The summer of 1999 was a bad one for reading the back pages if you supported Arsenal, because for most of July and August there was really only one story, and it kept not ending. Nicolas Anelka, twenty years old, PFA Young Player of the Year for the season just finished, a double winner at nineteen, had made it plain he wanted out, and the noise around the story had the particular ugliness of a transfer saga conducted through leaks rather than statements — agents, brothers, "sources," the works. When it finally concluded, on 20 August, the fee was £22.3 million, from Real Madrid, and the general feeling among people who wrote letters to the football weeklies that autumn was that Arsenal had sold the best young forward in England to fund the manager's next experiment. Which, as it turned out, was exactly what had happened, and the experiment was called Thierry Henry.

I don't have a programme for this one, which is a confession worth making before I go any further, because a transfer is not a match and there is no ninety minutes to keep the ticket stub from. What I have instead is the newsprint version of memory, which is nearly as unreliable and rather harder to check. The version that gets retold now, whenever Anelka's name comes up in an "underrated Arsenal careers" thread, is that the club sold a generational talent for money and got lucky. That is not quite what the record shows, though it isn't a lie either — it's a tidying-up, the same tidying every sale of a young player gets once the replacement works out. Anelka had already made himself close to unsellable through his own conduct that summer, refusing to train, refusing to speak to journalists, refusing more or less everything except the move itself. The fee was very good business. Whether Arsenal had much choice by August is a separate question, and the retelling tends to skip it.

What isn't in dispute, because it's in the club's own accounts, is where some of that money went. Arsenal did not spend £22.3 million on one replacement. They spent a good deal less than half of it on Thierry Henry, signed that same window from Juventus, a forward who had not yet worked out what kind of player he was and who Arsène Wenger apparently thought he already knew. The rest of the Anelka fee did not vanish into wages or a director's bonus. Some of it, per the club's own reporting a few years later, went toward £32.3 million in land at Ashburton Grove — the site, eventually, of the stadium Arsenal play in now. A boy who wouldn't speak to journalists financed, in part, both Henry's arrival and the ground under the Emirates before the Emirates existed. That is a better story than "sold a talent, got lucky," and it has the advantage of being true.

It is worth pausing on Real Madrid's side of this, because it undercuts the idea that the big money always knows what it is doing. Anelka did not settle in Spain — he went five months without a league goal for the champions of Europe, to the point where the tabloid Marca ran the headline "Anelka finally scores a goal… on a video game" after he'd been photographed playing FIFA with their staff. He won the Champions League with Real Madrid that season regardless, scored twice in the semi-final against Bayern, and was gone to Paris within weeks of lifting it. And years later, Arsenal's own annual report revealed that Madrid had missed an instalment on the transfer fee altogether — a club that would go on to spend £37 million on Luis Figo and £45 million on Zinedine Zidane in successive summers, while reportedly selling its own training ground to manage the debt. The moral isn't that Madrid were fools. It's that a fee is only ever half the story; the discipline is in what a selling club does with the money, and Arsenal in 1999, whether by foresight or necessity, did something with theirs.

The fair case against all this is not hard to make, and it was made loudly enough at the time. Selling your best young forward looks like defeatism dressed up as strategy, and supporters who watched Anelka go were not comforted by the promise of prudent reinvestment — they wanted the player. Every sale carries the risk that the replacement doesn't arrive, or arrives and doesn't work, and nobody in August 1999 could have told you with any confidence that the unproven Frenchman from Juventus would outscore the one just sold to European champions. Selling well is not a repeatable formula just because it worked once. Henry is the exception that gets quoted precisely because he is exceptional; for every Anelka-into-Henry there is a sale that simply weakens the team and buys nothing worth having.

Which is roughly the argument being had in the Arsenal boardroom, in various coded forms, this summer. The vocabulary has changed — nobody in 1999 said "PSR compliance" — but the underlying calculation hasn't: that a club aiming to compete at the top has to fund its ambitions partly by selling shrewdly rather than simply spending, and that the trick is knowing which sale is Anelka-to-Henry and which is just a hole in the squad with a good press release attached. The current talk of "improving the selling strategy" is the 2026 dialect of a sentence that could have been written in the Highbury boardroom in 1999: we need the money to become somebody else's better player. Arsenal have been here before, and the archive's verdict on it is not romantic. It is that the sale looked like weakness in August and read as method by May, and that the gap between those two readings is where the actual skill lives — not in refusing to sell, and not in selling for its own sake, but in knowing, at the moment the letters pages are furious, whether you are funding a stadium or just balancing a book.