Ryerson Holding Corporation (RYZ) | The Buildout — AI Infrastructure
The Verdict
Ryerson buys industrial metal from mills and turns it into something a factory can use — cutting, slitting, leveling, tempering, welding, and fabricating steel, stainless, and aluminum, then holding it in local inventory and delivering it just-in-time. In the AI build-out it is a picks-and-shovels supplier, not a technology company. Its processed metals go into customers' power, IT hardware, cooling, and fabrication work on data center and power generation projects, and that demand shows up inside several existing end markets rather than in a segment of its own. Management describes the company as an AI infrastructure partner to its customers.
| Market Cap | — |
| Revenue (TTM) | $5.8B |
| Revenue Growth | +31.5% |
| EBITDA Margin (TTM) | 2.0% |
| Net Debt | $1.3B |
| Earnings Beats | 1 of 2 |
| P/E (TTM) | — |
| EV/EBITDA (TTM) | — |
What We Like
- Data center and power generation projects were about 7% of Q2 2026 revenue and grew roughly 30% sequentially, per management's estimate, with demand described as continuing to build.
- Synergy execution is running ahead of plan: Q3 realization of $13–14M implies a $52–56M annual run-rate, above the $40M first-year target, at roughly $1.2M of one-time costs through Q2.
- Leverage fell from 5.1x to 4.0x in one quarter and liquidity rose to $757M from $618M, with management pointing to closer to 3.0x by end-2026.
- Same-store North American tons rose 5.8% in H1 2026 against MSCI industry growth of 2.9%, and the contract book improved year-over-year for the first time since 2022.
- Operating leverage is visible: same-store WDSG&A fell to 15.2% of sales from 16.8%, and same-store cost per ton fell to $402 from $416.
What We’re Watching
- Q3 2026 is guided down: adjusted EBITDA excluding LIFO of $88–92M versus $101M in Q2, on "some compression" from program pricing lags and cost inflation. Management calls the headwinds temporary and says they need about one inventory turn to cycle through.
- Non-ferrous prices reverted roughly 15% at the end of Q2 and into early Q3; aluminum and stainless three-month averages stepped down 10–15%.
- LIFO expense has come in above guidance twice — $10M versus $6–8M expected in Q1, and $17M versus $14–16M guided in Q2.
- Leverage at 4.0x, down from 5.1x in Q1, remains above management's stated goal of moving closer to 3.0x by end-2026; the new $100 million repurchase authorization is available through April 2028.
The merger thesis is intact and, on the operating lines, strengthening. Synergies are being delivered ahead of schedule, leverage fell a full turn in one quarter, share gains are running at roughly twice the industry rate, and the contract book turned positive year-over-year for the first time since 2022. What is new is a specific, quantified margin risk: the spread between transactional and program pricing widened to a three-year-high 700–800 basis points against a 600–700 basis point historical norm, and management says the program book has not caught equilibrium yet. The open question is whether that program-pricing lag is genuinely a one-inventory-turn transition, or a structural feature of a book that is now 60% contract in a rising-cost environment.
Earnings
Q2 2026 was Ryerson's first full quarter as a combined company with Olympic Steel. Net sales were $2.01 billion, described as a record, on just over 800,000 tons shipped — above the guided $1.86–1.93 billion. Gross margin was 17.8%, while adjusted gross margin excluding LIFO expanded 20 basis points to 19.3%. Adjusted EBITDA excluding LIFO of $101 million beat the guided $88–92 million. GAAP net income of $15.5 million included a $15.7 million purchase accounting charge; adjusted net income was $27.6 million.
| Metric | Q2 FY2026 | Q1 FY2026 | Q2 FY2025 | YoY |
|---|---|---|---|---|
| Revenue | $2.0B | $1.6B | $1.2B | +71.6% |
| Gross margin | 17.8% | 18.1% | 17.8% | +0bps |
| EBITDA | $69M | $41M | $24M | +184.0% |
| EPS | $0.29 | $0.10 | $0.06 | +402.5% |
| Data center & power generation revenue share | ~7% of Q2 revenue | n/a | n/a | Sales +~30% sequentially |
| Synergy realization | ~$5M | n/a | n/a | Q3 guided $13–14M |
Second quarter results also continue to benefit from secular demand tied to data center and power generation projects, which we estimate represented approximately 7% of our second quarter revenues. Sales tied to these applications continue to accelerate during the quarter, increasing approximately 30% sequentially.— Richard (Rick) Marabito, 2026-07-30
Management tone: Management stayed promotional about the multi-quarter arc — "escape velocity," "stacking wins and building synergy momentum" — while guiding Q3 down on margin compression rather than burying it. Synergy language escalated from "well on our way" on the Q1 call to "ahead of schedule" on Q2. Commodity commentary turned more mixed: carbon was called strongest, then stainless, then aluminum, with non-ferrous described as reverting roughly 15%. Leverage framing improved from "expect it to move lower" to a stated year-end target near 3.0x. The Q&A was light — three questions, after which the operator reported no further questions — leaving balance-sheet durability and contract-book durability largely unexplored.
Management Guidance
Q3 2026 guidance: volumes −3% to −5% sequentially on normal seasonality, average selling prices flat to up 2%, revenue $1.87–1.95B, adjusted EBITDA excluding LIFO $88–92M, and synergy realization of $13–14M. Management guided to "some compression" in margins, driven by rising material costs, ongoing program customer pricing lags, and inflationary pressure across labor and delivery. It also flagged $5–7M of additional inventory purchase accounting adjustments through year-end, net leverage "closer to" 3.0x by end-2026, and FY2026 capital expenditures of about $75M ($50M same-store plus $25M Olympic). The two-year $120M annual run-rate synergy target was reaffirmed.
Trajectory
Revenue stepped up sharply with the merger: $1,104.8 million in Q4 2025, then $1,566.5 million in Q1 2026 (+41.8%) and $2,006.2 million in Q2 (+28.1%). The step is mostly the deal — Q1 contained only six weeks of Olympic Steel and Q2 the first full quarter. Underneath, same-store volume grew at roughly twice the industry rate, with H1 2026 North American tons up 5.8% same-store against MSCI industry growth of 2.9%, and same-store average selling prices up 7.2% sequentially in Q2. Reported gross margin was 17.8% in Q2, but adjusted gross margin excluding LIFO expanded 20 basis points to 19.3%; the gap is two mechanical items, a $17 million LIFO charge and a $15.7 million purchase accounting adjustment. Both are finite, and management guided to margin compression in Q3 as program pricing catches up.
The Model
The model projects FY+1 revenue of $7,353 million and EBITDA of $287 million, a 3.9% margin. For FY+2 it projects revenue of $7,950 million and EBITDA of $370 million, a 4.65% margin. The near-term year reflects a full twelve months of the combined Ryerson and Olympic Steel platform plus the synergy run-rate management expects to exceed its $40 million first-year target. The second-year step-up reflects continued synergy capture toward the $120 million two-year annual run-rate target reaffirmed for early 2028, and the commercial mix shift management describes as moving from 40-60 toward 45-55 transactional-to-contract near term and 60-40 longer term.
| Metric | FY2025 | Next FY (E) | Following FY (E) |
|---|---|---|---|
| Revenue | $4.6B | $7.4B | $8.0B |
| YoY Growth | — | +60.9% | +8.1% |
| EBITDA | $46M | $287M | $370M |
| EBITDA Margin | 1.0% | 3.9% | 4.7% |
Projections are the median of 5 independent model runs. The model’s revenue sits 8.4% above analyst consensus.
Q3 2026 guidance: volumes −3% to −5% sequentially on normal seasonality, average selling prices flat to up 2%, revenue $1.87–1.95B, adjusted EBITDA excluding LIFO $88–92M, and synergy realization of $13–14M. Management guided to "some compression" in margins, driven by rising material costs, ongoing program customer pricing lags, and inflationary pressure across labor and delivery. It also flagged $5–7M of additional inventory purchase accounting adjustments through year-end, net leverage "closer to" 3.0x by end-2026, and FY2026 capital expenditures of about $75M ($50M same-store plus $25M Olympic). The two-year $120M annual run-rate synergy target was reaffirmed.
What Could Go Right — and Wrong
- Data center and power generation revenue share holds above about 7% or keeps growing, with the roughly 30% sequential rate proving repeatable.
- The transactional-to-program spread narrows back toward the 600–700 basis point historical norm, lifting commercial margins.
- Synergy realization tracks to or through the two-year $120 million annual run-rate target by early 2028.
- The contract book's year-over-year improvement persists, turning the larger half of the book positive.
- Net leverage reaches closer to 3.0x by end-2026, freeing the $100 million repurchase authorization.
- The program-pricing lag proves structural rather than a one-inventory-turn transition, extending margin compression past Q3.
- The roughly 15% non-ferrous price reversion flows further into realized average selling prices, removing the price support the Q3 guide relies on.
- Delivery and labor cost inflation outruns pass-through for more than a quarter, with program contracts setting the timing of any increase.
- Net leverage stalls above 4x as the trailing-EBITDA seasoning from a full quarter of Olympic Steel fades.
- Demand stays narrow: agriculture, consumer discretionary, and residential construction remain recessed while large-OEM program demand is stagnant.
Looking Ahead
Over the next twelve months the story turns on three things: whether Q3's margin compression proves to be the trough as management's one-inventory-turn framing implies, whether synergy realization keeps climbing from $13–14M in Q3 toward the $120 million two-year annual run-rate target by early 2028, and whether net leverage reaches closer to 3.0x by year-end. The Connecticut consolidation of Olympic Milford and Ryerson Specialty Alloys is guided to complete in Q1 2027. Management expects data center and power generation demand to keep building, and says early Q3 indicators showed the contract book improving year-over-year for the first time since 2022.
- September 2026Quarterly dividend payment — $0.1875 per share; record date Sept 3, payment Sept 17.
- Q3 2026Q3 2026 results call — Tests whether margin compression is transitory and synergy lands.
- End of 2026Net leverage target — Management points to a ratio closer to 3.0x by year-end.
- Q1 2027Connecticut consolidation completes — Consolidates Olympic Milford and Ryerson Specialty Alloys.
- Early 2028$120M synergy run-rate — Two-year annual run-rate synergy target reaffirmed.
Financials
Annual Summary
| Metric | FY2024 | FY2025 | TTM | YoY |
|---|---|---|---|---|
| Revenue | $4.6B | $4.6B | $5.8B | -0.6% |
| Gross Margin | 18.0% | 17.0% | 17.3% | 105bps |
| EBITDA | $104M | $46M | $116M | -56.1% |
| EBITDA Margin | 2.3% | 1.0% | 2.0% | 126bps |
| Net Income | −$9M | −$56M | −$33M | -555.8% |
| Free Cash Flow | $105M | $36M | −$152M | — |
| Net Cash | — | — | — | — |
Key Ratios (Trailing)
- P/E TTM—
- EV/EBITDA TTM—
- EV/Revenue TTM—
- Price/FCF TTM—
- Gross Margin (TTM)17.3%
- EBITDA Margin (TTM)2.0%
- Net Margin (TTM)-0.6%
- ROIC0.6%
- FCF Conversion-131.1%
- SBC / Revenue0.0%
The Company
Ryerson is a value-added processor and distributor of industrial metals — a metals service center. It buys metal from mills, then cuts, slits, levels, tempers, and fabricates it before holding it as inventory and delivering it just-in-time. The FY2025 10-K describes approximately 75,000 products across a full line of carbon steel, stainless steel, alloy steels, and aluminum, plus a limited line of nickel and red metals. The economics are working-capital-heavy and spread-driven: revenue is roughly tons shipped times average selling price, and profit depends on how much of the spread survives the fixed cost base of facilities, fleet, and sales organization.
The company operates approximately 4,300 employees across 103 facilities in North America and three in China, serving approximately 40,000 customers as of 2025-12-31. Its top 25 suppliers, including subcontractors, accounted for approximately 76% of purchase dollars in 2025. The business was re-baselined by the merger with Olympic Steel, announced 2025-10-28 and closed 2026-02-13; management describes the combined entity as one of North America's largest metal service center platforms. Q1 2026 included six weeks of Olympic Steel and Q2 was the first full quarter. Management tracks two customer classes — transactional spot business and program/contract business — at a combined 40-60 mix, with a stated goal of 45-55 near term and 60-40 longer term.
Business Segments
Competitive Landscape
The source material does not cleanly disclose Ryerson's competitor set. Olympic Steel was a competitor before the merger — the 10-K supply record still tags it "competitor," a pre-close classification the intel file flags as stale. Reliance Steel & Aluminum appears as a customer in the supply-chain file but reads as a peer metals service center. No competitive share data is disclosed; the only share-adjacent statistic in the material is the volume comparison against the Metal Service Center Institute benchmark. Management's description of the edge is operational rather than contractual: product availability, reliability, processing capabilities, and speed of response, backed by a 103-facility North American network, a 75,000-SKU breadth, and a service-level standard of 95% for A1A items. None of that is a patent, a sole-source contract, or a switching cost the filings quantify.
- Olympic SteelWas a competitor before the merger; the 10-K supply record still carries a pre-close "competitor" classification that the intel file flags as a stale artifact.
- Reliance Steel & AluminumListed as a customer in the generation-sourced supply-chain file but read as a peer metals service center in the supply-chain intelligence.
Supply Chain
Ryerson sits between the mills and a very long tail of industrial customers: it buys from a concentrated supplier base and sells into roughly 40,000 accounts. The source's supply-chain neighbors are all labeled inferred ecosystem neighbors, and none of them names Ryerson directly.
More on RYZ: Earnings recap