Reliance's $8.02 Quarter: The Border Wall Grew Up, and the Market Shrugged

A 25% EPS beat, record tons, and a keyword that fell from #1 to #18 — why Reliance's tight-supply windfall is being priced as a peak rather than a cycle

RS · Earnings Call · 2026-10-22

The beat nobody is paying for

Reliance printed the kind of quarter that usually moves a stock. Non-GAAP diluted EPS of $8.02 against a guidance range of $6.40 to $6.60. Sales up 43% year-over-year. Record quarterly tons sold. The second-highest gross profit dollars in company history at $1.5 billion. And the stock? Up 1.1% over the last 90 trading days, dead flat on the tape. The dislocation between the print and the price is the story here, and it starts with a single line item. “non-GAAP earnings per diluted share of $8.02 significantly exceeded our guidance range of $6.40 to $6.60 despite considerably higher-than-anticipated LIFO expense” — Arthur Ajemyan, CFO or similar financial executive · 2026-10-22 That LIFO expense was $187.5 million in the quarter versus a $75 million internal estimate. On a per-share basis it was $2.74 against the $1.10 baked into guidance — an incremental $1.64 headwind. Reported gross margin of 27.9% actually fell 20 basis points sequentially, which is the number that shows up in screenshots and feeds the "peak margin" reflex. Strip out LIFO and the picture inverts: FIFO gross margin expanded to 31.6% from 30.5% in Q2 and 29.0% a year ago. Non-GAAP pretax income of $544 million more than doubled. This is the structural quirk of Reliance's accounting. Rising metal prices manufacture a paper expense precisely when the underlying business is strongest. A year ago the company was telling investors the opposite story — that gross margin had bottomed:

we do see momentum coming out of that. We think like our gross profit margin troughed in Q3 based on the factors we see today.
Karla Lewis, President and Chief Executive Officer · 2025-10-23

They called it. Three quarters later, FIFO margins are 260 basis points higher.

What actually changed: price, not volume

The genuinely new fact in this quarter is pricing power. Management guided average selling price to flat-to-up 2% sequentially and delivered 8.2%. “Our third quarter average selling price increased 8.2% from the second quarter of 2026, significantly exceeding our expectation of flat to up 2%.” — Stephen Koch, Executive Vice President and Chief Operating Officer · 2026-10-22 The mechanism is the sector's defining theme this quarter — tight supply. Karla Lewis's team estimates the industry-wide shipment increase at 10.1% per the Metals Service Center Institute; Reliance grew tons 15.6%. That gap is market share gains, and it is being won on availability rather than discounting, because in a supply-constrained market the metal is the moat. Carbon plate, hot-rolled coil, specialty wide-flange beams and heat-treated aluminum plate are all described as difficult. Carbon steel products led the volume and price advance. Contrast that with the summer of 2025, when the same management team was explaining why mill price increases were not sticking: “The tariffs and the unknown around the tariffs gave our suppliers an opportunity to increase prices on some products. But on the other side, our customers also were facing that uncertainty and so we're holding back on buying.” — Karla R. Lewis, President and Chief Executive Officer · 2025-07-24 That gap between a price announcement and a price realization has closed. It is the cleanest evidence in the transcript that something has genuinely shifted, and it flows straight into the guidance: revenue has compounded from roughly $2.1 billion a quarter a decade ago to a $4.6 billion base in the last reported period, and Q3 built on that again.

The border wall grew up — and fell down the keyword ranks

Here is the most interesting tension in the dataset. The U.S. border wall project was the single highest-momentum keyword in Reliance's own transcript this time last quarter, and in Q2 it was the top-ranked phrase on the call. This quarter it sits well down the list. A careless read says the theme is fading. The opposite is true. The project has stopped being news and started being a revenue line. It contributed $0.74 per share against an expected $0.60, and the second phase — which was not contractually secured when Reliance first announced the award — is now substantially booked:

we have started to receive bookings under Phase 2, so we're at over half of the estimated $800 million Phase 2 with orders already secured for over half of that. So we think this should run through 2027, depending if they execute the full amount of Phase 2, it could run into early 2028.
Karla Lewis, President and Chief Executive Officer · 2026-10-22

Phase 1 is $1.45 billion, originally slated through Q2 2027; at the current run rate the company thinks it clears sometime in Q2. That visibility is why the keyword's momentum cooled — the uncertainty premium collapsed. The mix impact is the subtle part, and it explains the gross-margin optics better than LIFO does. Border wall tonnage is structurally dilutive to gross margin percentage yet accretive to pretax margin, because it runs through existing infrastructure at a cost per ton far below the company average. Roughly 60 basis points of the year-over-year FIFO margin reduction is the project; 40 basis points of pretax margin accretion is the offset. Analysts pressed on this twice — the leverage is real but management declined to quantify it, and that reticence is itself worth noting.

What Reliance is not saying

Across the market's curated themes this quarter, the loudest single motif is the IEEPA tariff refund windfall — a one-time cash benefit that retailers and distributors have been rushing to book. Reliance barely touches it. Instead, trade policy appears only as a risk factor, and the tariff regime is implicitly framed as a supply constraint rather than a refund opportunity. On imports, Lewis was blunt: “we don't believe that the market is ripe for a flood of inventory at this point.” — Karla Lewis, President and Chief Executive Officer · 2026-10-22 That is a beneficiary talking, not a victim. The same dynamic shows up in stainless, where specialty grades are described as having hit bottom and customers are restocking, and in data center and semiconductor demand — the latter a meaningful reversal. In April, semiconductors were "a small part of the business, but it has been lagging." This quarter the language is that momentum is "accelerating." Nonresidential construction, a little over a third of sales, is being carried by data center and energy infrastructure projects alongside heavy civil work. That confluence matters because the market's own price tape on the AI data center complex has gone cold — down mid-single digits over both 90 and 30 days — even as operating margin has climbed back to 9.5% from the 2025 trough. Reliance is levered to a theme the tape has stopped rewarding.

The setup into 2027

Guidance for Q4 is $7.40 to $7.60, embedding $112.5 million of LIFO expense and roughly $0.70 of border wall contribution. Note what the raised full-year LIFO estimate — $450 million from $300 million — actually signals: management expects metal prices to stay high enough that it must mark up inventory carrying cost. That is a bullish read wearing a bearish costume. The nearly $890 million LIFO reserve is a stored asset that flows back into income when prices eventually decline. The balance sheet is doing nothing to restrain them. effective net cash of roughly -$1.4 billion sits against 0.8x net debt-to-EBITDA and interest coverage that has averaged well above 20x. On October 20 the board lifted the buyback authorization to $2.5 billion; since 2021 Reliance has retired 23% of its shares at an average of about $236, against roughly $395 in the third quarter. And yet the multiple is not cheap. Price to net income of 21.3x reflects a market that has watched EPS fall for three years and does not yet trust the recovery. If the tight-supply read through 2027 is right, that multiple compresses on its own without the stock moving a dollar. That is the whole tension in one line: Reliance just told you its pricing power is better than it has been in years, its largest single contract has year-and-a-half visibility, and its biggest accounting headwind reverses the moment metal prices stop rising. The tape's answer, so far, is customer optimism is high but peak-cycle multiples get no credit. Someone is going to be wrong.