WPP's £2.8 Billion 2020 Write-Down Traced Back to a Deal It Closed in 2000

In the spring of 2000, WPP swallowed Young & Rubicam in a stock swap worth around $5.5 billion — the largest deal advertising had ever seen, the one that made a company that began as a shopping-basket manufacturer¹⁰ the biggest agency group on earth.⁷ There was no five-year earn-out attached, no target the sellers had to hit to get paid. It was a fixed-price takeover, and the difference between what WPP paid and what Y&R's tangible assets were worth simply parked itself on the balance sheet as goodwill. Twenty years later, in the worst quarter the industry had seen in a generation, that number came due — all at once, in a single £2.8 billion charge.⁵

The story everyone tells is that WPP was a machine built on earn-outs — buy an agency cheap, hold back most of the price, pay it out only if the founders keep hitting their numbers, and walk away clean if they don't. That story is mostly true, and it mostly worked. But it describes the wrong transactions. The deals that actually built the empire had no earn-out at all, and the biggest write-down in WPP's history didn't come from an earn-out target missed. It came from the part of the playbook nobody talks about.

The earn-out was the safe part of the machine: a structure that pays only if the numbers come, and quietly refunds you when they don't

Here is what an earn-out actually does, and why it was the disciplined half of Sorrell's playbook. When WPP bolted a smaller agency onto one of its verticals, it recorded the future contingent payments at fair value — the present value of the cash it expected to hand over, but only if the acquired business improved profits in line with the estimate.² The mechanic was in the filings for years: directors set a best estimate of what the sellers would earn, discounted it, and adjusted as reality arrived.³ If the acquired shop underperformed, the liability shrank — and the excess flowed back through profit as a credit. That isn't a hidden landmine. It's a self-correcting instrument. In fiscal 2009, WPP booked £19.4 million of operating-profit credits precisely from releasing excess provisions set up on earlier acquisitions — the earn-out machine handing money back because targets weren't hit.⁴ The risk sat with the seller, who only got paid for performance actually delivered.

The marquee deals were the opposite of an earn-out: the acquisitions that built the scale were paid in full, up front, at the top of the market

The transactions that made WPP famous were not earn-outs at all. J. Walter Thompson fell in a $566 million hostile takeover in 1987; the Ogilvy Group in an $864 million contested deal in 1989; and Young & Rubicam in the roughly $5.5–5.7 billion all-stock swap of May 2000 that crowned WPP the largest advertising group in the world.⁷¹¹ None of these had a five-year holdback tied to the sellers' performance. WPP paid the full price at the peak of each deal's logic, and the premium over hard assets became goodwill — an intangible that just sits there, immune to the self-correction that quietly protected the bolt-on deals. An earn-out shrinks when the business disappoints. Goodwill doesn't shrink. It waits for a test, and then it detonates.

How a deal from 2000 blew up in 2020: goodwill doesn't miss a target — it fails a math test, and the math turned against it all at once

Goodwill is tested, not earned. Every year WPP had to ask whether the future cash the Young & Rubicam assets could throw off still justified the number carried on the books. For two decades the answer was close enough to yes. Then COVID-19 arrived and moved three dials at once: discount rates went up, the 2020 profit base went down, and the industry's expected growth rate fell.⁵ Run those inputs through the impairment model and the value of the acquired businesses collapsed — not because Y&R had failed a performance clause, but because the arithmetic of what its future was worth had changed. WPP recognised £3.1 billion of impairments that year, £2.8 billion of it goodwill, and its own filing states the majority related to businesses acquired as part of the 2000 Young & Rubicam deal.⁵ The warning had been visible for a while. Back in 2018, WPP had already taken a £148.0 million impairment on VMLY&R, the agency it built by merging its own VML into Young & Rubicam⁶ — the same asset, softening early.

Isn't the earn-out label close enough to true?: WPP really did buy hundreds of agencies on earn-outs — but the label points at the wrong risk

The fair objection is that WPP genuinely was an acquisition machine that ran on earn-outs — Sorrell said so himself, the accounting policy for contingent consideration was baked into the group's filings for years, and the bolt-on strategy really did stitch the empire together.¹² All true. And the framing isn't dishonest so much as misaimed: it takes the disciplined, self-correcting part of the playbook and hangs the group's biggest failure on it. The instrument that got the headlines is the one that behaved. The instrument that quietly caused the damage — full-price goodwill on fixed-price takeovers — never got a catchy name. Notice, too, that the earn-out was closely identified with Sorrell personally, not just with WPP as an institution — though at his next venture, S4 Capital, he explicitly broke from that pattern, structuring the 2018 MediaMonks acquisition as an upfront cash-and-equity deal with no earn-out at all.⁹ The playbook wasn't broken. The part of it that carried the risk was just the part nobody was watching.

The final proof came in 2023, when WPP folded J. Walter Thompson, Young & Rubicam, and Wunderman into a single agency called VML — retiring three of the most storied names in advertising. By one inflation-adjusted estimate, WPP had paid roughly $11.5 billion for the brands it was now erasing, more than the entire company was worth on the market.⁸ That is the true shape of the fork. Earn-outs were the deals WPP could afford to be wrong about, because being wrong refunded you. The takeovers were the deals it couldn't — because a fixed price at the top of the market doesn't self-correct. It just waits for the test, and when the test finally comes, it charges you everything at once.