Actionable insights — The Westinghouse IPO

1. The profitable-incumbent screen: when the earner arrives after the story stocks

1. In any theme where speculative names have already re-rated, keep a list of the private incumbents that own the theme's core, proven asset — the approved design, the licensed technology, the installed base.
2. Score each on the one dimension the upstarts fail: does it have real, audited revenue and profit today? That single test separates an operating business from an option on a technology.
3. When such an incumbent files or announces, treat it as a valuation event for the whole theme, not just for itself: it gives the market a profitable yardstick against which pre-revenue peers must be marked.
4. Position two ways: consider the incumbent on its own merits, and check whether the arrival of a credible earner should compress the multiples of the story names competing for the same capital.

• The S-1 when it goes public (revenue, margins, backlog); whether the listed pre-revenue nuclear names de-rate as a credible earner absorbs sector capital; other profitable privates in the same theme that could follow.

2. Price a private asset through its listed part-owner's disclosures

1. When a private company matters to your thesis, find every public company that owns a piece of it — their filings are legally obliged to disclose their proportional share of revenue, earnings, or carrying value.
2. Gross the disclosed share up by the ownership percentage to estimate the whole (a 49% share of revenue × ~2 ≈ total revenue), and note the growth rate, which is usually disclosed alongside.
3. Use the same filings to check what the owner carries the stake at on the balance sheet — the gap between carrying value and a plausible IPO valuation is the mark-up the listing would crystallize for the parent.
4. Watch the currency: a Canadian or European part-owner reports in its own currency, so convert before comparing to a U.S. comp.

• The part-owners' next quarterly disclosures on the stake; equity-method vs consolidated presentation; whether the pre-IPO carrying value is disclosed; how much of the listed owner's market cap the stake could represent post-listing.

3. Anchor the valuation on a listed peer with a similar revenue base

1. Find the closest listed business doing the same thing at roughly the same scale, and reduce it to one ratio you can carry across: enterprise value (or market cap) ÷ revenue.
2. Apply that ratio to your grossed-up estimate of the private company's revenue for a first-pass valuation, before any narrative premium.
3. Then list the reasons the private name deserves a premium or discount to the comp — growth rate, contract mix, backlog visibility, government support, litigation/bankruptcy history — and adjust deliberately rather than by feel.
4. Sanity-check the headline valuation being floated in the press against that arithmetic; a big gap is either an unpriced growth story or hype you can now name.

• BWXT's own multiple drifting with nuclear sentiment (it moves the anchor); Westinghouse's growth rate vs BWXT's; margin disclosure in the S-1, which converts the sales multiple into an earnings multiple.

4. Check who bears construction risk after a post-bankruptcy model change

1. For any company in a large-project industry, identify the single line of business that has historically destroyed it — usually fixed-price construction, where cost overruns land on the contractor.
2. Ask precisely where that risk sits now: does the company still build, or has it moved to licensing, equipment supply, and services while others build?
3. An asset-light licensing/services model converts lumpy, loss-making project revenue into recurring, higher-margin fees — but also caps the upside per project and makes the company dependent on someone else being willing to take construction risk.
4. So test the counterparty side: are there constructors and utilities actually signing up to build, and on what terms? A licensing model with no willing builders earns nothing new.

• Contract structure in the S-1 (fixed-price vs cost-plus vs pure royalty); named EPC partners for the 10 planned U.S. reactors; whether any construction risk creeps back in via guarantees or equity stakes in projects.

5. Separate an announced pipeline from a started one

1. Split any project pipeline into three buckets: announced intentions, financed/committed orders, and construction actually started. Only the last two produce revenue on a schedule.
2. For a design-licensor, note that revenue often begins well before concrete is poured (engineering and licensing fees) but scales only when units are truly under way — so date the cash flows, don't just count the units.
3. Check the current base: if the company is profitable today from an existing installed base, the pipeline is upside rather than the whole thesis — a materially safer setup than a pre-revenue developer.
4. Set explicit dated milestones for the conversion (first order, first FID, first concrete) and treat slippage as the primary risk to the growth case.

• The first firm U.S. AP1000 order and its utility counterparty; NRC filings; whether "start construction by 2030" survives in the S-1 language; the mix of overseas vs domestic revenue disclosed.

6. Read government support as a two-sided term sheet — subsidy and claim

1. When a company benefits from state backing, write down both sides: the support (loans, financing, offtake, tariffs) and the consideration the government takes (equity, warrants, revenue share, price controls).
2. Quantify the claim as dilution: warrants for a fixed percentage of the stock are a direct haircut to every other shareholder, and a share of distributions above a threshold caps the cash yield.
3. Check the status of each leg. Support that is "preliminary," "in limbo," or contingent on a trade deal is not yet a cash flow — but the government's claim attached to it may still be live, so the downside can arrive before the upside.
4. Distinguish demand-side support (loans to customers to buy the product) from balance-sheet support to the company — the former is cleaner, since it lifts orders without diluting equity.

• Whether the Japan trade deal is finalized or dropped; the final warrant/stake language in the S-1 (and whether the 20% is struck at the IPO price); loan drawdowns by utilities as the leading indicator of real orders.

7. Track the supply of new paper in a hot theme

1. In a theme running hot, keep a running count of listings, filings, and secondaries — sponsors sell into strength, and the queue is a real-time read on how attractive sponsors find the bid.
2. Note the size trend: when the offerings escalate from small speculative deals to a potential sector-defining mega-listing, late-cycle supply is beginning to test demand.
3. Compare each new deal's aftermarket performance against the last — a sequence of downsized or broken deals tells you the bid is exhausted well before the theme's narrative changes.
4. Size positions in the theme against that supply calendar, not against the headlines.

• Other nuclear IPO filings; the pricing and first-day performance of each; whether Westinghouse prices inside or outside the BWXT-implied anchor — a rich price into a soft tape is the classic top-of-supply signal.