Melrose After Garden Grove: What the $100m Does Not Cover

Melrose Industries told the market on 25 August that the Orange County District Attorney's Office had closed its criminal investigation into the May incident at GKN Aerospace's Garden Grove plant, with the company's filing stating that no criminal charges will be filed. Alongside it came a $100m claims programme — about £73m — for residents and businesses caught in the evacuation, and a target of 28 September to resume full manufacturing. The shares rose around 8%, among the strongest FTSE 100 performers of the session.

That is the good news, and it is real. But the $100m is a capped number attached to one part of the liability, and the release is explicit that it is not the whole bill: claims from emergency services agencies and any civil enforcement penalties are handled separately from the programme. Meanwhile the company's own £175m buyback remains paused after only £12m was executed, and full-year guidance is still framed as excluding the Garden Grove impact. On the evidence Melrose has published, the criminal question is closed and the financial question is not.

What actually happened

On the afternoon of 21 May 2026, personnel at GKN Aerospace's Garden Grove site in California noticed that a storage tank holding methyl methacrylate — a chemical used to make high-strength aerospace acrylic materials — was rising in temperature. The compound releases energy exothermically; contained, that can build pressure to the point of rupture. Officials ordered an evacuation, and roughly 50,000 residents left their homes for several days. Production stopped.

The site matters more than its size suggests. Operated by GKN Aerospace Transparency Systems, Garden Grove designs and manufactures transparencies — canopies, windshields and windows — for military and commercial aircraft. It employs around 500 people and generated £136m of sales in the 2025 financial year. Those are specialised parts with qualified tooling and long requalification cycles, which is why an outage at a site representing under 4% of group revenue has held attention for three months.

Monday's announcement did three things. It closed the criminal exposure. It put a ceiling of $100m on the community claims programme, which opens in the coming weeks and runs into 2027, covering hotel stays, meals, transport, lost wages and loss of use. And it set 28 September as the target for resuming full manufacturing, with a phased return to full capacity and customer deliveries running through the fourth quarter.

None of this was the first disclosure of a financial impact. At the half-year results on 31 July, Melrose had already quantified what the outage cost: £16m of revenue and £9m of adjusted operating profit in the first half, with a further £25m to £30m of exceptional costs expected in the second half. It also paused the £175m buyback after £12m, explicitly pending clarity on Garden Grove.

Why the market reacted the way it did

An 8% move in a FTSE 100 industrial is a large one, and it was not really about the $100m. It was about the removal of a tail.

A live criminal investigation into a US industrial incident is the kind of overhang that resists modelling. It carries an unknown magnitude, an unknown timeline and — critically for an aerospace supplier — a potential effect on customer qualification and government contracting that has nothing to do with the size of any fine. Investors cannot discount it sensibly, so they tend to apply a blanket risk premium instead. The DA's decision removed that premium in one line.

The restart date did similar work in a different register. Since late May the market has had no anchor for when Garden Grove produces again. A dated target converts an open-ended outage into a bounded one, which is what allows the second-half numbers to be modelled at all.

What the market appears to have done, though, is treat the resolution of the criminal question as resolution of the whole question. The release does not support that reading, and the company's own capital allocation does not either. If management believed the liability were now bounded at $100m, the natural signal would be to restart the buyback. It has not.

The Openbook read

Our five-factor framework splits cleanly here, and the split is the point: this is a case where Momentum and Reward/Risk are pulling in opposite directions on the same day's news.

Momentum improves, and legitimately so. An 8% relief rally on the removal of a genuine binary risk is not a sentiment blip; it reflects a real reduction in the distribution of outcomes. The underlying operating trend supports it too — first-half revenue rose 10% on a constant-currency basis to £1,878m, adjusted operating profit rose 16% to £347m, and the interim margin reached 18.5%, up from 18.0% for the 2025 full year — with engines and aftermarket leading the growth. Momentum is being earned by the business, not only by the news.

Growth is the factor least disturbed by any of this. Full-year 2026 guidance — revenue of £3.75bn to £3.95bn and adjusted operating profit of £700m to £750m — was reiterated at the half year, and the aftermarket and engines demand driving it is unrelated to a transparencies plant in California. Against 2025's £3,589m of revenue and £647m of adjusted operating profit, the guided range implies continued double-digit profit growth. Garden Grove is a subtraction from that, not a challenge to it.

Profitability holds up better than the headlines imply, and the arithmetic is worth doing. A full quarter of lost Garden Grove output on a £136m annual revenue base is roughly £34m of revenue. At divisional margins, the profit effect is a fraction of that — consistent with the £9m of adjusted operating profit the first-half stoppage actually cost. The exceptional costs are the larger number, and they sit below the adjusted line. The site is material to its customers and immaterial to group margin, and those two facts are not in conflict.

Solvency is where the uncapped tail lands, and it is also where Melrose has the most cover. The company generated £125m of free cash flow in 2025 and guided to underlying cash flow of £150m–£200m for 2026, with a stated path toward £600m by 2029. A $100m programme spread across the remainder of 2026 and into 2027 is affordable against that. The question is not whether Melrose can pay $100m. It is what the separately-handled items add — emergency services claims from a 50,000-person evacuation, and civil enforcement penalties from agencies that have not yet concluded their own processes. Neither has a stated ceiling or a stated date.

Reward/Risk is therefore the factor that has to arbitrate, and the paused buyback is the cleanest evidence available. Management holds better information about the residual liability than the market does, and it chose to keep £163m of authorised repurchase capacity unspent rather than deploy it into an 8% rally. That is a considered signal about how bounded the remaining exposure actually is. Buying a relief rally on a partial resolution, while the company that knows most about the exposure declines to buy its own shares, is a specific and identifiable asymmetry. Readers can compare the factor profile against aerospace peers on the Melrose equity page and the Openbook screener.

The read-across

The narrowest read-across is the supply chain. Aircraft transparencies are single-source, qualified components; a three-to-four month outage at the dominant Western supplier propagates into airframers' build schedules regardless of who is at fault. Primes with military transparency content — the platforms Garden Grove serves — carry the schedule risk even though they carry none of the legal risk. The 28 September target and the phased Q4 ramp are therefore watched as closely by customers as by shareholders.

The broader read-across is to how UK-listed industrials with US operating footprints are priced for regulatory tail risk. Melrose is a British company whose profit centre is American manufacturing, and this episode has demonstrated the full sequence: an incident, a mass evacuation, a district attorney investigation, continuing regulatory engagement, and a civil claims programme — with only the criminal strand now closed. Rolls-Royce and BAE Systems carry structurally similar geographies and structurally similar exposure to US enforcement processes that do not run on a UK reporting calendar.

For the FTSE 100 itself, the effect is limited. Melrose is a mid-weight constituent and an 8% move contributes a fraction of an index point. The signal value is higher than the index value: it is a reminder that for aerospace suppliers, the operational disclosure and the legal disclosure move on different clocks, and the market is much better at pricing the first than the second.

What to watch next

• 28 September — the restart target. It is a target, not a commitment, and it is followed by a phased ramp rather than a switch. Slippage would extend the outage into a second full quarter.
• The claims programme opening, expected within weeks. Take-up rate against the $100m ceiling is the first hard evidence of whether the cap is generous or tight.
• Whether the £175m buyback restarts. With £163m still unspent, resumption would be management's clearest statement that it considers the residual liability bounded. Continued suspension says the opposite.
• The separately-handled items. Emergency services claims and any civil enforcement penalties sit outside the programme and outside the guidance. These are the numbers with no ceiling and no date.
• Full-year 2026 results. Melrose reported its 2025 audited results on 27 February 2026; the equivalent 2026 disclosure is where the total Garden Grove cost — exceptional charges, claims and lost output combined — is finally presented as one figure rather than several.

The market bought the end of the criminal investigation. What it has not yet priced is the part of the release that says the $100m does not cover everything.