You Priced That Job Before You Owned the Steel
Run the tape on one grade of steel for a single year. CRU had US hot-rolled coil at $694 a ton on the first of January 2025. By early April it was $967. Mid-year it had backed off to $804, then climbed again to $908 by December, and it opened 2026 north of a thousand at $1,002. One grade of steel in one country, four big moves in 12 months, and every one of them landed on some fabricator's open quote.
That's the exposure most shops carry without naming it. A quote is a promise about a price you don't control yet. You bid the job in April off the steel you're looking at in April, the customer sits on the number for a few weeks, the PO finally lands in June, and you go to buy plate against a market that has moved while you waited. The estimate didn't change. The steel did.
The move is supply, not demand
What made 2025 unusual isn't only the size of the swings. It's what drove them. Steel Market Update and others tie the strength to Section 232 tariffs running at 50 percent on imported steel, which choked off foreign supply and pushed buyers onto domestic mills. When a price runs up on demand, you usually feel the demand yourself in your own order book, and you can price into it. When it runs up on supply, the way this one did, the cost lands on your material line whether or not your customers got any busier. You're paying more for the same tons to make the same parts for the same jobs.
A shop that treats its quoted material price as a fixed cost is really running a commodity position it never chose to take. On a job with steel at a third or more of the sell price, $100 a ton in the wrong direction between quote and purchase can wipe out the profit you thought you'd locked. Nobody in the building made a bad part. The margin left through the calendar.
Stop quoting the price and start quoting the terms
You can't forecast the coil market, and you shouldn't try. What you can do is stop writing quotes that pretend the price you see today is the price you'll pay in 90 days.
Put a validity window on every quote, in writing, and make it short. "Pricing firm for ten business days, material subject to mill pricing at time of order" is a normal sentence in this trade, and the customers worth keeping already expect it. The shop that leaves quotes open-ended for a quarter is handing the buyer a free option on steel, and options cost money whether or not anyone charges for them.
For longer backlogs, tie the material line to something outside your own goodwill. An escalation clause that adjusts the steel component to a published index at time of purchase moves the price risk to where it belongs, onto the market, instead of leaving it parked on your margin. On big or long-lead jobs, buy the steel to the order rather than to the quote, so the day you commit to a sell price is close to the day you commit to a buy price. The tighter you pull those two dates together, the less room the market has to get between them.
And know your own number cold. What share of each job is steel, by grade, and how far can that grade move before the job stops making money? A fabricator who can answer that in a sentence is running the material risk on purpose. One who can't is carrying it anyway, just blind.
The quote is a position
Here's the reframe worth holding onto. The moment you sign a fixed-price quote with a material cost baked in, you've taken a position on steel, same as if you'd bought a futures contract, except you did it by accident and you're not getting paid to carry it. In a flat market that position is harmless and you'll never notice. In a year like the one steel just had, it's the difference between the margin you quoted and the margin you keep.
Look at your three biggest open quotes today and find the oldest one. Then ask what $100 a ton does to it if the PO shows up next month. If you don't like the answer, the fix isn't a better forecast. It's a shorter window and a clause, and both of them can go in the next quote you send.