Reliance, Inc. (NYSE: RS) — A Great Distributor Priced as if Tariffs Are Forever

Report date: July 25, 2026 Price (July 24, 2026 close): $407.83 · Market cap: ~$20.9B · Enterprise value: ~$22.4B Sector: Materials — Metals Service Centers & Distribution Coverage status: Initiation

Sections 1–15 below contain no investment recommendation and no price target. The single exception is the Claude's Take block immediately below, which is explicitly labeled as such.

⚡ Claude’s Take

This is the author’s own independent opinion, offered as general information only — not investment advice, and not a recommendation to buy or sell any security. The analysis in Sections 1–15 below carries no position and no price target.

Verdict: HOLD if owned — AVOID initiating here at ~$408. Accumulate on weakness in the ~$270–310 zone (roughly 2.0–2.3x book value and 13–15x a ~$20 mid-cycle EPS).

Tag: “The best house in the neighborhood, listed at the top of the market.”

Reliance is a genuinely good business and I want to be clear about that before I say anything critical. It has been profitable in every single year since its 1994 IPO, through two financial crises and a pandemic. It has paid a dividend for 66 consecutive years. It has retired 34% of its shares since 2014 — and, crucially, it bought that stock well, averaging roughly $234 a share since 2021 against today’s $408. Its gross margin has held in a 25–32% band while revenue swung from $8.8B to $17.0B, which is the financial fingerprint of a real cost advantage rather than a story about one. Management is unusually disciplined: the CEO’s target cash bonus sits at the 25th percentile of the peer group, there are no employment contracts, no change-in-control agreements, no perquisites, no tax gross-ups, and the 2024 acquisitions were bought for roughly 0.96x sales with only 16% of consideration booked as goodwill. This is not a company with a governance problem or a capital-allocation problem.

It has a price problem. At $407.83 the stock trades at the 99.86th percentile of its own ten-year history on P/E, P/B and P/S simultaneously — 26.6x, 2.98x and 1.45x respectively, each an all-time high in the available record. And the earnings being capitalized at that record multiple are, by management’s own description, price-driven rather than demand-driven: Q2 average selling price rose 14.5% year-over-year against volume up 10.8%, and that price move traces directly to Section 232 tariffs raised to 50% in June 2025, with aluminum pricing that the CFO says has “nearly doubled” from pre-tariff levels. Roughly 9–10% of current EPS comes from a single Department of Homeland Security border wall contract whose Phase 1 runs only to mid-2027 and whose Phase 2 is explicitly not guaranteed. To justify $408 on this business’s historical 12–14x mid-cycle multiple, you must believe normalized EPS is $29–34 — that is, that a 50% tariff regime set by executive action is the new mid-cycle. That is a policy bet dressed as a quality-compounder bet.

The framing is late-cycle momentum, not falling knife and not value. The tape is orderly and fundamentally earned — a 1.34 Sharpe over the last year with only a 19% maximum drawdown, and the factor model reads RS as a low-beta dividend/shareholder-yield vehicle (DividendYield beta 0.79, market beta 0.65) rather than a speculative cyclical. I am not calling a top and I would not short this. But two independent groups of people who know more than the market do not appear to be buying: no officer or director has made a single open-market purchase in two years while the stock ran from ~$280 to ~$408, and the company itself bought $234M of stock in Q1 2026 and then bought exactly zero in Q2 — while reporting record results, holding $529M of unused authorization and running at 0.9x net debt/EBITDA. When the best-informed buyer with capacity, authorization and a demonstrated 74% gain on prior repurchases declines to buy, that is a revealed valuation opinion worth more than any multiple I can compute.

Conviction: medium-high on the valuation call, high on the business quality. What would flip me bullish: evidence that Section 232 at 50% has become structurally entrenched (codified by legislation rather than executive order) combined with the border wall Phase 2 being exercised and RS resuming buybacks at these levels — that would convert a policy rent into a durable earnings base. What would flip me bearish: a material easing of Section 232, or a reversal in the aluminum/carbon spread that shows up as FIFO gross margin breaking below ~29%, which would expose how much of 2026 was inventory-holding gain rather than franchise earnings.

📈 Stock Price Action — Five-Year Event Map

Over five years RS has gone from $106.11 (January 29, 2021, the five-year low) to an all-time high of $413.43 on June 12, 2026, closing July 24, 2026 at $407.83 — just 1.4% off that high, inside a 52-week range of $262.77–$413.43. Almost none of that gain came in 2025: the stock spent the entire year range-bound between roughly $277 and $309. The re-rating is a 2026 event, and it is worth roughly 42% since December 2025.

Cycle narrative. (1) The 2021 surge was a pure metal-price event as post-COVID restocking collided with constrained mill capacity. (2) February 2022’s 24.8% single-month gain followed the invasion of Ukraine, which removed Russian and Ukrainian semi-finished steel from world markets. (3) By mid-2022 the market began discounting recession and the steel price rolled over, though RS still printed a record $29.92 in EPS for the year. (4) 2023’s advance reflected the market’s belated recognition that service-center margins had structurally reset higher. (5)–(6) 2024 was a two-step derating as average selling price fell and EPS declined 10.2%, with December’s 16.2% drop the sharpest month in the five-year window. (7) All of 2025 was a stalemate: record tons sold (+6.2%, beating the MSCI industry benchmark by over 7 percentage points) were exactly offset by a 2.6% ASP decline and a $258M swing in LIFO accounting. (8) The 2026 move began with the January tariff-driven pricing inflection, accelerated 19.3% in April on the Q1 beat and the DHS border wall award, and added a further 9.2% in July on the Q2 beat — record tons sold, ASP up 14.5% year-over-year and non-GAAP EPS of $6.27, up 42%.

Price moves are Fact; attributed causes are Interpretation. No recommendation, price target, or technical level is expressed or implied in this section.

1. Executive Summary

Reliance, Inc. is the largest metals service center company in North America, operating approximately 310 locations across 41 U.S. states and 10 foreign countries, distributing over 100,000 metal products to more than 125,000 customers with 15,700 employees. It is a single-segment intermediary: it buys carbon steel, aluminum, stainless, alloy and specialty metals from mills, performs first-stage processing on roughly half its orders, and delivers small, fast, high-service-level orders to fabricators, machine shops and OEMs. Carbon steel is 53% of sales, aluminum 17%, stainless 13%. Non-residential construction and general manufacturing are each roughly one-third of revenue.

The business is better than its industry. RS has been profitable every year since its 1994 IPO and has paid dividends for 66 consecutive years. Its gross margin has held within a 25–32% band while revenue swung from $8.8B (2020) to $17.0B (2022) — evidence of a genuine cost and supply-access advantage, not a narrative one. In 2025 it grew tons sold 6.2% (5.3% same-store) while the MSCI industry benchmark declined 1.0%, an outperformance of over 7 percentage points. Management is disciplined on capital: 34% of shares retired since 2014 at an average of roughly $234 since 2021, four 2024 acquisitions bought at about 0.96x sales with only 16% of consideration in goodwill, and a compensation structure with the CEO’s target bonus at the 25th percentile of peers and no employment or change-in-control agreements.

The industry, however, is structurally mediocre. Service centers do not control their principal input cost, sell largely commoditized product, and earn returns that swing from 6.7% ROIC at the trough to 22.3% at the peak. Management’s own proxy concedes there are no closely comparable public companies, and on the Q2 call management said it hopes consolidation will produce “a more disciplined environment” — an admission that pricing discipline is currently weak.

2026 has been exceptional, and the reason matters. Q2 net sales rose 26.5% to $4,630.0M on record tons sold of 1,790.1 thousand (+10.8%) and an average selling price of $2,602/ton (+14.5%). Non-GAAP EPS of $6.27 was up 42%. But the driver is price, and the price is policy: Section 232 tariffs on steel and aluminum were raised to 50% in June 2025 and expanded to derivative articles, with aluminum “nearly doubled” from pre-tariff levels per the CFO. A further ~9–10% of EPS comes from a single DHS border wall contract ($0.41 in Q2, ~$0.60 guided for Q3) whose Phase 1 ends mid-2027 and whose $800–900M Phase 2 is explicitly optional. Quality of earnings requires care: a $258M LIFO swing between 2024 and 2025 caused reported gross margin to fall 100bp while economic (FIFO) margin rose 80bp, and management has now doubled its FY2026 LIFO expense estimate from $150M to $300M.

The valuation is the crux. At $407.83, RS trades at 26.6x earnings, 2.98x book and 1.45x sales — each at the 99.86th percentile of its own ten-year history, and each above every prior observation in an eleven-year record. Peers are also elevated (STLD 97th percentile, NUE 93rd, GWW 99.8th), so this is a sector-wide re-rating rather than an RS-specific anomaly — but RS is the most extreme case. On a mid-cycle EPS of roughly $20, the stock is at ~20x versus a historical mid-cycle multiple of 12–14x. Two informed constituencies are conspicuously absent as buyers: no insider has bought on the open market in two years, and the company itself repurchased $234M in Q1 2026 and nothing at all in Q2 despite $529M of authorization and 0.9x leverage.

2. Business Overview

What the company does. Reliance is a metals service center — an intermediary between mills and metal-consuming manufacturers. It purchases metal in mill-standard forms and quantities, holds it in inventory across a dense branch network, performs first-stage processing (cutting to length, sawing, shearing, leveling, machining, bending, welding, grinding, thermal processing), and delivers small orders quickly. Roughly half of its orders involve value-added processing; the other half are basic distribution.

The economic function is inventory and convenience. A mill will sell a fabricator 200 tons of one grade on an eight-week lead time. A fabricator often needs two tons of six different grades tomorrow. Reliance bridges that gap and is paid a gross spread for doing so. The FY2025 10-K describes the customer rationale precisely: customers buy from service centers for “the ability to obtain value-added metals processing services, readily available inventory” and quick turnaround.

Scale and footprint. Approximately 310 locations in 41 U.S. states and 10 foreign countries as of December 31, 2025 (down slightly from 315 in 2021 as the network is optimized). Over 100,000 SKUs. More than 125,000 customers. Roughly 75 subsidiaries operating under their own brand names — Phoenix Metals, United Pipe & Steel, Metals USA, EMJ, Yarde Metals, National Specialty Alloys and others. Founded 1939 in Los Angeles as a single rebar fabricator; renamed from Reliance Steel & Aluminum Co. to Reliance, Inc. on February 15, 2024 to signal a “diversified metal solutions provider” identity. Headquarters is now Phoenix/Scottsdale, Arizona.

Revenue segmentation. One reportable segment (metals service centers). Product mix has been stable for three years:

End markets (Q2 2026). Non-residential construction ~one-third (carbon tubing, plate and structural; driven by data centers, energy infrastructure, heavy civil and public infrastructure). General manufacturing ~one-third (industrial machinery including data-center equipment, shipbuilding, military, consumer products, construction machinery). Aerospace ~9% (commercial recovering as OEM build rates rise; defense and space strong). Automotive ~4%, served primarily through toll processing where RS processes customer-owned metal for a fee — toll tons are excluded from reported tons sold. Also energy, electronics and semiconductor fabrication (management flags semiconductors as “meaningfully improved”).

How it makes money. Gross profit per ton, multiplied by tons. The critical mechanic — and the one most often misunderstood — is that Reliance buys and sells almost entirely in the spot market. Per the 10-K: “Since we primarily purchase and sell our inventories in the spot market, our average selling prices generally fluctuate with the changes in replacement costs of the various metals we purchase.” When metal prices rise, RS sells inventory acquired at lower cost into a higher market and its realized spread widens. When prices fall, the reverse. This is why management states plainly in the 10-K that “Pricing for our products generally has a much more significant impact on our results of operations than customer demand.”

Recurring vs. non-recurring. There is no contractual recurring revenue in the SaaS sense — no subscriptions, no long-term take-or-pay. But the revenue is habitual: 125,000 customers placing small, frequent, low-ticket orders creates a highly diversified and repeatable order flow with no meaningful customer concentration. The one genuinely non-recurring item today is material: the DHS border wall contract, Phase 1 of which is $1.4B of sales running through mid-2027.

Verdict. A well-run, extremely diversified, low-concentration distribution business whose revenue quality is high in terms of customer diversification and low in terms of pricing control. The company does not set its own selling price; the metal market does.

3. Industry Dynamics

Structure. The North American metals service center industry is fragmented and, by management’s own account, lacks a genuine public comparable. The 2026 proxy states: “There are no public companies in the metals service center industry that are closely comparable to the Company in terms of size, stock market capitalization, complexity and financial performance.” RS is the largest by revenue at $14.29B, and yet the industry’s shipments are tracked by a trade body (MSCI) precisely because they are dispersed across hundreds of independent operators.

Market size and profit pools. Industry-wide shipments declined 1.0% in 2025 per MSCI while RS grew tons 6.2%. This is the essential industry fact: the pool is not growing, and RS’s growth is share capture from weaker competitors. The profit pool itself is highly variable, expanding when metal prices rise and contracting when they fall, because the intermediary’s spread is a function of price direction rather than volume.

Competitive intensity. High. Product is largely commoditized; competition is on availability, service level, delivered cost and price. Barriers to entry at the local level are modest — a warehouse, a saw, a truck and working capital. Barriers at national scale are meaningful: a 310-branch network, mill relationships built over decades, and the balance sheet to carry inventory through a downturn are not quickly replicated. Management’s own hope, expressed on the Q2 2026 call when asked whether recent consolidation would reduce competition, is revealing: “we are hoping that it will create a more disciplined environment, you know, with fewer competitors. We hope that it takes one bidder, one competitive bidder out, hopefully, if they focus on more pricing discipline.” Companies in structurally disciplined industries do not hope for discipline; they have it.

Regulation — the dominant variable. This industry’s economics are currently being set in Washington, not in the market. The sequence, per the FY2025 10-K:

• Early 2025: executive orders imposed Section 232 duties on steel and aluminum from Canada and Mexico, eliminated the tariff-rate quotas that had partially exempted certain countries, and eliminated product-specific exclusions.
• March 12, 2025: aluminum tariffs raised from 10% to 25%.
• June 2025: Section 232 steel and aluminum tariffs generally raised to 50%, and expanded to cover a range of additional “derivative” steel-containing articles.
• February 20, 2026: the Supreme Court ruled the president is not authorized to impose tariffs under the International Emergency Economic Powers Act (IEEPA). Critically, per the 10-K: “The Supreme Court’s ruling has no direct impact on the tariffs in place under Section 232, including tariffs on steel and aluminum.”

The distinction matters enormously and is the single most important thing to understand about RS today. IEEPA tariffs were struck down; Section 232 tariffs — the ones that govern steel and aluminum — survive, because they rest on a different statutory basis requiring a Commerce Department national-security investigation and finding. That makes them more durable than IEEPA tariffs. It does not make them permanent. They remain executive-branch policy, adjustable without legislation, and the 10-K itself concedes: “The current system of tariffs is fluid and the ultimate impacts of such tariffs on our revenues, financial results and cash flows will be based on a number of variables that are not known at this time.”

The effect on RS’s economics is direct and large. The CFO on the Q2 2026 call: “aluminum pricing has nearly doubled from the pre-tariff level.” Q2 aluminum price per ton was up 31.7% year-over-year. Domestic carbon steel prices are supported because imports are effectively priced out and mill lead times have extended. Management is explicit: “Trade policy continues to limit imports and support strong domestic pricing, which is further strengthened by extended lead times.”

Value-chain position. RS sits between concentrated, capital-intensive mills (Nucor, Steel Dynamics, Cleveland-Cliffs, Commercial Metals) and a highly fragmented customer base. It has weak bargaining power upstream on price — though, as discussed below, meaningful power on allocation — and moderate power downstream because its orders are small, urgent and a modest share of any customer’s cost base.

Switching costs and barriers. Customer switching costs are low in theory: another service center can quote the same grade. In practice they are non-trivial for small fabricators who value a single supplier that reliably has the material, will process it, and will deliver tomorrow — and who are, in RS’s phrase, ordering “more frequently and in smaller quantities.” That is convenience-based stickiness, not contractual lock-in.

Marathon capital-cycle read. The sector is currently earning tariff-protected returns well above cost of capital. Under Marathon’s supply-side framework, that condition attracts capital — mill capacity additions, service-center capacity expansion, and new entrants — which in time compresses the spread. RS itself is adding capacity: FY2026 capex of approximately $300M with about half allocated to growth. The offsetting factor, and it is a real one, is that the tariff wall constrains the most important source of supply response (imports). The capital cycle is being held open by policy. That is exactly why the durability of Section 232 is the thesis.

Verdict: a structurally mediocre industry currently enjoying an exceptional, policy-created cyclical peak. Fragmented, price-competitive, no input-cost control, returns that oscillate between 6.7% and 22.3% ROIC. Its redeeming features are low capital intensity relative to mills, counter-cyclical cash generation (inventory liquidates into cash in a downturn — RS generated $1,173M of operating cash flow in the 2020 collapse), and fragmentation that rewards a disciplined consolidator. Reliance is a very good operator in a bad-to-average industry, and investors should not confuse the quality of the operator with the quality of the industry.

4. Competitive Position

Does a moat exist? Yes — a real but bounded cost-and-access advantage. It is not customer captivity and it is not a network effect.

Naming the mechanism precisely matters, because the wrong label leads to the wrong multiple.

Mechanism 1 — Purchasing scale and privileged mill allocation. This is the strongest and most underappreciated element, and 2026 has made it visible. In a tight market, mills ration. COO Stephen Koch, Q2 2026 call: “based on our market position in beams and plate and some of our service centers that have been in this business for a long time, when supply gets a little bit tight, we get what we have been getting in the past years. You know, people do not like to use the word allocation. But when it is really tight, we get what we got in the past and also, when we had need of some favors or have some jobs to come up, we do get preferential treatment. So I would say that just the long track record really helps us in a market like this.”

This is a genuine, hard-to-replicate supply-side advantage. It is worth most precisely when it is scarcest — in a tight market — and it converts directly into share gain, because a competitor who cannot get metal cannot serve a customer at any price. It is the reason RS grew tons 6.2% while the industry shrank 1.0%.

Mechanism 2 — Balance sheet as a competitive weapon. Koch again, on why RS is taking share: “a lot of our competitors with higher interest rates and the higher cost of carrying inventory… there is a lot of holes in inventories, and our inventory levels are pretty robust. So I think that there is just great opportunity for us to capture more market share.” At 0.9x net debt/EBITDA with investment-grade access, RS can hold inventory through a price spike that forces leveraged competitors to run thin. Cost of capital is the moat in an inventory business.

Mechanism 3 — Network density. 310 branches produce shorter delivery radii, lower freight cost per order, and the ability to promise same- or next-day service on small orders. A new entrant can match one branch’s economics; it cannot match the delivered-cost curve of a national network without decades of investment.

Mechanism 4 — Decentralized operating model. RS runs ~75 subsidiaries under their own brands with local pricing and purchasing autonomy. Sales staff are paid base plus commission on gross profit from their territory — an unusually direct alignment that pushes margin discipline to the point of sale. This is a management-system advantage that is culturally hard to copy.

The financial test — does the moat show up in the numbers? Yes, and this is the strongest evidence. Gross margin by year:

Revenue moved by a factor of two; gross margin moved within a 6.8-point band and has averaged roughly 29.4%. That stability under extreme volume and price stress is precisely the outcome that should deteriorate if the purchasing-scale and service advantages were illusory. It does not deteriorate. That satisfies the test that a claimed moat must tie to a financial outcome that would degrade without it.

The Greenwald share-stability test — and why RS “fails” it informatively. Greenwald holds that stable market shares over long periods signal high barriers to entry, while shifting shares signal their absence. RS’s share is not stable — it is consistently rising, by more than 7 percentage points relative to the MSCI benchmark in 2025 alone. Strictly applied, that indicates the industry lacks strong barriers. The correct reading is not that RS has no advantage but that RS has a firm-specific advantage inside a low-barrier industry: it is a consolidator taking share from sub-scale, under-capitalized rivals, not an incumbent defending a fortified position. That distinction is important for valuation. Consolidator economics are real and can compound for a long time, but they do not command the multiple of a structurally protected franchise, because the underlying industry keeps regenerating competitors and the returns keep mean-reverting with the metal cycle.

Direct competitive comparison. RS’s true peers are largely private or divisions of larger groups (Ryerson, Russel Metals, Olympic Steel, O’Neal, Kloeckner, plus mill-owned distribution arms). Against the listed metals complex, the distinction is one of business model rather than degree: Nucor, Steel Dynamics, Commercial Metals and Cleveland-Cliffs are producers with mill assets, high fixed costs and direct exposure to spread between scrap/iron and finished steel. RS is asset-lighter, carries no blast furnace or EAF risk, and generates cash in downturns as inventory liquidates. That is a genuinely superior cash-flow profile through a cycle — RS produced $1,173M of operating cash flow in 2020 when revenue collapsed 37%. It is the single best argument for RS deserving a premium multiple to the producers.

Verdict: a durable but bounded competitive advantage — a cost-and-access advantage of the supply-side and scale-economies type, not customer captivity. It is real, it is visible in stable gross margins across a doubling of revenue, and it is currently producing measurable share gains. It is not, however, an advantage that immunizes earnings from the metal price cycle, and no amount of operational excellence changes the fact that RS does not set the price of what it sells.

5. Growth History and Forward Opportunities

Historical growth. Reported revenue growth is a poor guide to this business because it conflates price and volume. Decomposed:

*2023 tons/ASP derived; 2024–25 and Q2’26 figures are as disclosed.

The pattern is unmistakable. Volume growth is steady, modest and genuinely earned (roughly 5–11% per year, consistently above the industry). Price is the swing factor and it produces essentially all of the earnings volatility. In 2025, record volume was almost entirely offset by lower price. In 2026, volume and price are moving up together — an unusually favorable combination that has produced a 42% EPS increase.

Organic vs. acquired. RS discloses same-store metrics, which is analytically valuable. In 2025: total tons +6.2% versus same-store +5.3%; total net sales +3.3% versus same-store +2.6%. So roughly 85% of volume growth was organic and the four 2024 acquisitions contributed the remainder ($389.2M of 2025 sales). This is a business that grows primarily by taking share, supplemented by tuck-in M&A — not a roll-up dependent on deal flow.

Market-share capture is the core growth engine and it is verifiable. 2025 tons +6.2% against MSCI industry shipments −1.0% is a 7.2-point spread; on a same-store basis, +5.3% versus −1.0% is a 6.3-point spread that cannot be attributed to acquisitions. Management attributes this to scale, processing breadth, service levels and product/geographic diversity, and adds a specific and credible mechanism: trade-policy uncertainty “has led our customers to purchase more frequently and in smaller quantities which are core tenets of our operational strategy.” When customers de-stock and buy just-in-time, they need the supplier who always has material — which advantages the best-capitalized inventory holder.

Forward opportunities.

1. Data centers and electrical infrastructure. Repeatedly cited as the strongest demand vector in non-residential construction, with related pull-through in general manufacturing (“industrial machinery, including data center equipment”) and now semiconductors, where management sees “momentum accelerate, supported by increasing data center activities.” This is the most durable of the current tailwinds because it is capex-cycle driven rather than policy driven.

2. The DHS border wall contract. Phase 1 is approximately $1.4B of sales through mid-2027. Phase 2 is a further $800–900M that is at the customer’s option and explicitly not guaranteed — CEO Karla Lewis: “it is not guaranteed. We believe that they will probably execute that extension for phase 2.” Shipments began April 2026, contributed 5.1 percentage points of the Q2 sequential tons increase and $0.41 of Q2 EPS (against guidance of $0.15–0.20), and are expected to contribute ~$0.60 in Q3 at roughly 120,000 tons. Economically it is attractive: it carries a ~40bp negative gross-margin impact (low-processing, high-tonnage product) but a +30bp positive pretax margin impact, because RS services it off existing infrastructure at a variable cost per ton well below the company average. It is high-return incremental business — and it is finite.

3. Reshoring and defense. Shipbuilding, military and space activity are called out as strong; aerospace backlogs are growing as OEM build rates recover.

4. Tuck-in M&A. RS has completed acquisitions consistently for decades and closed four in 2024. None closed in Q2 2026 — consistent with a company that is finding assets expensive at the same moment it finds its own stock expensive.

5. Capacity investment. FY2026 capex of ~$300M, about half for growth: processing capability, footprint expansion, and volume growth in attractive markets.

Growth quality assessment. The volume growth is high quality: organic, share-based, verified against an independent industry benchmark, achieved without price concession (gross margin has not deteriorated while share was gained), and reinvested at attractive incremental returns. The earnings growth in 2026 is substantially lower quality, because roughly half of it comes from price, which RS does not control, and ~10% from a finite government contract. An investor buying today’s earnings run-rate is buying a volume franchise wrapped in a price windfall.

Verdict: high-quality volume growth, low-quality earnings growth. The share gains are real, repeatable and evidence of competitive strength. The 2026 profit surge is largely a policy-driven price event plus a one-off contract, and should not be extrapolated as a growth rate.

6. Financial Quality

Revenue and margin structure. Revenue is volatile and largely outside management’s control; gross margin is stable and largely within it. This is the defining financial signature of the company and is documented in the twelve-year table in Section 4. Gross margin has averaged roughly 29.4% over twelve years with a standard deviation of under two points, across a revenue range of $8.6B to $17.0B.

Operating leverage is real but modest. SG&A is roughly 19–20% of sales and is substantially variable (warehousing, delivery, freight, and commission tied to gross profit). In 2025, same-store SG&A rose 4.2% while same-store tons grew 5.3%, producing a 1.0% decline in same-store SG&A per ton — genuine operating leverage, though the dollar-margin effect is small relative to price swings. In Q2 2026, non-GAAP SG&A per ton was flat year-over-year despite inflation, aided by border wall volume.

Quality of earnings — the LIFO question is central and is where most analysis of this company goes wrong.

Reliance accounts for inventory on LIFO. In a rising-price environment, LIFO charges the most recent (highest) costs to COGS, depressing reported gross margin and earnings; in a falling-price environment it does the opposite. The swings are enormous:

The 2024→2025 swing was $258.1M. Reported gross margin fell 100bp to 28.7% while economic gross margin rose 80bp to 29.5%. An analyst reading only the reported line would conclude the business deteriorated in 2025; the opposite is true. Conversely in 2024, $144.4M of LIFO income flattered results.

In 2026 the effect is running the other way and has intensified. Management raised its full-year 2026 LIFO expense estimate from $150M to $300M, of which roughly $100M is aluminum-related. Q2 LIFO expense of $112.5M was triple the $37.5M guided, equal to $1.64 per share. Q3 is guided at $75M ($1.10/share). The LIFO reserve has grown from $548.6M at year-end 2025 to approximately $700M at June 30, 2026.

Three implications follow, and they cut in different directions:

1. Reported 2026 earnings understate current-period economics. Q2 non-GAAP EPS was $6.27 on a LIFO basis and $7.91 on a FIFO basis — a 26% difference.
2. The $700M LIFO reserve is a genuine, unrecognized economic asset. Inventory is carried roughly $700M below replacement cost. The CFO frames it correctly: it “remains available to support future operating results and help mitigate the impact of future metal price declines.” When prices eventually fall, that reserve unwinds into LIFO income, cushioning reported earnings. It is also, in effect, a deferred tax shelter.
3. But LIFO is the more honest measure for this business, and management’s preference for FIFO deserves scrutiny. The CFO says FIFO is “how we evaluate our ongoing performance.” For a distributor that must continuously replace inventory at current market cost to stay in business, LIFO — which charges COGS at approximately current replacement cost — is arguably the better representation of sustainable earning power. FIFO earnings in a rising-price market include inventory holding gains that are not repeatable. Investors capitalizing the $7.91 FIFO quarterly figure are capitalizing a windfall.

A second quality-of-earnings item — company-owned life insurance. RS carries COLI policies with associated policy loans. In 2025: investment income from life insurance policies of $107.3M, offset by interest expense on policy loans of $109.3M, cost of insurance of $18.8M, and income from policy redemptions of $13.0M, netting to a $7.8M expense. This is economically minor but it creates a data trap: at least one major aggregator (ROIC.ai) reports RS’s FY2025 interest expense as $165.0M, apparently by combining COLI policy-loan interest with debt interest. The 10-K reports actual interest expense of $55.7M (2025), $40.3M (2024) and $40.1M (2023). All interest figures in this memo are taken from the filing. Anyone screening RS on aggregator data will materially overstate its interest burden and understate its coverage.

Free cash flow. Reliance is cash-generative, but the pattern is counter-intuitive and must be understood correctly:

The inverse relationship between growth and cash is the key feature: in 2021, when revenue grew 60%, working capital consumed $909.7M and FCF fell to $562.8M despite record profit. In 2020, when revenue collapsed 37%, working capital released $400.3M and FCF hit $1,001M on the lowest net income of the period. This is a self-financing business in a downturn — a genuinely valuable and underappreciated characteristic that supports a premium to mill-owning producers. The corollary is that 2026’s volume-and-price surge is consuming working capital: H1 2026 operating cash flow was only $313.6M against H1 pretax income of $779.3M.

Capital intensity. Capex has exceeded D&A for four consecutive years ($468.8M, $430.6M, $328.9M versus D&A of $245.4M, $268.7M, $278.2M), reflecting a genuine investment cycle in processing capability. FY2026 is guided to ~$300M, roughly half growth. Maintenance capex appears to run around $150–170M, so the business is not capital-hungry by industrial standards — roughly 1.1–1.2% of sales at maintenance levels.

Balance sheet. Conservative and a competitive asset. At June 30, 2026: cash $235.4M, total debt $1,670.0M, net debt $1,428.5M, net debt/total capital 16.2%, net debt/EBITDA 0.9x. The revolver is drawn $520M of $1,500M available. Total equity was $7,179.5M at year-end 2025 against $10,373.3M of assets; goodwill of $2,169.9M and other intangibles of $960.1M mean tangible book is roughly $4.0B, or about $78/share — so the stock trades at roughly 5.2x tangible book. Working capital metrics are strong: inventory turns of 5.2x in Q2 2026 against a company goal of 4.7x, and DSO of 42 days, consistent year-over-year.

Dilution and SBC. Effectively nil dilution — the opposite. Stock-based compensation was $55.6M in 2025 (0.4% of sales, ~7.5% of net income), modest by any standard, and vastly outweighed by buybacks. Diluted shares fell from 78.6M (2014) to 51.4M (Q2 2026), a 34% reduction.

Returns on capital. ROIC ranged 6.7% (2014–16) to 22.3% (2022), landing at 8.8% in 2025 and rising in 2026. ROE 7.6% to 27.6%, at 10.1% in 2025. The honest assessment is that RS earns roughly its cost of capital at the trough and well above it at the peak, averaging perhaps 12–13% through a cycle. That is a decent but not exceptional return profile, and it is the single most important corrective to the “high-quality compounder” framing that the current multiple implies. This is not a 20%-ROIC franchise; it is a low-teens-through-cycle business that touches 20%+ when metal prices spike.

Verdict: do economics improve with scale? Modestly yes, but the cycle dominates. Scale delivers stable gross margin, improving SG&A per ton, superior mill access and a cost-of-capital edge — all real. But those advantages move operating margin by one to two points; the metal price cycle moves it by seven. Financial quality is high in terms of balance-sheet strength, cash conversion, working-capital discipline and the near-total absence of dilution or accounting aggression. It is moderate in terms of returns on capital, which are cyclical and average only low-teens.

7. Capital Allocation

This is the strongest section of the Reliance story, and it deserves credit before criticism.

Share repurchases — excellent execution, and a loud current signal. Since 2021 RS has repurchased approximately $3.4B of stock at an average price of roughly $234 per share, reducing shares outstanding by about 22%. Against a current price of $407.83, that is a realized gain of roughly 74% on deployed capital. Over the longer arc, diluted shares fell from 78.6M in 2014 to 51.4M in Q2 2026 — a 34% reduction, meaning a shareholder who did nothing now owns 1.53x as much of the company as in 2014.

Annual repurchases: $323.5M (2021), $630.3M (2022), $479.5M (2023), $1,093.7M (2024), $594.1M (2025), $234.2M (Q1 2026), and $0 in Q2 2026.

That last figure is the most important number in this memo after the valuation percentile. In the quarter in which Reliance reported record tons sold, a 42% EPS increase, and raised guidance — with $529M of remaining authorization, 0.9x net debt/EBITDA, and an investment-grade revolver drawn only $520M of $1,500M — the company bought no stock at all. The CFO’s language was deliberate: “We did not repurchase any shares of our issued or outstanding common stock during the quarter… We remain opportunistic in our approach.”

Management has demonstrated over five years that it knows how to buy its own stock well, at an average of $234. It bought $234.2M in Q1 2026, when the stock traded roughly $300–360. It stopped completely once the stock cleared ~$360. The best-informed buyer of RS equity, with capacity, authorization and a proven track record, has declined to buy at current levels. No external valuation work is more informative than that.

Dividends — exceptional consistency. 66 consecutive years of dividend payments, never suspended or cut, spanning every recession since 1960. The quarterly rate rose 10.0% in 2024 and 9.1% in 2025; the most recently declared dividend is $1.25/share payable August 28, 2026. Dividend per share has grown from $1.40 (2014) to $4.85 (2025), a 3.5x increase, while the payout ratio has stayed conservative at 34.2% in 2025 (and only 11.7% at the 2022 earnings peak). This is a sustainable, well-covered and genuinely long-duration commitment.

M&A — disciplined and, unusually, verifiable as such. Four acquisitions closed in 2024: Cooksey Iron & Metal (February 1), American Alloy Steel (April 1), Mid-West Materials (April 1), and certain assets of Ferragon’s FerrouSouth division (August 16). Aggregate net assets acquired: $373.0M, of which only $59.5M was goodwill — 16% of consideration. The acquired businesses generated $389.2M of net sales in 2025, implying roughly 0.96x sales paid.

This is the antithesis of the serial-acquirer pattern that destroys value. RS is buying inventory ($109.9M), property/plant/equipment ($107.5M) and receivables ($44.9M) — hard assets at close to fair value — rather than paying strategic premiums that inflate goodwill and depress future returns on capital. The 10-K adds a discipline statement worth quoting: “We have not diversified outside our core business of providing metal distribution and processing solutions since inception.” Eighty-seven years without strategic drift is a meaningful data point about management character. No acquisitions closed in Q2 2026 — consistent with a buyer that finds assets expensive at the same time it finds its own stock expensive.

One legitimate criticism — capital returns outran free cash flow in 2024–25. Combined 2024–2025: buybacks $1,687.8M + dividends $504.4M = $2,192.2M returned against combined free cash flow of $1,501.7M. The $690M gap was funded by drawing cash from $1,080.2M (year-end 2023) to $216.6M (year-end 2025) and raising net debt from $62.0M to $1,204.3M (now $1,428.5M).

This was defensible — leverage remains trivial at 0.9x, and the stock was demonstrably cheap at an average of $234 — but it is a pro-cyclical pattern. RS spent its balance-sheet flexibility buying stock at $234, which was right, and now holds less dry powder at a moment when management itself judges the shares too expensive to buy. Had metal prices stayed depressed through 2026 instead of inflecting, the 2024 pace of repurchase would look aggressive rather than opportunistic. The outcome was excellent; the process carried more risk than the outcome reveals.

Incentive alignment — above average, with two identifiable flaws.

Annual cash incentive: 80% weighted to Pretax Income Margin, 20% to Tons Sold Growth measured against the MSCI industry benchmark (raised from 10% in 2024 to 20% in 2025). NEO target is 150% of base salary. 2025 outcomes: Pretax Income Margin of 6.91% earned 96.4% of base salary; Tons Sold Growth of 7.27% above the MSCI benchmark earned the maximum 60% of base salary. No payout occurs on the tons metric unless RS both grows tons and beats the industry benchmark.

Long-term equity: 80% of CEO/CFO/COO target equity is performance-based (70% for other NEOs), tied to a three-year ROA target: threshold 7% (25% vesting), target 10% (100%), maximum 13% (200%). All performance awards use three-year measurement periods.

Governance hygiene: No employment agreements. No change-in-control or golden-parachute agreements. No perquisites. No tax gross-ups. No hedging permitted. Clawback policy covering all incentive compensation. Independent non-executive Chair. CEO target annual cash incentive is set at approximately the 25th percentile of the peer group — genuinely unusual restraint. CEO stock ownership is 26.9x base salary against a 5x requirement; COO and CFO at 8.2x against 4x.

The two flaws. First, Pretax Income Margin — 80% of the annual bonus — is largely determined by the metal price cycle rather than by management action. In 2025, margin fell to 6.91% because ASP declined and LIFO swung by $258M; in 2026 it will rise sharply because tariffs lifted prices. Management is being paid, in the main, on the tape. Second, the long-term ROA metric has a low bar and is an accounting return. A 7% threshold against a twelve-year average ROA of roughly 9–10% (and 18.5% in 2022) means threshold vesting is close to automatic in all but a severe downturn; even the 10% target was exceeded in both 2024 (11.57%) and 2025 (10.15%). ROA also uses total assets including $2.17B of goodwill and $960M of intangibles, which at least does penalize overpaying for acquisitions — a genuine virtue — but it is not a cash-on-cash return and there is no explicit ROIC or relative-TSR hurdle anywhere in the plan.

The redeeming feature is the Tons Sold Growth versus MSCI metric: it is a true relative measure that cannot be earned by riding the cycle, and the Committee deliberately doubled its weight. That is the right direction of travel.

Verdict: management has allocated capital intelligently — this is a clear yes. Buybacks executed at prices that produced a 74% gain and were then halted on valuation discipline; 66 years of uninterrupted dividends with a conservative payout; tuck-in M&A at ~0.96x sales with minimal goodwill; no strategic diversification in 87 years; genuinely restrained executive pay. The criticisms are real but second-order: capital returns temporarily outran FCF in 2024–25, and 80% of the annual bonus keys off a margin metric that the metal cycle largely sets. Against the base rate for cyclical industrials, this is top-decile stewardship.

8. Changes and Headwinds — Last Two Years

1. The tariff regime became the dominant economic variable (2025–26). Early 2025 brought Section 232 duties on Canadian and Mexican steel and aluminum, elimination of tariff-rate quotas and product exclusions; aluminum went 10%→25% on March 12, 2025; and in June 2025 steel and aluminum tariffs were generally raised to 50% and extended to derivative articles. Aluminum pricing has “nearly doubled” from pre-tariff levels. Thesis impact: strongly positive near-term, and the single largest source of risk.

2. The Supreme Court’s February 20, 2026 IEEPA ruling. The Court held the president lacks authority to impose tariffs under IEEPA. Section 232 tariffs were explicitly unaffected. Thesis impact: net positive but with a warning attached — it removed one legal threat while demonstrating that tariff authority is judicially contestable. Section 232 rests on firmer statutory ground but remains executive policy.

3. The DHS border wall contract (awarded early 2026, shipping from April 2026). Phase 1 ~$1.4B of sales through mid-2027; Phase 2 a further $800–900M at customer option, not guaranteed. Contributed $0.41 of Q2 EPS against $0.15–0.20 guided; ~$0.60 expected in Q3 on ~120,000 tons. Margin-dilutive at the gross line (−40bp) but pretax-accretive (+30bp) because it runs off existing infrastructure. Thesis impact: positive but finite, and being capitalized as if permanent.

4. The LIFO whipsaw (2024→2026). $144.4M of income in 2024, $113.7M of expense in 2025, and a full-year 2026 estimate doubled from $150M to $300M mid-year, with the reserve building from $548.6M to ~$700M. Thesis impact: neutral economically, materially negative for reported-earnings legibility. It has made reported results a poor guide to underlying performance in both directions.

5. Sustained market-share capture. 2025 tons +6.2% (same-store +5.3%) against MSCI industry shipments −1.0%; Q2 2026 tons +10.8% and again “significantly exceeding” industry trends. Driven by inventory availability while leveraged competitors run thin, and by customers shifting to smaller, more frequent orders. Thesis impact: positive and the most durable of the current tailwinds.

6. The 2024 acquisition program and the 2026 pause. Four deals in 2024 for $373.0M at ~0.96x sales; zero in Q2 2026. Thesis impact: neutral-to-positive — evidence of price discipline rather than a lack of opportunity.

7. Company identity repositioning (February 15, 2024). Renamed from Reliance Steel & Aluminum Co. to Reliance, Inc. to signal a “diversified metal solutions provider.” Thesis impact: cosmetic. The business mix is unchanged — carbon steel has been exactly 53% of sales for three consecutive years. Investors should not read a business-model change into a name change.

8. End-market rotation. Non-residential construction has been carried by data centers and energy infrastructure rather than traditional commercial building. Semiconductors have gone from weak to “meaningfully improved.” Aerospace is recovering with growing backlogs but persistent elevated inventories. Automotive toll processing is steady. Thesis impact: positive, and healthily diversified — no single end market exceeds one-third.

9. Cost headwinds. Q2 2026 non-GAAP SG&A rose 11% year-over-year on higher incentive compensation (a function of higher profit), wage inflation, and — a new and specific item — freight and fuel cost inflation attributed to the U.S.–Iran conflict, which management also names as a forward risk to the Q3 outlook. Thesis impact: modestly negative, and a reminder of exogenous exposure.

10. Rising capital intensity. Capex ran $468.8M (2023), $430.6M (2024), $328.9M (2025) against D&A of $245.4M/$268.7M/$278.2M — four consecutive years above D&A. FY2026 guided to ~$300M, about half growth. Thesis impact: neutral — consistent with a company investing into share gains, but it does compress near-term FCF.

Verdict: on balance these developments have strengthened the business and weakened the margin of safety. The share gains, the M&A discipline and the buyback execution are genuine improvements in franchise quality. But the earnings power that the market is now capitalizing at a record multiple rests on two developments — a 50% tariff regime and a finite government contract — that are, respectively, reversible policy and explicitly time-limited. The company got better; the price got better faster.

9. Risk Analysis

Catastrophic-loss assessment. The probability of permanent capital impairment from a business failure is very low. RS has been profitable every year since 1994 through two financial crises and a pandemic; it holds investment-grade credit at 0.9x leverage; its inventory liquidates into cash during downturns (2020: $1,173M of operating cash flow on collapsing revenue); it has no single-product, single-customer or single-technology dependency; and it has no meaningful off-balance-sheet or pension exposure ($24.9M pension liability). The dominant risk here is not solvency or franchise destruction — it is paying too much. The realistic bear outcome is a multi-year period of flat-to-negative returns as earnings normalize from a policy-inflated peak and a record multiple compresses toward its historical range, not a permanent loss of capital.

10. Valuation Discussion

No price target and no recommendation appears in this section. The analysis addresses only what the current price embeds.

Where the stock trades. At $407.83 (July 24, 2026) on approximately 51.375M diluted shares, market capitalization is roughly $20.9B; with net debt of $1,428.5M, enterprise value is approximately $22.4B.

The central valuation fact. On trailing data, RS trades at:

Reliance has never, in the available decade of history, been this expensive on any of the three primary multiples — let alone all three simultaneously. This is corroborated by an independent eleven-year series. Year-end P/B: 1.07 (2015), 1.35 (2016), 1.32 (2017), 1.07 (2018), 1.51 (2019), 1.48 (2020), 1.67 (2021), 1.71 (2022), 2.09 (2023), 2.05 (2024), 2.09 (2025). The highest intra-year closing P/B in that entire record is 2.60 (2024). Today’s 2.98x exceeds every observation. The same holds on P/S: the highest prior intra-year close was 1.38 (2024) against 1.45 today.

Important context that cuts against an RS-specific mispricing claim. The whole complex is elevated on own-history terms: GWW 99.8th percentile, STLD 97.3rd, FAST 96.1st, WCC 93.0th, NUE 92.9th, CMC 83.1st. RS is the most extreme, but this is a sector-wide re-rating, not an idiosyncratic RS anomaly. That matters: it means the market is pricing a durable, tariff-protected North American metals pricing regime across the entire value chain. An investor betting against RS’s multiple is betting against that thesis broadly, not against a single mispriced security. It also means relative-value screens will not flag RS, because its peers have re-rated alongside it.

Earnings basis — and which number to use.

FY2026E is built from disclosed H1 non-GAAP EPS of $11.42 plus the Q3 guidance midpoint of $6.50 plus a normal seasonal Q4. The FIFO figure adds back the guided $300M of FY2026 LIFO expense (roughly $4.47/share after tax).

Which basis is right is the crux of the valuation debate. Bulls point to ~14.6x FIFO earnings and call the stock cheap. That is the wrong lens for a going-concern distributor. FIFO earnings in a rising-price market embed inventory holding gains — profit earned on metal bought before the price rose — which by definition cannot repeat unless prices keep rising at the same rate. LIFO, which charges cost of sales at approximately current replacement cost, is the better proxy for sustainable earning power in a business that must continuously restock at market. Management’s stated preference for FIFO (“how we evaluate our ongoing performance”) is understandable but self-serving in a year when FIFO flatters results by 26%.

Embedded-expectations analysis — what must be true at $407.83?

Reliance’s historical mid-cycle P/E is 12–14x (year-end P/E averaged roughly 13x across 2015–2023, spanning 6.7x at the 2022 earnings peak to 20.9x at the 2020 trough). Applying that historical multiple to today’s price implies the market is underwriting normalized EPS of approximately $29–34.

For that to be correct, essentially all of the following must hold:

1. Section 232 at 50% persists indefinitely, keeping domestic steel and aluminum prices structurally elevated and imports excluded;
2. Average selling price stays at or above the ~$2,600/ton Q2 2026 level rather than reverting toward the $2,244 of 2025 or the $2,303 of 2024;
3. Volume growth continues at mid-single digits or better as RS keeps taking share from the industry;
4. The border wall contract’s Phase 2 is exercised and comparable large contracts recur;
5. Share count continues shrinking at 3–4% annually.

Points 3 and 5 are well supported by evidence. Point 4 is explicitly uncertain by management’s own statement. Points 1 and 2 are the load-bearing assumptions, and they are not company-specific — they are bets on U.S. trade policy.

Alternatively, one may accept a higher mid-cycle multiple on the argument that the business has genuinely improved — a 34% smaller share count, a structurally higher gross margin (29–31% recently versus 25–28% in 2014–18), verified share gains, and demonstrated double-digit ROIC through a downturn. That argument has real merit, and it is likely that the 2.09x year-end P/B of 2023–2025 understated fair value. But moving from a historical 1.5–2.1x P/B to today’s 2.98x requires more than “somewhat better” — it requires a step-change in the character of the business, and carbon steel is still exactly 53% of sales, exactly as it was in 2023.

What the market is getting right. The volume franchise is real and improving. Working capital converts to cash in downturns, which genuinely deserves a premium to mill-owning producers. Capital allocation has been top-decile. Leverage is trivial. Consolidation of a fragmented industry by the best-capitalized player is a multi-year compounding opportunity.

What the market may be getting wrong. First, capitalizing a finite government contract — roughly 9–10% of current EPS, with Phase 1 ending mid-2027 and Phase 2 unguaranteed — at 26x. Second, treating a 50% tariff set by executive order as a permanent feature of normalized earnings. Third, the classic cyclical error of applying a peak multiple to peak-spread earnings, which is how cyclicals inflict their worst damage. Fourth, potentially anchoring on FIFO earnings that embed non-repeatable inventory holding gains.

Scenario analysis (illustrative; explicitly not price targets).

The salient observation is that the bull case is required merely to justify the current quote, not to profit from it. That is the definition of a fully-valued security.

Cash-flow cross-check. 2025 FCF of $502.5M on a $20.9B market cap is a 2.4% FCF yield; 2024’s $999.2M would be 4.8%; the 2022 peak of $1,776.8M would be 8.5%. Even using the best FCF year of the last five, the yield does not compensate for cyclical earnings risk. The dividend yield is approximately 1.2%.

Sum-of-the-parts. Not warranted. RS operates one reportable segment with a single economic engine; the ~75 subsidiary brands share purchasing, credit and capital allocation and are not separable in any economically meaningful way.

Verdict on valuation. The current price embeds an assumption that the 2026 tariff-supported pricing environment is the new normal rather than a cyclical and policy-driven peak, and it applies a record multiple to that assumption. The business quality justifies a premium to its own history; the magnitude of the premium currently on offer requires the bull case to be correct in full.

11. Variant Perception

The consensus view. Sell-side and financial-media coverage is constructive-to-bullish and has been rising with the tape: record shipments, market-share gains, a 42% EPS increase, a Q3 guide of +76–81% year-over-year, and an “acquisitions plus infrastructure demand” growth narrative. The stock is up ~34% year-to-date and ~42% since December 2025 and sits 1.4% off an all-time high. Consensus holds that Reliance is a high-quality, well-managed consolidator whose earnings are inflecting upward on reshoring, data-center construction and favorable trade policy, and that share gains plus buybacks justify the re-rating. Notably, one gatekeeper flagged the stock as “still overvalued” on a valuation-model basis in late June, so the view is not unanimous.

The strongest bull case. Reliance is being re-rated because it has genuinely become a different quality of business, and the market is correct to notice. Consider what the last five years demonstrate: a 22% reduction in share count executed at an average of $234 (a 74% realized gain); gross margin that has held 29–31% versus 25–28% in 2014–2018; tons sold growth exceeding the industry benchmark by more than 7 points in a year the industry shrank; and ROIC that troughed at 8.8% in 2025 — a genuinely bad year — versus 6.7% in the equivalent 2014–16 trough. That is a structurally higher floor. Layer on a durable North American reshoring cycle, data-center-driven infrastructure demand, a tariff wall that has repriced domestic metal upward and shows no sign of political vulnerability (having survived the Supreme Court’s IEEPA ruling), and a fragmented industry consolidating into the hands of the best-capitalized operator — and a business that used to deserve 12x might reasonably deserve 17–18x. On FY2026 FIFO earnings the stock is at 14.6x, which for a share-gaining, cash-generative, 0.9x-levered compounder is not obviously expensive. The bull adds: valuation percentiles are backward-looking and will always say “expensive” at the moment a business re-rates for good reasons.

The strongest bear case. Reliance does not control the price of what it sells, and roughly all of the incremental 2026 earnings come from price. Q2 ASP rose 14.5% against volume up 10.8%, and that price move is a direct consequence of a 50% Section 232 tariff imposed by executive order in June 2025 — a policy, not a competitive advantage. Management states in its own 10-K that pricing “has a much more significant impact on our results of operations than customer demand.” Roughly 9–10% of EPS comes from a border wall contract that ends mid-2027 unless a customer exercises an option management concedes “is not guaranteed.” The market is capitalizing this at 26.6x earnings, 2.98x book and 1.45x sales — each the highest in the ten-year record, all three at the 99.86th percentile simultaneously. Meanwhile the two most informed constituencies are not buying: no officer or director has made an open-market purchase in two years, and the company itself, holding $529M of authorization and 0.9x leverage after reporting record results, repurchased zero shares in Q2. This is the textbook cyclical trap — peak-spread earnings on a peak multiple — with the added fragility that the peak is policy-created and therefore revisable by an act that requires no legislation.

The 3–5 assumptions that actually matter.

1. Does Section 232 at 50% persist? The single highest-leverage variable. It is not a company question; it is a policy question, and no amount of fundamental work resolves it.
2. Is ~$2,600/ton the new normal ASP, or is ~$2,250–2,400 the mid-cycle? This determines whether normalized EPS is ~$28 or ~$20 — a 40% difference that maps almost one-for-one into fair value.
3. Is the market-share gain durable or a function of a tight market? RS gains share partly because it can obtain metal when competitors cannot. If supply loosens, the advantage attenuates. The counter-evidence is that RS also gained share in 2025, a loose year — which suggests the advantage is at least partly structural.
4. Does the border wall recur, or is it a one-off? Phase 2 is $800–900M at customer option. Management “believes they will probably execute” — a belief, not a contract.
5. Has the business earned a structurally higher multiple? A smaller share count, higher gross margin and a higher trough ROIC argue yes. Unchanged product mix (carbon steel exactly 53% for three years) and unchanged cyclicality argue the re-rating has overshot.

Falsification tests.

What would falsify the bear case: Section 232 codified in legislation rather than executive order; FIFO gross margin sustaining above 30% through a period of flat or falling metal prices (proving spread expansion is structural, not a holding gain); border wall Phase 2 exercised and a second comparable large contract awarded; and, most decisively, RS resuming meaningful share repurchases at $380–420 — which would signal that the people with the best information consider these prices reasonable.

What would falsify the bull case: FIFO gross margin breaking below ~29% while volumes hold, indicating the spread was an inventory-timing artifact; a material Section 232 reduction or broad exclusion regime; ASP reverting toward $2,300 with EPS following down toward $15–17; continued absence of both insider and corporate buying through another two quarters of record results; or LIFO expense guidance being raised again, which would signal input costs are rising faster than RS can pass through.

Where consensus may be offsides — the positioning read. The factor evidence complicates the simple contrarian story and deserves honest treatment. Statistically RS does not look like a crowded speculative trade: the factor model reads it as a low-beta dividend/shareholder-yield vehicle (DividendYield beta 0.794 is its single largest loading; market beta 0.650–0.704), its factor-similar peers include shareholder-yield ETFs (SYLD 0.918, SDVY 0.911, REGL 0.907) alongside CMC (0.937), NUE (0.911) and STLD (0.908), and its risk-adjusted record is excellent and orderly: a 1.34 Sharpe over the past year with a maximum drawdown of only 19.0%, and a 0.82 Sharpe over five years. The three-month annualized return of 83.3% at a 3.09 Sharpe is extreme, but it is earnings-supported rather than narrative-driven.

The honest synthesis is this. This is not a bubble and it is not a falling knife. It is a good business, held largely by yield- and quality-oriented owners rather than momentum speculators, that has delivered real earnings and re-rated to the top of its historical range on the back of a policy-created pricing environment. The variant perception is not “the market is wrong about the business” — the market is broadly right about the business. It is that the market has quietly converted a cyclical spread business into a quality compounder in its own mind, and is paying a quality-compounder multiple for earnings whose principal driver is a 50% tariff and a border wall contract. The mispricing, if it exists, is in the durability assumption, not the quality assumption. And the sharpest tell that this may be so is that the company itself — the most informed, best-incentivized and historically most skilled buyer of its own shares — stopped buying.

12. Fact vs. Interpretation

13. Open Questions

1. How durable is Section 232 at 50%? The dominant variable in the thesis and one that fundamental analysis cannot resolve. Section 232 rests on firmer statutory ground than the IEEPA authority struck down in February 2026, but it remains executive policy adjustable without legislation. Is there any live proceeding, Commerce review, or trade negotiation that could modify the rate or reintroduce exclusions?

2. Why exactly did the buyback stop in Q2 2026? Valuation discipline is the natural reading and the one adopted here, but alternatives exist: preserving capacity for a large acquisition, funding the border wall working-capital build, or a blackout-period technicality. Management said only “we remain opportunistic.” Direct confirmation would materially sharpen the signal.

3. What is the true normalized average selling price? 2024 was $2,303, 2025 was $2,244, Q2 2026 was $2,602. Whether mid-cycle is $2,300 or $2,600 determines whether normalized EPS is ~$20 or ~$28. This turns almost entirely on question 1.

4. What is the gross profit per ton trend, stripped of price effects? Management asserts gross profit per ton is “up significantly from a couple of years ago” and that the percentage-margin optics are distorted by aluminum. The company does not disclose gross profit per ton directly. This single disclosure would separate genuine spread improvement from inventory-holding gains and is the most valuable missing datum.

5. Will border wall Phase 2 be exercised? $800–900M of incremental sales at the customer’s option. Management “believes they will probably execute” it. What are the conditions precedent, and is there any contractual visibility?

6. How much of the 2025–26 share gain persists if supply loosens? RS gains share partly because it can source metal when competitors cannot. The encouraging counter-evidence is that it also gained 7 points of relative share in 2025, a loose year — but the mechanism may differ.

7. What is normalized maintenance capex? Capex has exceeded D&A for four straight years. The split between maintenance and growth (~50/50 per FY2026 guidance) implies maintenance of roughly $150M, but this is inferred rather than disclosed, and it materially affects any owner-earnings calculation.

8. What is the economic rationale for the COLI program? $107.3M of investment income offset by $109.3M of policy-loan interest plus $18.8M cost of insurance nets to an expense. It is likely a deferred-compensation funding vehicle, but the structure obscures reported interest expense and has already caused at least one major data vendor to misstate RS’s interest burden by 3x.

9. How is the aluminum LIFO distortion expected to resolve? Aluminum is 17% of sales but ~one-third of the $300M FY2026 LIFO estimate. If aluminum prices plateau, does LIFO expense fall sharply in 2027, mechanically boosting reported EPS?

10. What is the acquisition pipeline at current asset prices? Zero deals closed in Q2 2026. Is this discipline in an expensive market, or an absence of available targets?

14. What Must Be True

Bull case — what must be true

1. Section 232 tariffs at or near 50% remain in force through at least 2028, keeping domestic steel and aluminum prices structurally elevated and imports substantially excluded.
2. Average selling price sustains at or above roughly $2,500/ton, versus $2,244 in 2025.
3. Volume growth continues at mid-single digits or better, with RS beating the MSCI industry benchmark by 3+ points annually.
4. FIFO gross margin holds at or above 30%, demonstrating that spread expansion is structural rather than an inventory-timing artifact.
5. Border wall Phase 2 is exercised, and comparable large infrastructure or defense contracts recur.
6. Share count continues shrinking 3–4% annually, requiring management to resume buybacks at current or higher prices.

Falsification test for the bull case: Watch FIFO gross margin against average selling price direction over the next two to three quarters. If ASP flattens or declines while FIFO gross margin holds above 30%, the bull case is validated — the spread is structural and RS has genuinely repriced its service. If FIFO gross margin falls below ~29% as ASP flattens, the 2026 margin was an inventory holding gain and normalized EPS is nearer $20 than $28. A second, cleaner test: does the company resume repurchasing stock at $380–420? If it does, management’s revealed valuation opinion has changed and the single strongest bear datapoint is removed.

Bear case — what must be true

1. The tariff-driven price level proves temporary or partially reverses, whether by Section 232 modification, exclusion reintroduction, negotiated country carve-outs, or simple demand destruction at 50% tariff-inflated prices.
2. ASP reverts toward $2,250–2,400, compressing both revenue and the realized spread.
3. The border wall contribution ends with Phase 1 in mid-2027 and Phase 2 is not exercised, removing ~9–10% of EPS with no offsetting replacement.
4. The market re-rates RS back toward its historical multiple range (1.5–2.2x book, 13–16x mid-cycle earnings) as the earnings driver is recognized as cyclical and policy-dependent.
5. Management’s own reticence proves prescient — the buyback pause and the two-year absence of insider buying mark a valuation peak rather than a pause.

Falsification test for the bear case: Watch for Section 232 being placed on a durable statutory footing (codifying legislation, or a formal long-horizon Commerce determination with no exclusion process) combined with RS resuming buybacks at prices above $380. Either alone is suggestive; together they would demonstrate that (a) the policy rent is structural, not transient, and (b) the best-informed buyer regards current prices as reasonable — which would substantially dismantle the bear case. A supporting confirmation would be FY2027 guidance built on ASP at or above $2,600 without a demand concession.

15. Source Appendix

See Appendix B — Source Appendix below, which lists every public primary source relied upon.

Sections 1–15 contain no investment recommendation and no price target; the Claude's Take block is explicitly labeled as the author’s own subjective view. This article is general information only and is not investment advice. Every factual claim is sourced to the public filings and company disclosures listed in Appendix B.

APPENDIX A — Standard Diligence Questionnaire

Reliance, Inc. (NYSE: RS) — July 25, 2026

A standard diligence questionnaire applied to Reliance. Fact / Interpretation / Assumption labels are applied where the distinction matters.

General

What thoughtful questions have other investors asked about this company?

From the Q2 2026 earnings call, the sell-side pressed on four themes, and the quality of the questions is a fair guide to where the debate sits:

• Border wall durability and margin structure (Bank of America, Goldman Sachs, BMO). Analysts probed whether the ~$0.60/share Q3 run-rate holds, whether the low-opex structure persists at higher volumes, and whether Phase 2 extends beyond mid-2027. Management confirmed Phase 1 is ~$1.4B through mid-2027 and that Phase 2 ($800–900M) “is up to our customer to opt in… it is not guaranteed.” (FACT)
• Inventory positioning in a tight market (KeyBanc). Why did inventory rise less than $100M despite a $600M revenue increase? Answer: turns ran 5.2x versus a 4.7x company goal, and industry-wide inventories are thin. (FACT)
• The LIFO/aluminum distortion (JPMorgan). Why did LIFO expense triple versus guidance? The CFO quantified aluminum as ~one-third of the $300M full-year estimate on 17% of sales, and disaggregated roughly 100bp of “margin compression noise” into ~50bp FIFO and ~50bp LIFO. (FACT)
• Whether buyers are pushing back on price (JPMorgan, Wells Fargo). Management said customers accept prices while they can pass them through, and that competitors’ inventory gaps are creating share-capture opportunities. (FACT)

The question the sell-side did not ask, and which is arguably the most important: why did the company repurchase zero shares in the quarter while reporting record results with $529M of authorization and 0.9x leverage? (INTERPRETATION: this omission is itself notable — the market is focused on the earnings trajectory rather than on the price being paid for it.)

A second under-asked question: management does not disclose gross profit per ton, which is the cleanest measure of whether spread improvement is structural or an inventory-timing artifact. Management asserts it is “up significantly” but provides no series.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low?

Closer to a high, though not at the 2022 extreme. (INTERPRETATION, well supported.) Twelve-year pretax margin history: 4.9% (2015), 5.0% (2016), 6.0% (2017), 7.4% (2018), 8.5% (2019), 5.4% (2020), 13.4% (2021), 14.3% (2022 peak), 11.8% (2023), 8.2% (2024), 6.8% (2025), 9.3% (Q2 2026). Twelve-year average is roughly 8.4%. Q2 2026 is comfortably above average and rising; Q3 guidance of $6.40–6.60 (+76–81% YoY) implies further expansion. So earnings are above mid-cycle and accelerating, but still well below the 2022 peak. Diluted EPS: $29.92 (2022) → $22.64 (2023) → $15.56 (2024) → $13.98 (2025) → ~$23–24 estimated for 2026.

Driven by the external environment or internal actions?

Predominantly external. (FACT, per management’s own disclosure.) The FY2025 10-K states: “Pricing for our products generally has a much more significant impact on our results of operations than customer demand.” Q2 2026 average selling price rose 14.5% year-over-year against volume up 10.8%, and the price move traces to Section 232 tariffs raised to 50% in June 2025. Aluminum pricing has “nearly doubled” from pre-tariff levels per the CFO.

The genuinely internal contributions are real but smaller: market-share capture (2025 tons +6.2% against an MSCI industry decline of 1.0%), SG&A per ton discipline (−1.0% same-store in 2025), the border wall contract win, and a 22% share-count reduction since 2021. (FACT) A reasonable decomposition is that roughly two-thirds of the 2026 earnings surge is price/policy and one-third is company execution. (ASSUMPTION)

How stable are revenues?

Volatile at the top line, stable at the gross-margin line — this is the defining characteristic. Revenue ranged $8.6B (2016) to $17.0B (2022) over twelve years. Gross margin over the same period ranged only 25.1% to 31.9%, averaging ~29.4%. (FACT) Customer diversification is exceptional: 125,000+ customers, no disclosed concentration, no single end market above one-third.

Outlook for products/services?

Constructive near-term. Q3 2026 guidance is non-GAAP EPS of $6.40–6.60. Management cites continued strength in data centers, energy infrastructure and public infrastructure; “meaningfully improved” semiconductor demand; recovering commercial aerospace with growing backlogs; steady automotive toll processing; and healthy general manufacturing subject to normal seasonality. Risks named by management: trade policy and the U.S.–Iran conflict. (FACT)

How big will this market be — growing, shrinking, domestic or international?

Overwhelmingly domestic: approximately 310 locations, 41 U.S. states, 10 foreign countries, with international presence limited (Canada, Mexico, UK, Singapore, Malaysia, Australia). The addressable market is not growing in volume terms — MSCI industry shipments declined 1.0% in 2025. (FACT) RS grows by taking share from a fragmented field, not by riding market expansion. (INTERPRETATION) Dollar market size expands and contracts with metal prices independent of tonnage.

Business Quality & Competitive Moat

Is the industry getting more or less competitive?

Marginally less, but from a highly competitive base, and management’s framing is revealing. Asked directly about consolidation on the Q2 call, CEO Lewis said: “we are hoping that it will create a more disciplined environment, you know, with fewer competitors. We hope that it takes one bidder, one competitive bidder out.” (FACT) Companies operating in structurally disciplined industries do not hope for discipline. (INTERPRETATION) Separately, competitors’ balance-sheet stress — “higher interest rates and the higher cost of carrying inventory… a lot of holes in inventories” — is currently reducing effective competitive intensity in a way that may not persist when rates fall.

How profitable is the business (ROIC, ROE)?

Cyclically, and only adequately through the cycle. (FACT)

The honest reading: RS earns roughly its cost of capital at the trough and well above it at the peak. (INTERPRETATION) This is the single most important corrective to the “high-quality compounder” framing implied by a 26.6x multiple. It is a low-teens-through-cycle business, not a 20%-ROIC franchise.

How profitable is the industry — how many competitors, what barriers to entry?

Fragmented with hundreds of participants; RS is the largest in North America at $14.29B of revenue and management’s own proxy states there are “no public companies in the metals service center industry that are closely comparable.” (FACT) Barriers at the local level are low (a warehouse, a saw, a truck, working capital). Barriers at national scale are material: a 310-branch network, decades-old mill relationships that translate into allocation preference in tight markets, and a cost of capital that permits carrying inventory through a cycle. (INTERPRETATION)

Can the business be easily understood?

Yes — this is a genuine strength. Buy metal, hold it, cut it, deliver it fast, earn a spread. There are exactly two analytical complications, and both are addressed in the memo: (1) LIFO accounting, which caused a $258M earnings swing between 2024 and 2025 and made reported margin move opposite to economic margin; and (2) the company-owned life-insurance program, whose policy-loan interest has caused at least one major data vendor to overstate RS’s interest expense by roughly 3x ($165M reported versus $55.7M actual). (FACT)

Can it be undermined by foreign low-cost labor?

No — the opposite. The business is structurally protected by trade barriers. Service centers are inherently local: metal is heavy, freight-cost-sensitive, and customers need same- or next-day delivery of small orders. Offshoring the distribution function is not economically feasible. Foreign metal can and does compete, which is precisely why Section 232 tariffs have been such a powerful tailwind. (INTERPRETATION) The genuine exposure is that RS’s customers — fabricators and manufacturers — can be offshored, which would erode domestic demand over time.

Do brands matter?

Modestly, and in an unusual way. RS operates ~75 subsidiaries under their own long-established local brands (Phoenix Metals, Metals USA, EMJ, Yarde Metals, National Specialty Alloys, United Pipe & Steel). (FACT) These carry local reputational equity with fabricators, but the purchase decision is driven by availability, price, processing capability and delivery speed. Brand is a modest reinforcer, not a moat. (INTERPRETATION)

What is the nature of competition?

Availability first, service second, price third — and in a tight market like 2026, availability dominates almost entirely. This is why mill allocation preference is RS’s most valuable asset right now. COO Koch: “when it is really tight, we get what we got in the past and also… we do get preferential treatment.” (FACT) In loose markets the competitive axis shifts toward price, which compresses spreads industry-wide.

Customers’ switching costs?

Low in theory, moderate in practice. (INTERPRETATION) No contracts, no technical lock-in, no integration. Another service center can quote the same grade tomorrow. In practice, small fabricators placing frequent low-ticket orders value a single reliable source that always has material, will process it and will deliver next day — and RS notes customers have shifted to buying “more frequently and in smaller quantities,” which increases dependence on supplier reliability. This is convenience-based stickiness. It would not survive RS being persistently out of stock or materially more expensive.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet?

Yes — and one is large. The LIFO reserve was approximately $700M at June 30, 2026 (up from $548.6M at December 31, 2025). (FACT) Inventory is carried roughly $700M below current replacement cost. Pre-tax that is about $13.50 per share; after tax at 23.5%, roughly $10.40 per share of unrecognized economic value. The CFO characterizes it correctly as available to “help mitigate the impact of future metal price declines” — when prices fall, the reserve unwinds into LIFO income, cushioning reported earnings.

Two further items: (a) the ~75 subsidiary brands and 87 years of mill relationships carry no balance-sheet value but produce the allocation preference that drives share gains (INTERPRETATION); (b) real estate — $299.5M of land and $1,797.6M of gross buildings carried at depreciated historical cost across ~310 locations, much acquired decades ago. (FACT) Market value likely exceeds book, though this is not quantified in the filings.

Off-balance-sheet liabilities?

Minimal. Operating and finance leases are capitalized on balance sheet under current standards ($318.6M of total capital leases at year-end 2025). Pension liabilities are trivial at $24.9M. No securitizations, no VIEs, no material guarantees disclosed. (FACT) The COLI program involves policy loans whose interest ($109.3M in 2025) is disclosed and offset by investment income ($107.3M), netting to a $7.8M expense — economically immaterial but presentationally confusing.

How conservative is the accounting?

Conservative, and notably so. (INTERPRETATION, well supported.)

• LIFO inventory — the conservative choice, charging COGS at approximately current replacement cost. It depresses reported earnings in rising-price environments (as now) and builds an unrecognized reserve.
• No restatements, no NT (late-filing) notifications, and no adverse ICFR opinion identified across the five-year corpus of 10-Ks, 10-Qs and 8-Ks. (FACT)
• Impairments are taken promptly and are modest: $9.9M (2025), $11.7M (2024), $108.0M (2020, during the COVID collapse). (FACT)
• Goodwill of $2,169.9M on $10,373.3M of assets is meaningful but reflects decades of acquisitions; the 2024 deals added only $59.5M of goodwill on $373.0M of consideration.
• SBC of $55.6M (2025) is 0.4% of sales — low, and non-GAAP adjustments are minimal (GAAP EPS of $6.29 versus non-GAAP of $6.27 in Q2 2026 — the adjustment runs against management’s favor, which is unusual and a mark of good faith). (FACT)

How CapEx-hungry is the business?

Moderately, and currently elevated by choice rather than necessity. Capex has exceeded D&A for four consecutive years: $468.8M (2023), $430.6M (2024), $328.9M (2025) against D&A of $245.4M/$268.7M/$278.2M. FY2026 is guided to ~$300M with about half allocated to growth, implying maintenance capex of roughly $150M — about 1.1% of sales. (FACT for capex; ASSUMPTION for the maintenance split, based on management’s stated 50/50 allocation.) That is light by industrial standards. The real capital intensity in this business is working capital, not fixed assets: inventory of $2,187.8M and receivables of $1,571.1M dwarf annual capex.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy?

FCF is substantial but counter-cyclically patterned: $1,001.0M (2020), $562.8M (2021), $1,776.8M (2022), $1,202.5M (2023), $999.2M (2024), $502.5M (2025). (FACT) The crucial and underappreciated feature is that FCF is highest when the business is worst — in 2020 revenue collapsed 37% yet FCF hit $1,001M because inventory liquidation released $400.3M of working capital. Conversely 2021’s 60% revenue growth consumed $909.7M of working capital and cut FCF to $562.8M. This is a self-financing business in a downturn.

The philosophy, in management’s stated order: organic growth investment, acquisitions, then shareholder returns via a growing dividend and opportunistic buybacks. In practice, shareholder returns have dominated: 2024–25 combined saw $1,687.8M of buybacks plus $504.4M of dividends — $2,192.2M returned against $1,501.7M of FCF, with the gap funded by drawing cash from $1,080.2M to $216.6M and raising net debt from $62.0M to $1,204.3M. (FACT)

Significant acquisitions recently?

Four in 2024 — Cooksey Iron & Metal, American Alloy Steel, Mid-West Materials, and certain FerrouSouth assets of Ferragon — for $373.0M of net assets acquired, of which only $59.5M was goodwill (16% of consideration). Those businesses produced $389.2M of 2025 net sales, implying roughly 0.96x sales paid. (FACT) No acquisitions closed in Q2 2026. (FACT)

(INTERPRETATION: this is genuinely disciplined M&A. Buying tangible assets — $109.9M of inventory, $107.5M of PP&E, $44.9M of receivables — near fair value rather than paying strategic premiums is the opposite of the goodwill-inflating serial-acquirer pattern that destroys returns on capital. The 10-K’s statement that “we have not diversified outside our core business… since inception” is 87 years of strategic focus.)

Buying back shares?

Historically yes, aggressively and well — and right now, no. Since 2021: ~$3.4B repurchased at an average of roughly $234 per share, a 22% share-count reduction, against a current price of $407.83 — a realized gain of about 74%. Over 2014–2026, diluted shares fell from 78.6M to 51.4M, a 34% reduction. (FACT)

By year: $323.5M (2021), $630.3M (2022), $479.5M (2023), $1,093.7M (2024), $594.1M (2025), $234.2M (Q1 2026), $0 (Q2 2026).

(INTERPRETATION — and this is the single most important governance datapoint in the file: with $529M of remaining authorization, 0.9x net-debt/EBITDA, $980M of undrawn revolver, and a quarter of record results, management bought nothing. The CFO said only: “We remain opportunistic in our approach.” The most informed, best-incentivized and demonstrably most skilled buyer of RS stock declined to buy at these prices.)

Issuing large amounts of new shares to insiders?

No. SBC was $55.6M in 2025 (0.4% of sales, ~7.5% of net income) — modest by any standard and overwhelmed by buybacks. Equity awards outstanding capacity is 1,248,825 shares (~2.4% of shares outstanding). All employee equity awards carry a minimum one-year vesting period. (FACT)

Compensation policy of directors/management?

Above average, with two identifiable structural flaws. (FACT for the structure; INTERPRETATION for the assessment.)

Structure: Annual cash incentive = 80% Pretax Income Margin + 20% Tons Sold Growth measured against the MSCI industry benchmark (weight doubled from 10% in 2024). NEO target is 150% of base salary. Long-term equity = three-year ROA target with threshold 7% (25% vesting), target 10% (100%), maximum 13% (200%); 80% of CEO/CFO/COO target equity is performance-based.

2025 outcomes: Pretax Income Margin of 6.91% earned 96.4% of base salary; Tons Sold Growth of 7.27% above the MSCI benchmark earned the maximum 60% of base salary.

Governance hygiene — genuinely strong: No employment agreements. No change-in-control or golden-parachute agreements. No perquisites. No tax gross-ups. No hedging permitted. Clawback covering all incentive compensation. Independent non-executive Chair. CEO target annual cash incentive is set at approximately the 25th percentile of the peer group (other NEOs at top quartile). CEO holds 26.9x base salary in stock against a 5x requirement; COO and CFO at 8.2x against 4x.

The two flaws: (1) Pretax Income Margin — 80% of the annual bonus — is largely set by the metal price cycle, not by management. It fell to 6.91% in 2025 because ASP declined and LIFO swung $258M; it will rise sharply in 2026 because tariffs lifted prices. Management is substantially paid on the tape. (2) The long-term ROA metric has a low bar and is an accounting return. A 7% threshold against a twelve-year average ROA of ~9–10% (18.5% in 2022) makes threshold vesting close to automatic outside a severe downturn; even the 10% target was cleared in 2024 (11.57%) and 2025 (10.15%). There is no ROIC hurdle and no relative-TSR component anywhere in the plan.

The redeeming feature: the Tons Sold Growth versus MSCI metric is a true relative measure that cannot be earned by riding the cycle — and the Committee deliberately doubled its weight.

Motivations of management?

Well aligned by ownership and modest by pay. (INTERPRETATION) CEO Karla Lewis holds 26.9x base salary in stock — far beyond any requirement — and accepts a target cash bonus at the 25th percentile of peers. There are no employment contracts, so executives serve at will with no severance protection. The absence of golden parachutes, gross-ups and perquisites is genuinely unusual for a $21B-cap company.

The offsetting observation is behavioural: over the trailing two years, no officer or director has made a single open-market purchase, while insiders sold 97,720 shares for approximately $30.2M (all code-S; sellers include the CEO, COO, CFO and General Counsel). (FACT) Sales are small relative to holdings and partly routine, so this is not a distress signal — but it removes any “insiders are buying” support for the current price.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer?

No. Reliance, Inc. is a U.S. domestic C-corporation listed on the NYSE under “RS” since its September 16, 1994 IPO. Ordinary common stock, standard Form 1099 dividend reporting, no K-1, no ADR structure, no MLP complications. CUSIP 759509102, ISIN US7595091023. (FACT)

One practical caution: the ticker “RS” also trades on the TSX for Real Estate Split Corp., an unrelated Canadian issuer. News feeds and screeners routinely conflate the two. Separately, at least one major vendor classifies RS under “Mining — Miscellaneous,” which is a mapping error — it is a distributor, not a miner. (FACT)

Dividend policy?

66 consecutive years of dividends — never suspended, never cut, spanning every recession since 1960. (FACT) The most recently declared dividend is $1.25/share payable August 28, 2026. The quarterly rate rose 10.0% in 2024 and 9.1% in 2025. Dividend per share has grown from $1.40 (2014) to $4.85 (2025), a 3.5x increase. The payout ratio was 34.2% in 2025 and only 11.7% at the 2022 earnings peak — comfortably covered even at trough earnings. Current yield is approximately 1.2%. (FACT)

How profitable is the business?

Answered above under Business Quality. Summary: 2025 gross margin 28.7% (LIFO) / 29.5% (FIFO); operating margin 7.0%; net margin 5.2%; ROIC 8.8%; ROE 10.1%. Q2 2026: gross margin 28.1% (LIFO) / 30.5% (FIFO); pretax margin 9.3%. Through-cycle ROIC of roughly 12–13%. (FACT)

Is net income diverging from cash from operations?

Yes, and the divergence is directionally informative rather than alarming. CFO/NI ratios: 3.18x (2020), 0.57x (2021), 1.15x (2022), 1.25x (2023), 1.63x (2024), 1.12x (2025). (FACT)

The pattern is mechanical and benign: cash conversion exceeds net income when the business shrinks (inventory liquidates) and falls below it when the business grows rapidly (inventory builds). The 0.57x in 2021 reflected a $909.7M working-capital build during 60% revenue growth, not an accrual-quality problem.

The current period bears watching. H1 2026 operating cash flow was only $313.6M against H1 pretax income of $779.3M — a ratio well below 1.0x, reflecting the working-capital absorption of simultaneous volume and price growth. (FACT) This is an expected consequence of the growth pattern, not an earnings-quality red flag, but it means reported EPS growth is materially outrunning cash generation right now. FCF of $156.0M in H1 2026 against a ~$20.9B market cap annualizes to well under a 2% yield.

Risks & Downside

What factors would cause the stock to decline?

In descending order of expected impact: (INTERPRETATION, grounded in the memo’s risk matrix.)

1. A reduction or partial rollback of Section 232 tariffs. Would compress domestic metal prices, average selling price and realized spread simultaneously. The dominant variable, and it is a policy question rather than a business question.
2. Multiple compression alone. At the 99.86th percentile of ten-year history on P/E, P/B and P/S, a reversion toward even the upper end of the historical range (2.1–2.6x book versus 2.98x today) is a material decline with no change in earnings whatsoever.
3. A metal price reversal independent of policy. 2024–25 demonstrated the sensitivity: a 2.6% ASP decline drove a 10.2% EPS decline.
4. Border wall roll-off. ~9–10% of EPS ends mid-2027 unless Phase 2 is exercised, which management says is “not guaranteed.”
5. A cyclical downturn in non-residential construction or general manufacturing (each ~one-third of sales).
6. Continued absence of buyback support through further quarters, which would confirm management’s valuation reticence.

Risk of a catastrophic loss?

Low. (INTERPRETATION, well supported.) No customer concentration (125,000+ customers). No single-product or single-technology dependency. No meaningful litigation, environmental or regulatory-sanction exposure disclosed. Net debt/EBITDA of 0.9x with $980M of undrawn revolver capacity and $235.4M of cash. Pension liability of $24.9M is trivial. No going-concern language, no covenant stress, no adverse ICFR opinion in the five-year corpus. Critically, the business generates cash in a downturn — 2020 produced $1,173M of operating cash flow on revenue that fell 37%.

Chance of a total loss?

Negligible over any reasonable horizon. RS has been profitable in every year since its 1994 IPO — through the dot-com bust, the 2008–09 financial crisis, the 2015–16 industrial recession, and the 2020 pandemic — and has paid dividends for 66 consecutive years. It holds hard assets: $2,633.3M of net PP&E, $2,187.8M of inventory (carried ~$700M below replacement cost) and $1,571.1M of receivables against $1,670.0M of total debt. (FACT) A total loss would require a simultaneous collapse of North American metals demand and an inability to liquidate inventory — an implausible combination for a business whose working capital self-liquidates in downturns.

The realistic downside is not loss of capital but loss of time: a multi-year period of flat or negative returns as policy-inflated earnings normalize and a record multiple compresses toward its historical range. (INTERPRETATION)

Recent News & Events

Has the business environment changed recently?

Yes, materially, and in two directions. (FACT)

Favorably: Section 232 steel and aluminum tariffs were raised to 50% in June 2025 and expanded to derivative articles, following the March 2025 increase in aluminum tariffs from 10% to 25% and the elimination of tariff-rate quotas and product exclusions. This has lifted domestic pricing sharply (Q2 2026 ASP +14.5% YoY; aluminum “nearly doubled” from pre-tariff levels), extended mill lead times, and — because RS enjoys allocation preference — accelerated market-share capture. Demand has broadened: data centers and energy infrastructure are driving non-residential construction, semiconductors have gone from weak to “meaningfully improved,” and aerospace backlogs are growing.

Unfavorably: the February 20, 2026 Supreme Court ruling that the president lacks authority to impose tariffs under IEEPA demonstrated that tariff authority is judicially contestable — though Section 232 was explicitly unaffected. Input-cost inflation from the tariffs has tripled quarterly LIFO expense versus guidance and forced management to double its FY2026 LIFO estimate from $150M to $300M. Freight and fuel inflation attributable to the U.S.–Iran conflict was named by the CFO as an SG&A driver and a forward risk to the Q3 outlook.

Significant acquisitions?

None in Q2 2026. The last program was the four 2024 deals ($373.0M, ~0.96x sales, $59.5M goodwill). (FACT)

Change in accounting policies?

None. LIFO inventory accounting is unchanged and long-standing. There have been no restatements, no NT filings and no auditor changes in the five-year corpus. The only presentational change of note is segment-reporting disclosure enhancement, which added CODM detail without altering the single-segment structure. (FACT)

Recent changes — new markets, facilities, management?

• New market/contract: the DHS border wall contract, awarded early 2026, first shipments April 2026 — Phase 1 ~$1.4B through mid-2027, Phase 2 $800–900M at customer option. This meaningfully deepened RS’s non-residential construction and carbon tubing position (tubing rose from 9% to 12% of mix quarter-over-quarter). (FACT)
• Facilities: ~310 locations, modestly consolidated from 315 in 2021 as the network is optimized. FY2026 capex of ~$300M with about half for growth — processing capability, footprint expansion, and volume growth in attractive markets. (FACT)
• Management: stable. Karla R. Lewis (President and CEO), Stephen Koch (EVP and COO), Arthur Ajemyan (SVP and CFO), William A. Smith II (SVP and General Counsel). No executive departures identified in the 8-K corpus. Board refreshed via standard annual director equity grants (469 shares each on May 20, 2026). (FACT)
• Identity: renamed from Reliance Steel & Aluminum Co. to Reliance, Inc. on February 15, 2024 to signal a “diversified metal solutions provider” positioning. (FACT — but the business mix is unchanged: carbon steel has been exactly 53% of sales for three consecutive years. Investors should not read a business-model change into a name change.) (INTERPRETATION)

APPENDIX B — Source Appendix

Reliance, Inc. (NYSE: RS) — July 25, 2026

All sources accessed July 25, 2026 unless otherwise noted. Primary sources (SEC filings, company releases) take precedence over aggregated third-party data throughout the memo; where the two disagreed, the filing was used and the discrepancy is noted below.

1. Primary — SEC Filings (Reliance, Inc., CIK 0000861884)

The trailing 60-month SEC corpus was enumerated and reviewed in full: 5 × 10-K, 15 × 10-Q, 36 × 8-K, 5 × DEF 14A, 5 × DEFA14A, 10 × 11-K, 5 × SD, 6 × Form 3, 164 × Form 4. Form 144 filings (34) were excluded as noise per the filing-corpus standard.

All filings are publicly available via SEC EDGAR under CIK 0000861884.

2. Primary — Company Releases and Management Commentary

3. Quantitative Data Sources (third-party; reconciled to filings)

4. Industry and Regulatory Sources

5. Secondary / Trade Press (triage only; all material claims verified against primary sources)

Excluded as ticker collisions: Real Estate Split Corp. (TSX: RS and RS.PR.A) distribution announcements dated 2026-05-22 and 2026-06-23 appeared in the news feed and are a different issuer. Excluded.

All sources listed above are public: SEC filings (EDGAR, CIK 0000861884), company press releases and earnings-call transcripts, published industry shipment data, and public market-data services. No non-public information was used in preparing this article.