The 195,831 Shares IMI Lost (and the Buyback That Matters More)

In August 2026, IMI plc — a £10 billion British engineering company — announced that it had lost track of 195,831 of its own ordinary shares.

Not stolen. Not diluted away. Just never formally admitted to the London Stock Exchange. These were ordinary shares, carrying ordinary voting rights, issued under employee shares schemes, covering the range from 22 August 2018 up to 24 June 2026. They existed. Their holders could vote. But the company had never filed the paperwork to make them tradable on the market. After an internal review flagged the gap, IMI submitted the belated admission application on 3 August 2026. The shares went live on August 4.

That is weird. It is also, in a way, the point.

Because the odd thing about IMI's share capital right now is not that 195,000 shares were left in regulatory limbo for years. The odd thing is that IMI is spending £500 million to buy back and destroy its own shares while quietly issuing thousands more to employees — and the plumbing that connects these two flows is where the real capital allocation story lives.

What "admitted" actually means

On the London Stock Exchange, not all issued shares are created equal. A share can be legally issued by the company — it exists on the register, it carries a vote, it entitles the holder to dividends — but still not be "admitted to trading." Until admission happens, the share is a paper claim you can't easily sell on the open market. It is economically real but liquidity-dark.

This is a UK listing rule, not an accounting one. Under the UK's Public Offers and Admissions to Trading Regulations, any securities issued by a listed company must be admitted to the Official List and to trading on the LSE. Employee share plans routinely generate new shares, and companies are supposed to admit them as they come into existence. IMI apparently skipped that step for shares dating back to August 2018. Nobody got cheated — the employees held real equity. But the shares weren't fungible on the market until August 4.

The fix was a retroactive admission covering the full range. No prospectus was required (exempted under the regulations). The shares were "Fully fungible with existing Ordinary Shares" going forward. From the shareholder's perspective, nothing changed economically. From a regulatory-plumbing perspective, the company finally closed a gap.

The administrative story is boring. But it leads into a much more interesting mechanical question: if IMI's share base is expanding on one side and shrinking on the other, which flow wins?

The buyback machine

In March 2026, alongside full-year results, IMI announced a £500m share buyback programme, running through 31 December 2026. The shares are purchased for cancellation — not held in treasury, not reissued later. Once cancelled, they cease to exist.

The execution is methodical. Through Deutsche Bank's Deutsche Numis arm (the London arm of Deutsche Bank), IMI buys back tranches on LSE, Turquoise, Chi-X, and BATS — the fragmented venues where London equity liquidity now lives. At the end of August, a batch of 60,108 shares was bought at an average price of 3,079.29 pence per share. On September 2 and 3, two more tranches of 98,300 shares at an average price of 3,044.48 pence plus another batch of 98,600 shares followed on September 3. Small batches, multiple venues, steady pressure.

The scale matters. £500 million at roughly 3,050 pence per share is about 1.6 million shares — roughly 0.65% of the roughly 250 million shares in issue. That is not a round-number buyback that dramatically shrinks the float. But combined with what came before, the cumulative effect is meaningful. In 2025, IMI already spent £200.1m on share buybacks, nearly double the £97.1m in 2024.

The company built the buyback into its earnings guidance. Full-year 2026 adjusted EPS is guided at 136p and 142p, assuming a weighted average number of shares of 238m — which already assumes the £500 million buyback will largely execute. Without the buyback, EPS would be lower. The buyback is not a bonus; it is baked into the number the company is asking investors to believe.

The denominator

Here is where the plumbing becomes useful.

Under UK disclosure rules (the FCA's Disclosure Guidance and Transparency Rules), any shareholder who crosses certain ownership thresholds — 3%, 4%, 5%, and so on — must publicly disclose their holding. The percentage is always calculated against total voting rights: issued shares minus shares held in treasury.

As of September 1, 2026, IMI had 250,047,815 ordinary shares in issue and 11,648,836 in Treasury, for 238,398,979 voting rights — the denominator. As the buyback program continues, this denominator shrinks. As employee shares are issued, it grows. The net effect determines whether a given shareholder's ownership percentage drifts up or down.

On the buyback side, shares are being cancelled at a much higher rate than employee shares are being issued. The 195,831 shares that went through the retroactive admission represent the cumulative total of roughly eight years of employee plan issuances. The buyback program alone will retire more shares in eight months than the employee plans have created in eight years.

So the denominator is going down. Which means every remaining shareholder's slice of the company is getting slightly larger, even if they do nothing. This is the mechanical effect of buybacks — not a story about "value creation" in the abstract, but a literal arithmetic shift in who owns what fraction of the cash flows.

The business behind the shares

All of this only matters if the underlying cash flows are worth owning a bigger fraction of. They are.

IMI is a fluid and motion control engineering company — valves, pumps, actuators, sensors — serving energy, automation, and healthcare markets. It grew revenue to £2,304m in 2025, marking a fifth consecutive year of mid-single digit organic revenue growth. Adjusted operating margin is 20.0%, up from 19.7% the prior year. Return on invested capital is 14.0%, up from 13.4%. Free cash flow was £290m, with 96% cash conversion. Net debt sits at 1.0x, at the lower end of the 1.0x–2.0x target range.

This is not a company with a dramatic growth curve. It is a well-run, diversified, mid-growth industrial with pricing power and high margins. The kind of company that generates steady free cash flow and can afford to return capital to shareholders without jeopardizing investment.

What the shareholder should actually think about

The missed admission of 195,831 employee shares is a regulatory housekeeping item, not a governance red flag. The shares were real, the employees weren't harmed, and the fix took one day. It is the kind of thing that happens when a company has been issuing small batches of shares for years and nobody noticed the paperwork slipped. The interesting story is not the error.

The interesting story is the capital structure being reshaped in real time. IMI is buying back shares at roughly 3,000 pence each, which translates to a P/E of roughly 23x on 2025 adjusted EPS of 132.3p. Is that cheap or expensive? That is a valuation question, not a plumbing one. But the mechanics are clear: the company is using free cash flow to shrink the share base, and it has priced that shrinkage into its own earnings guidance.

If the shares are worth more than the buyback price, the buyback transfers value from the company to remaining shareholders. If they are worth less, it transfers value in the other direction. The plumbing ensures the arithmetic happens. The investor's job is to decide whether £30 per share is a fair price for a 20% margin, 14% ROIC industrial compounder.

The 195,000 shares that were lost for years and then found are a curiosity. But they remind you of something useful: every share that exists or is destroyed changes everyone else's ownership. And at a company like IMI, where management is actively managing the denominator, watching the share count is not an afterthought. It is part of the investment case.