4 to 16
The four numbers and the fifth on one page, with Halverton’s figures beside a blank column. Value at risk one visible step at a time. The correlation memo, reported against crisis. The backtest log, with the zones for 250 and 500 observations and the lines a count leaves out. The horizon note. The three outcome levels to write down before a stress is run. The scenario calibration sheet. The financing lines and the call that lands the same morning. The reverse stress worksheet. The selling-policy decision record. The cost of the exit, sleeve by sleeve, and who pays it. The swing pricing memo and the queue. The liquidity report, after the next decision.
Download the PDF36 KBA workbook that marks itself. Answer in the yellow cells; the marking fills in as you go.
All parts of the book
Seven on value at risk and correlation, six on the backtest and the horizon, seven on the stress and its financing, six on the exit and the transfer, eight on liquidity and the selling policy, 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.
Chapter 7 · Case one
Chapter 7 objects that root-of-time scaling assumes you can trade, and that the book takes eleven days on average to sell. Given each sleeve its own days to sell, the value at risk is 24,029,172: the equivalent of 11.0 days, which the twenty-day 32,389,246 already covers. Sold in equal daily slices (ADDED) it is 14,978,648. Add the stressed half-spread on the whole book and the answer still reaches only 45.3 per cent of the stress loss. The horizon is not what is missing; the tail and the cost of exit are.
Chapter 6 · Case two
Six exceptions in 500 days, spread over two years or with five in one week (the dates are ADDED). The count test gives both a p-value of 0.66 and both are green. The independence test gives the spread pattern 0.00 and the clustered one 31.27, a p-value of 0.00000002. A working model shows five in a given week once in ten thousand million. The report should carry the largest count in any five days beside the total.
Chapters 9, 14 and 16 · Case three
Liquid-first cannot meet the fourth 10 per cent redemption inside a week: 30,889,484 of seven-day assets, 14,860,681 of them posted against the repo, against 54,436,048. The vertical slice meets all six. A floor at the next redemption plus the call (ADDED) meets all six too, for 6,795,363 over the six rounds: it keeps 42.3 per cent of the saving of liquid-first and none of its missed redemptions.
Download the three cases49 KB| Blue text | a hardcoded input — you may edit these |
| Yellow fill | an input cell; everything else on the sheet is a formula |
| Black text | a formula — do not overtype these |
| Checks sheet | the printed figure beside the computed one, with a PASS or a FAIL; in the blank set and the cases, each control states its own verdict and the file ends ALL OK |
| ADDED | a figure the book does not give, used in the cases and marked wherever it appears |
A workbook that agrees with a book proves nothing on its own — the author wrote both. What the Checks sheets do is different: they force the model to reproduce a number that was printed before the model existed, from a formula rather than from the number itself.
On this book the exercise did more than catch arithmetic. The six-round engine of workbook 4 was built to locate the point at which the cheap selling policy becomes the expensive one — the crossover the manuscript had asserted. There is no crossover. The chapter was rewritten around what the model actually says, which is a harder and more useful claim: the cheap policy is cheaper in every round, and what it spends instead is the fund’s ability to meet the next redemption at all.
Where a printed figure depends on a convention rather than on arithmetic — the confidence multiplier at 99 per cent, the trading-day count, the assumed severity in the stress — the workbook shows the convention as an input rather than burying it in a formula.
Financial Risk Management develops the complete framework behind these models, with the assumptions, trade-offs and worked examples tied together.
View the book on Amazon Read the related analysisAlso as an ebook on Apple Books · Google Play.
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 | The four workbooks corrected, 96 checks. The first-mover advantage of workbook 4 is 0.2611 per unit, not 0.0804; on 500 observations the green zone ends at 8 exceptions, not 9; the impact scale, the correlation switch and the financing sheet this page described are now in the workbooks, with the swing factors in workbook 4. Every other figure on the page was re-verified. Added: a blank set, thirteen working documents, forty questions that mark themselves, and three cases. |
More in corporate finance, with free companion files. The full list of titles is on the author page.
These files accompany Financial Risk Management. 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.
Also available as a standalone free template: VaR and expected shortfall model.