Every modeller knows the 21:00 version of themselves. Dinner is done, the house is quiet, and the laptop is open again because the debt sculpting still doesn't balance and the lender's comments are due Thursday. You tell yourself it's twenty minutes. It's never twenty minutes.
I've written a fair amount on this blog about circular references and DSCRs. This post is about the other thing, the hour when you don't open the file, and a book I wrote for it.
What The Idle Hour is
The Idle Hour: Buildings, objects, rakes and other weekend reading is a book of short pieces, mostly two pages each, meant for the hour after dinner or a quiet Saturday morning. The introduction tells you not to read it properly: open it at random, read for twenty minutes, put it down without guilt.
There's no argument and no framework. There are buildings in New York and London and the slightly ridiculous stories behind them, the origins of things people wear without thinking (loafers, brogues, the navy blazer), dandies and engineers and billionaires, a lot of lists, and a cabinet of numbers, charts and myths at the back. Where a story is legend rather than record, the text says so, and there's a section at the end listing the claims I'd check before repeating them at dinner.
It won't make you a better modeller by Monday. That's rather the point.
The day job leaks in
I couldn't quite leave work alone, so there's a part called Money and Machines. It has the Franciscan friar Luca Pacioli, who in 1494 wrote down the double-entry system the Venetian merchants were already using and so made it possible, five centuries later, to tell whether a company in Johannesburg is solvent. He also shared a house with Leonardo da Vinci, which I mention to anyone who tells me accounting is dull.
There's a piece called The £10 million spreadsheet, which is the one modellers will read with a slightly cold feeling. A dropped minus sign at Fidelity Magellan, the copy-and-paste VaR model behind the London Whale, the Reinhart and Rogoff average that missed five countries off the bottom of the range. It ends with the dull rules nobody follows: separate inputs from calculations, never hard-code a number in a formula, build a check row that should equal zero and actually look at it.
And there are three Excel problems, set for fun. Here's the first:
You have a column of dates in A2 to A500, unsorted, spanning several years. Count how many of them fall on the last Friday of their month. One formula, no helper column, no VBA.
The clean answer is one SUMPRODUCT and a small insight about what "last" means. It's on page 203, followed by a harder one from the day job: somebody has calculated a project's return with IRR on irregular cash flows, and you have to say why the answer is wrong. I've seen a model go through three committees with IRR in it before anyone asked what the dates were.
And then it leaves the office
Most of the book has nothing to do with spreadsheets, which is why I enjoyed writing it. Why the Panama hat isn't from Panama. The holes in your brogues. Brunel, who was right about the railway gauge and lost anyway, because being right isn't enough if you're right after the standard is set. John Harrison, a Lincolnshire carpenter who solved longitude, proved it, and then waited decades to be paid by a board that had bet its reputation on the other answer. (Anyone who has argued with a lender's model auditor will feel for him.) Shackleton, who lost his ship in the Weddell Sea and got all 28 men home.
Then the lists: ten walks that cost nothing, ten drinks a man should be able to make, ten questions that make a dinner, ten rules for email. None of it requires a laptop, and all of it fits in the gap between dinner and bed.
Where to get it
The Idle Hour is the third book in the series after A Stylish Life and Ten Times, and it's out now on Amazon: Search for The Idle Hour on Amazon.
Buy it for yourself, or for the modeller in your team who replies to emails at 23:00. Then close the laptop. The model will still be wrong in the morning, and you'll be in a better state to find out why.