Long-Term Thesis

Long-Term Thesis: Underwriting a Niche Monopoly for the Next Decade

Figures converted from GBP at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged. Note: Somero is a US company that already reports its financial statements in US dollars, so the figures below are as-reported with no conversion applied; only the AIM share price (quoted in GBP pence) is translated to USD at the prevailing rate for market-capitalisation and valuation.

The 5-to-10-year question for Somero is not "is this a good business" — the moat, the margins and the balance sheet settle that. It is narrower and harder: can a category monopoly that has already saturated the high end of a small, deeply cyclical niche convert its cash machine into a larger one — by widening the market down-market and through acquisition — without breaking the discipline that made it worth owning? Underwrite that, and you are buying a debt-free, 52%-gross-margin franchise at roughly 4x mid-cycle EBITDA with net cash worth ~23% of the equity. Get it wrong, and you own a no-growth cash box whose owners cannot pry the cash loose.

This page does three things: it states what must be true for SOM to be a superior 10-year holding, it tests each condition against the multi-year primary record, and it names the leading signals that would prove the thesis working or breaking long before the share price tells you.

Source: FY2025 Annual Report, Financial Review and Chairman's and CEO's Statement (revenue, gross margin, net cash and cash flow) [2] [3].

1. What has to be true over 5 to 10 years

A long-term holder is implicitly betting on five conditions, ranked below by how load-bearing each is to the return and how much the multi-year record already supports it. The discipline of this page is to separate the conditions that are already evidenced (the moat, the balance sheet) from those that are still hypotheses (the down-market expansion, M and A, the governance unlock).

Sources: gross-margin band from the 17-year history in the 2025 Results Presentation [1]; pre-boom base and FY2026 outlook from the FY2025 Annual Report [2] [3]; the 40,000-contractor addressable market and M and A framework from the 2025 Results Presentation [10] [11]; AGM outcome per company RNS, 17 June 2026 (post-dates the filing corpus).

The shape of the bet is unusual: conditions 1 and the safety in 2 are already paid for in the price (you are buying a proven moat at a trough multiple), so the upside is concentrated in conditions 3–4 (does the arena widen and does capital get deployed well) and the trap risk in condition 5 (does the cash ever reach owners). A PM should weight evidence on conditions 3 and 5 far more heavily than the next quarterly print.

2. What you are actually underwriting: a moat the numbers have already stress-tested

The durable core of the thesis is the one thing a single snapshot cannot show and the multi-year record proves cleanly: gross margin that refused to move while revenue swung more than five-fold. Across seventeen years — through a 53% revenue collapse in 2009 and a three-year, 33% slide into 2025 — Somero's gross margin never left a 47–58% band, and it sat at 52% in the worst trading year of the past decade because the company raised price into the downturn [1] [2]. A commodity machine-maker cannot do that; a price-setting category franchise can.

Source: 2025 Results Presentation, Historical Results — 17-year revenue and gross-margin series [1].

That flat line is the byproduct of four advantages that an investor can each tie to a mechanism, not an adjective: a category franchise Somero created in 1986 and still leads, protected by 140-plus patents and end-customer flatness specs written by blue-chip occupiers (Amazon, Walmart, Costco, Tesla, Prologis) [5]; an operational switching cost built on 24/7 global service that management calls "difficult to replicate" because an idle machine on a job site burns money fast [6]; a counter-cyclical aftermarket annuity on an installed base in 90-plus countries (parts and service fell only ~11% in 2025 against a ~19% drop in machine sales) [3]; and a micro-niche too small for large OEMs to bother attacking, with no listed pure-play rival anywhere. The signature shows up in returns: even at the 2025 trough, ROCE of ~16% beat every diversified machinery peer's good-year return.

Source: derived from reported financials, FY2022-FY2025 (operating income, equity and capital employed) [2].

The durable conclusion: underwrite the moat for the through-cycle margins and the returns, not as a growth engine. It defends share and pricing, which it has done; it does not defend revenue against the construction cycle, and it cannot manufacture growth in a flat end-market. The one place the moat is genuinely thinner is Europe (~10% of revenue), where management has repriced its new Hammerhead ride-on "to compete more effectively" against fragmented local and emerging Chinese rivals — an admission it is a price-taker at that margin [18]. The single most informative durability signal for the next decade is therefore gross margin against its ~47% seventeen-year floor: breaking below it while volumes are stable would mean an attacker has set a competing price and the moat — not the cycle — has reset.

3. The central long-term tension: a fortress balance sheet with a capped reinvestment runway

Here is the structural problem a 10-year holder must confront head-on. Somero generates cash almost perfectly — FY2025 free cash flow of $17.0m exceeded net income of $10.2m on capex of just $0.8m (under 1% of revenue) — but it has almost nowhere high-returning to reinvest it. The flip side of "a niche too small to attract large OEMs" is "a niche too small to compound a lot of capital." The plant tells the story bluntly: the Houghton, Michigan facility was expanded to a capacity management sized at "$175.0m" of revenue in 2021 and "over $200.0m" by 2022 [12] [13] — and revenue then fell to $88.9m. The company built capacity for a business more than twice its current size, so organic capacity is not the constraint on growth; demand is. That is why the cash comes back to shareholders rather than into the business.

Source: FY2025 Annual Report cash-flow disclosures and Dividends / Share Buyback notes [3] [8].

Over 2022–2025 Somero returned roughly $82m in dividends and buybacks — more than half its current market value — while never carrying debt, on a formulaic 50%-of-earnings ordinary dividend that flexed down honestly with the cycle (the FY2025 ordinary dividend fell to $0.1024 from $0.1693) plus opportunistic repurchases (856,785 shares bought in 2025) [8]. This is the cleanest piece of the long-term record: a disciplined cash-return machine that has paid owners every year through a brutal downturn without balance-sheet stress.

The decade-defining change is that management is pivoting from returning cash to deploying it. In 2025 it formalised a Capital Allocation Framework and, for the first time, a dedicated Mergers and Acquisitions Framework — prioritising a strong balance sheet, organic investment, strategic acquisitions, then shareholder returns — and signalled headroom to lever up to ~2.0x net-debt/EBITDA for deals, part-funded by suspending the supplemental dividend [4] [11].

4. The structural demand case: the trend is secular, the timing is cyclical

The reason the niche is worth widening at all is that its long-run demand drivers are genuinely structural, even though its revenue line is violently cyclical. Management points — consistently, across years — to four durable drivers, and they are the right ones: a chronic skilled-labour shortage that makes automation a necessity rather than a luxury (a laser screed lets a contractor pour more slab with fewer scarce workers); e-commerce and onshoring/reshoring driving warehouse and plant construction; and mega-project end markets — data centres, EV battery plants and semiconductor fabs — which management explicitly ties to US legislation such as the CHIPS Act and its roughly $280bn of support for domestic semiconductor manufacturing [9] [8].

The correct mental model — and the one a PM should hold for a decade — is that structural demand sets the trend; macro sets the timing. A contractor who believes in labour scarcity and onshoring will still defer a discretionary $200k+ machine purchase when rates, tariffs or sentiment turn. That is precisely how a structurally-growing niche produces a sharply cyclical revenue line, and why the FY2022 peak ($133.6m) and the FY2025 trough ($88.9m) can both be "normal" — the spike was a pandemic warehouse pull-forward, the trough is the pre-boom 2016–2020 base of $79–94m [1].

The genuinely new growth vector — and the most important long-term call beyond cyclical reversion — is down-market penetration. Somero historically served larger contractors running big pours; the new, lower-priced Hammerhead ride-on and Viper walk-behind are explicit moves to convert manual crews to mechanised ones, against a US total addressable market management frames as 40,000-plus concrete contractors, most of whom have never owned a laser screed [10]. If even a sliver of that base mechanises over a decade, it reframes Somero from a saturated high-end supplier into a market-developer — the difference between a cyclical and a compounder. This is unproven and slow-moving, which is exactly why it is a 10-year signal, not a quarterly one.

5. Through-cycle earnings power: the number the whole valuation turns on

For a cyclical, spot P/E is a trap — it looks expensive at the trough (depressed E) and cheap at the peak. The long-term thesis lives or dies on normalised through-cycle EBITDA, applied to an enterprise value flattered by net cash. The cycle of the last several years brackets the range: a $44.6m peak (FY2022), a mid-cycle of roughly $28m (the 2017–2020 average and the FY2024 actual of $27.7m), and a $17.5m trough (FY2025).

Source: adjusted EBITDA from FY2021-FY2025 results and the 17-year history [1] [2].

At a ~193p share price the equity is worth roughly $135m and, after ~$33m of net cash, the enterprise value is only ~$102m [3]. Hold that fixed EV against each rung of the cycle:

Source: EBITDA per FY2022-FY2025 Annual Reports [2]; EV/multiples derived using ~$135m market cap and ~$33.2m net cash [3]. Market capitalisation reflects market data, as reported.

You pay ~5.8x trough EBITDA, under 4x mid-cycle, and barely 2x peak — for a debt-free, 52%-gross-margin monopoly. The valuation only makes sense as a trap if you believe through-cycle EBITDA has been permanently reset toward $17m. The whole long-term debate compresses to that one variable, and the cleanest read on it is North American Boomed-screed revenue (the largest, highest-margin, most operating-levered line) inflecting up off its depressed base across the next several annual prints, or stepping down again.

6. The governance gate: can the value get out?

A cheap balance sheet is only a margin of safety if the cash can reach owners. Over a decade, that depends on governance — and this is the weakest pillar of the thesis. The skin-in-the-game has drained out of the cap table: founder-CEO Jack Cooney's 614,634-share stake left with his March 2025 retirement, his successor Tim Averkamp owns zero ordinary shares, and the entire board and management own roughly 0.3% of the company [17] [15]. Worse for alignment, the executive equity awards (RSUs) carry no performance conditions — the committee states expressly that "aside from service period requirements, performance criteria should not be applied" — so a meaningful slug of incentive pay rewards staying employed, not creating value [15].

Against that sits a concentrated, newly activist owner base that has effectively replaced insider alignment with external pressure: between end-2024 and end-2025 the top three holders roughly doubled their stakes — Brian Kelly to 14.1%, Regent Gas Holdings to 12.3%, VN Capital to 10.1% — so three holders now control well over a third of the company [16].

Source: significant-shareholder tables, FY2024 and FY2025 Annual Reports [16].

The collision came at the 17 June 2026 AGM, where shareholders voted down all seven resolutions — remuneration policy on just 38.65% support, the accounts on 49.76%, the auditor on 43.13% — yet the directors kept their seats, because as a Delaware corporation Somero elects directors by plurality, not majority, on a staggered, three-class board (per the company's RNS of 17 June 2026; this post-dates the filing corpus and is not in the citable filing set). That is the value-trap signature in its purest form: an owner base that wants change, structurally blocked from forcing it. The Board has since launched a governance review, begun shareholder consultation, and confirmed its 91-year-old longest-tenured non-executive will step down — constructive, but reactive.

For a 10-year holder this is the binary that matters most for unlocking (as opposed to compounding) value. Genuine reform — majority voting, a de-staggered board, performance-linked equity, or re-domiciling to match AIM norms — would convert the net cash and activist pressure into a real catalyst (a forced cash return, a sale, or disciplined deployment). Continued reliance on Delaware plurality voting to override owner sentiment would keep the cash trapped and the discount permanent.

7. Five-year scenarios: where the thesis lands

Sources: mid-cycle EBITDA and cycle range from the 17-year history [1]; down-market TAM and M and A framework from the 2025 Results Presentation [10] [11]; balance-sheet safety from the FY2025 Annual Report [3].

The asymmetry is favourable but conditional: the bear case is floored by net cash worth ~23% of the market value and a covered dividend, so the realistic downside is a de-rate rather than a wipeout, while the bull case requires several independent things to break right. This is why the holding is best framed as a high-quality, downside-protected option on cyclical reversion plus self-help — own it for the floor and the dividend, underwrite the upside on conditions 3–5 rather than assuming them.

8. What proves it working or breaking — the multi-year scorecard

These are the leading indicators a PM should track over years, not quarters — the signals that confirm or refute the thesis well before the share price resolves.

Sources: margin band and cycle range from the 17-year history [1]; pricing and Boomed-screed revenue from the FY2025 Annual Report Financial Review [2]; down-market strategy and M and A criteria from the 2025 Results Presentation [10] [11]; competitive and cyclical risk from the FY2025 Annual Report Principal Risks [7].

9. Synthesis: a proven moat, an unproven second act

Strip the cycle noise and the 10-year thesis resolves cleanly. What is durable and already proven — the moat, the ~50% through-cycle gross margin, the asset-light 100%+ cash conversion, the debt-free balance sheet and the disciplined cash-return record — is exactly what the market is not paying for at ~4x mid-cycle EBITDA with net cash worth ~23% of the equity. That asymmetry, floored by cash and a covered dividend, is the reason to own it.

What is unproven, and where the next decade is actually decided, is the second act: whether Somero can widen a niche it has already saturated — converting 40,000-plus manual contractors and making its first-ever acquisitions — without breaking the capital discipline that made it worth owning, and whether a governance structure the owners have voted against will let the value out at all [10] [11]. The honest verdict for a 5-to-10-year underwriter: believe the quality, the cash returns and the downside floor — they are seventeen years deep in the record; put the growth story and the M and A pivot on probation until there is share-gain and deal evidence to judge; and treat the governance review as the swing factor that decides whether this is a compounder, a coupon, or a trap. The top long-term driver is mid-cycle normalisation plus down-market arena expansion; the top failure mode is a value-destructive, governance-unchecked M and A pivot deploying the very cash that is today the margin of safety.