Building products and engineering and construction

Three businesses that candidates treat as one

Ask most candidates about building products and construction and you get a single blur: "companies that make stuff for houses." That will not survive a follow-up, because the coverage team sits on top of three genuinely different businesses that happen to share an end market.

The first layer is manufacturing: the companies that produce the physical product, including roofing shingles, insulation, windows and doors, wallboard, cement and aggregates, siding, plumbing fixtures, and HVAC equipment. They own plants, they buy raw materials and energy, and their earnings turn on price versus input cost and on how full the factories run.

The second layer is distribution. Wholesale distributors, pro dealers, and specialty suppliers buy from manufacturers, hold inventory across a branch network, and get product to a jobsite when a contractor needs it. They manufacture nothing. Their economics are working capital and logistics: inventory turns, delivery density, credit extended to small contractors, and the spread earned for having product two hours away instead of two weeks away.

The third layer is installation and construction. Specialty contractors (roofers, mechanical and electrical trades, concrete subcontractors) and large engineering and construction firms put the product in place, or design and build the facility around it. They own few hard assets. They own people, equipment fleets, and above all a book of contracts.

Capital intensity falls sharply as you move down that chain, and so do margins, but so does the capital you need to earn a return on. That is why a coverage team can have a cement business trading at one multiple and a contractor at half of it while both get described as "construction exposure" on the same pitch page. Where these sit relative to the rest of the coverage universe is laid out in the industrials sub-sectors map.

Demand comes from four separate places

Interviewers like this sub-sector precisely because candidates collapse demand into one variable. There are four, and they do not move together.

New residential construction is the one everybody names. It is the most volatile of the four because it depends on builders making forward-looking decisions about financing, land, and absorption. When that decision gets harder, starts fall fast and far. Products with high content per new home (framing lumber, wallboard, windows) carry the most exposure here.

Repair and remodel activity is the quieter and, for interview purposes, the more interesting driver. It tracks the installed base rather than new decisions: how many homes exist, how old they are, and how long the roof or the furnace in them has been in service. A roof at the end of its life gets replaced whether or not anyone is building new houses. That is why repair and remodel is structurally steadier than new construction, and why companies with a heavy replacement mix (roofing, water heaters, residential HVAC) can hold volume through a downturn in starts while a window supplier levered to new builds cannot. Name that split, new building decisions versus the aging of an existing stock, and you are ahead of most of the room.

Non-residential and commercial construction covers warehouses, offices, hospitals, data centers, schools, and plants. It typically lags residential, because commercial projects are longer-dated and start from committed capital budgets rather than a monthly sales decision. A company with both exposures does not see both halves turn in the same quarter.

Public infrastructure spending runs on a fourth clock entirely. Highways, bridges, water systems, and transit are funded by appropriations, not by private credit appetite, so this demand can rise while private construction falls. Aggregates producers, heavy civil contractors, and pipe and precast suppliers are levered to it. The classic trap is calling a company "cyclical" without asking which of these four cycles it is actually exposed to.

Aggregates and cement have a moat that does not travel

If one business here is worth knowing cold, it is aggregates. Crushed stone, sand, and gravel are heavy, low-value-per-ton commodities worth very little relative to the cost of moving them, and freight cost rises with every mile. Past a modest haul radius, delivered cost exceeds what a customer will pay, so a quarry cannot compete outside its local market and no distant competitor can compete inside it.

That physical constraint creates something rare in industrials: a genuine local franchise, with no patents, brands, or technology behind it. Inside its haul radius a quarry faces two or three competitors rather than a global field, and it can generally push price through the cycle, including when volumes are soft. Cement plants behave similarly, with a wider but still bounded radius from the plant and terminal network.

Permitting reinforces it. Opening a new quarry requires zoning approvals, environmental review, and local political consent, and the communities nearest the best reserves are usually least interested in granting it. The scarce asset is not the rock, which is common, but the permitted site with reserves close to a growing metro area. New supply cannot simply be built where the demand is.

The valuation consequence follows directly. Aggregates and cement assets typically command premium EBITDA multiples relative to other building products, and buyers underwrite reserve life (how many years of permitted reserves sit behind the quarry) alongside current earnings. When a bid for an aggregates business looks expensive against the rest of the sector, the buyer is usually paying for pricing power that a wallboard plant does not have.

Engineering and construction is a valuation problem of its own

This is where interviews get sharp.

Contractors do not sell a product off a shelf. They bid work, win it into a backlog, and recognize revenue over time as the job progresses, most commonly on a cost-to-cost percentage-of-completion basis: if you have incurred 40 percent of the total costs you expect to incur, you book 40 percent of the contract's revenue and margin. Everything that makes E and C hard flows from that one accounting fact, because the reported number depends on a total-cost estimate that management controls and that changes as the job runs.

Contract type determines who carries that risk.

Fixed-price and lump-sum turnkey contracts hand the contractor a single price for the finished work. Any cost above the estimate comes out of margin. On a repetitive, well-understood scope that is fine. On a large, first-of-a-kind project (a process plant, a major civil crossing, an energy facility with novel technology) it is where contractors have historically destroyed enormous amounts of shareholder value, and the industry has seen firms pushed to the brink by a small number of troubled lump-sum jobs.

Cost-plus and reimbursable contracts pay documented cost plus a fee, sometimes with an incentive tied to schedule or budget. Margins are thinner but far more predictable, and the customer carries cost risk. A contractor shifting mix toward reimbursable work is trading margin percentage for earnings quality, which you should read as de-risking rather than deterioration.

Two mechanisms deserve names in an interview. Change orders are customer-approved scope changes with agreed pricing, which are normal and healthy. Claims are amounts the contractor believes it is owed that the customer disputes, often for delay, differing site conditions, or design changes. Recognizing revenue on unapproved claims is aggressive, and disputes can sit unresolved for years. Unbilled receivables (costs and estimated earnings in excess of billings) are work performed but not yet invoiced, and they are a direct working capital drag: cash goes out to labor and subcontractors long before it comes back.

The trap to name out loud is this: a growing backlog can hide deteriorating job margins. Backlog is a gross bookings figure. It says nothing about the profitability of the work booked, and a contractor can grow it quickly by bidding aggressively. The tell is composition rather than the headline, the fixed-price share, the age of the book, and whether unbilled receivables and claims are growing faster than revenue. That is covered in backlog, book-to-bill, and cyclicality.

A hypothetical worth being able to run at a whiteboard: a contractor signs a fixed-price job at $500 million with an expected total cost of $450 million, so $50 million of margin. Costs incurred reach $270 million, 60 percent of the original estimate, so $30 million of margin has been booked. Management then reforecasts total cost to $520 million. The job is now an expected $20 million loss, and accounting requires the full anticipated loss to be recognized immediately rather than spread over the remaining work. The swing in that period is roughly $50 million, reversing the $30 million already booked plus taking the $20 million loss, on a project that looked on plan a quarter earlier.

How the layers compare

What the multiple spread is telling you

E and C firms trade at lower EBITDA multiples than product manufacturers for reasons you should list without hedging: idiosyncratic project execution risk an outsider cannot underwrite, thin margins where a small overrun percentage swamps net income, earnings that swing on a single reforecast, and no pricing power in a bid market where the low bidder wins.

There is also a structural trap in the enterprise value bridge. Contractors frequently report a large gross cash balance, and a careless analyst nets all of it against debt and lands on a small or negative net debt figure that makes the multiple look absurd. Much of that cash is not the company's to keep. It is customer advances and milestone billings collected ahead of cost incurred, float against work still owed, and it reverses as the job burns down. The disciplined fix is to treat part of that balance as operating rather than excess cash, or to use an average across a project cycle instead of the year-end snapshot.

Distributors get valued differently again. Because the asset base is working capital rather than plants, the lens is capital efficiency: inventory turns, cash conversion, and returns on invested capital. A distributor compounding at high returns on a modest capital base can justify a multiple that looks rich against its margin percentage, because margin percentage is the wrong denominator for a business that competes on turning capital quickly.

Mid-cycle normalization matters more here than almost anywhere else in industrials. Valuing a building products company off peak-cycle EBITDA is the most common modeling error in this sub-sector, and the fix is to normalize volumes and margins to a through-cycle level before applying a multiple, as set out in how industrials companies are valued.

Why deal flow here is roll-up territory

One line about M&A here: fragmentation plus local density economics equals consolidation, and sponsors noticed a long time ago.

Distribution, specialty contracting, and residential installation services are thousands of owner-operated businesses, many founder-owned and approaching a succession event. They also have a property sponsors love. A second roofing contractor in a metro area you already serve shares branches, trucks, crews, and purchasing scale, so the acquired earnings are worth more inside the platform than standing alone. Add multiple arbitrage, buying small businesses at mid single-digit multiples into a platform valued materially higher, and the strategy works without any change in end-market demand.

That shapes the junior seat. Bankers here spend much of their time on sponsor-backed platform sales, add-on acquisitions, and financing those platforms, which is why leverage vocabulary matters even in a coverage interview; the terminology is collected in leveraged finance terms. On the corporate side, manufacturers periodically separate a slower-growing line or a business with different cycle exposure, and that logic is covered in industrials M&A and spin-offs.

Practice question

Why do engineering and construction firms trade at lower EBITDA multiples than building products manufacturers, even when both are exposed to the same construction cycle?

Because the earnings are lower quality and the business carries risk that an outside investor cannot really underwrite. A contractor recognizes revenue over time on a percentage-of-completion basis, so the reported number depends on management's estimate of total cost to complete each job. On fixed-price work, if that estimate moves against them, the contractor absorbs the entire overrun, and accounting requires the full anticipated loss on a job to be recognized as soon as it is expected rather than spread out. So a single large project reforecast can erase a quarter with no change in end-market demand. On top of that, margins are thin, so a small percentage overrun is a large percentage of net income, and pricing power is weak because work is won by competitive bid. A product manufacturer, by contrast, sells a repeatable product with plant-level pricing and cost dynamics that you can actually model. I would also flag one bridge issue: contractors often carry a big gross cash balance that is really customer advances and milestone billings against work not yet performed, so netting all of it against debt understates enterprise value and makes the multiple look cheaper than it is. The right check is backlog composition, the fixed-price share, and whether unbilled receivables and claims are growing faster than revenue.

What the interviewer is listening for: Whether you connect the multiple to percentage-of-completion accounting and contract type rather than saying "construction is cyclical," which is true of both businesses. They also want to hear that you would not naively net a contractor's cash in the enterprise value bridge, and that you know backlog growth alone tells you nothing about job margins.