MYR Group: The 43% Crash Is a Rates Story — the 13% Margin Is the Real Question

MYR Group's stock more than tripled off its 52-week low near $172, peaked just above $500, and has since fallen roughly 43% to about $284 — all while the underlying company keeps posting the best numbers in its history. In the second quarter it delivered record revenue of $1.08 billion, record net income of $49.9 million, and $3.17 of earnings per share against roughly $2.62 expected, a business that had essentially no net debt afterward. What more do investors want?

The answer matters, because the pullback in this specialty electrical contractor — the company that builds the transmission lines, substations, and data-center wiring the electrification boom depends on — is being read two different ways. One camp sees a gift: a growth story on sale. The other sees the air coming out of a parabolic trade. The truth is closer to the second, but only half of it. The crash is mostly a rates story and a sector-wide repricing, not a broken company. The reason not to buy it yet is narrower and more specific — and it comes down to a single margin number that I'll argue hasn't proven itself.

First, the crash itself. MYR fell alongside the whole infrastructure contractor complex as Treasury yields spiked — the 30-year hit a 19-year high around 5.33% in early September — and as the rate scare hit the highest-multiple infrastructure names hardest. You can see the same de-rating everywhere: EMCOR, Primoris, Comfort Systems, MasTec, and Quanta have all rolled over together. This is the sector tide repricing an AI-and-grid trade that had gotten ahead of itself, not a MYR-specific fundamental collapse. The stock is now below both its 50-day and 200-day moving averages with an RSI near 31, technically in a downtrend and flirting with oversold — the classic shape of a name that ran too far, fast.

That framing matters because the fundamentals underneath are not just intact; they're improving. Backlog, the closest thing a contractor has to a forward-order book, hit a record $3.16 billion at the end of June, with backlog up 20% year over year. Revenue grew about 20% in the second quarter, and the balance sheet is close to cash-neutral after years of carrying debt. The demand that drives the business — electrification, grid modernization, data-center buildout — is real and structural. Electricity demand is surging, driven by data centers and electrification, and the grid work to connect them is measured in trillions.

So on the surface, this is the contrarian's dream: a high-growth, industry-leading contractor whose stock got cut in half while its order book grew. Which is why the decision to hold off buying deserves an honest test rather than a forced call.

Here is the crux — and for my money the reason the market is only half-wrong. The earnings that justify the stock price have been juiced by a margin spike that sits well above MYR's own historical range. Gross margin went from 8.6% in 2024 to 11.6% in 2025 and 13.2% in the second quarter of 2026. Segment operating margins roughly doubled in a year, with transmission and distribution climbing from 3.7% to 7.9% and commercial and industrial from 3.2% to 5.9%. Management itself attributes the jump not to new structural pricing power but to "better-than-anticipated productivity" and "favorable job closeouts" on fixed-price work — precisely the kind of items that get booked once and don't compound.

Contractors are lumpy by nature, and MYR's own history shows how violent the reversion can be. Gross margin was just 4.9% in one 2024 quarter, dragged down by problem fixed-price projects; changes in contract estimates lopped 4.4 percentage points off full-year 2024 margins. The current 13% gross margin, against an operating margin of about 5%, is a thin, mean-reverting edge running at the top of its range on closeout luck.

That single question — is ~13% gross margin the new structural normal from scale and mix, or a favorable-closeout artifact that fades toward the low double digits? — decides whether $284 is cheap or expensive. Strip the closeout lift out and the earnings power that a ~27-times-trailing-earnings stock is paying for is really worth a good deal less. On the current trajectory the stock is arguably near 20 times forward earnings — reasonable, not dirt cheap — but only if that profitability holds. If it reverts toward MYR's historic norm, the recent crash looks less like a bargain and more like the market pricing an undefeated business that had simply borrowed forward earnings.

This is why "worth watching but not buying yet" is the disciplined posture rather than a hedge. My usual contrarian gates — fundamental quality, valuation versus growth, and competitive durability — all hover at the edge of passing, but the margin variable is unresolved. There are also real demand-side hazards to respect: grid interconnection timelines stretching two to five years, more than 70 data-center projects rejected or restricted by local governments in the first four months of 2026, and roughly half of planned data centers facing delays. Those are headwinds to MYR's commercial-and-industrial segment, not existential threats — but they're a reminder that the demand tide can be lumpy.

So the setup resolves on evidence, and the reader can hold me to it. The buy argument becomes clean when the price stops making lower lows and reclaims the 200-day zone near $324 at the same time margins hold near current levels with management actually guiding to them in numbers. The contrarian case dies if gross margin rolls back toward the single digits — proof the spike was closeouts, not operating leverage — or if backlog growth stalls, which would be demand cracking, not just valuation.

I'm not buying the dip yet, and I don't think investors need to chase the crash. The market is arguably half-right: it's right to have taken the multiple out of a stock that tripled in a year, but it would be wrong to keep selling a record-backlog business into a rates-driven sector rout. The margin is the discriminator between a falling knife and a gift, and it hasn't spoken. Wait for it to do so.