Graham's Record Backlog Meets a 40% Drawdown

A strong quarter, a new power-density narrative, and a market that is no longer paying for the backlog alone.

GHM · Earnings Call · 2026-11-05

A good quarter into a bad tape

Graham's fiscal Q2 2027 print was, on its face, the kind of quarter a small-cap industrial dreams about. “Second quarter revenue was $69.8 million, up 6% from the prior year. Adjusted EBITDA increased 64% to $10.3 million” — Matt Malone, President and CEO · 2026-11-05, gross margin expanded 650 basis points to 28.2%, and backlog hit a record $561 million, the seventh consecutive record. Management reaffirmed FY27 revenue of $285-$295 million and adjusted EBITDA of $35-$40 million. Yet the stock is down roughly 40% over the last 90 trading days and 41.8% from its June 30 peak of $123.79. The report did not break the de-rating—it simply failed to reverse it. That gap is the story. The operating improvement is real, but the market is no longer paying for a backlog headline. It is asking what the backlog converts into, how much of the margin is structural, and whether Graham is still a Navy supplier with a cycle or a broader rotating machines platform. Even after the drawdown, the valuation is not cheap: price to operating income is still about 102x, and price to revenue is 5.6x. This is a stock that had been priced for a multi-year transformation and is now demanding proof.

Power density becomes the lead narrative

The freshest company-specific language in the quarter is not submarines. It is power dense machinery. Graham's keyword set puts rotating machines first and power density third, and the Q&A leans into the theme. Matt Malone described Barber-Nichols as a maker of some of the most power-dense rotating machines on the market: “Barber-Nichols does design some of the most, and produce some of the most power-dense rotating machines on the market.” — Matt Malone, President and CEO · 2026-11-05 He then walked through microreactors that must fit inside shipping containers, radar platforms moving from directional to 360-degree coverage, and space applications that include cooling launch vehicles, satellites, and even astronaut backpacks. The point is that Graham's engineering is being pulled into applications where performance per unit of volume matters more than legacy end-market labels. This is not wholly new—prior calls have discussed radar and directed-energy cooling, and the June Q&A already used the phrase “think more power dense solutions.” — Robert Brooks, Analyst · 2026-06-08 But it is being elevated from supporting color to the central equity story. That matters because it broadens the addressable market beyond the Defense backlog that already dominates the numbers. Defense is 66% of Q2 revenue and about 83% of backlog; the company needs adjacencies to justify a growth multiple. The new language is management's attempt to show that the adjacency is technical, not just acquisitive.

The margin beat has an asterisk

The 650-basis-point gross-margin expansion is the most eye-catching number in the release, and also the one that requires the most care. Chris Thome was explicit that the quarter benefited from comparison dynamics:

Margins were a little bit higher than our normal run rate for the quarter, and that really was a function of mix as well as higher volumes. But primarily, the impact on the margin was lower material receipts.
Chris Thome, Chief Financial Officer · 2026-11-05

That is the key caveat. In the prior-year quarter, Graham had about $10 million of unusually high material receipts, which carry lower margins. This year, lower material receipts flattered the comparison. Management raised full-year gross-margin guidance by only about 50 basis points, even after a 650-basis-point quarterly print. The market can do that math: a chunk of the beat is timing and mix, not a new run-rate. To be fair, the improvement is not only accounting. Malone cited better execution, pricing discipline on new work, and the unwinding of lower-margin jobs out of backlog. Those are real operational levers. But the burden of proof is now on the next few quarters to show that 28% gross margin is a floor rather than a peak. The Q3 guide implicitly acknowledges this—fiscal Q3 is seasonally the weakest because of holidays and direct-labor dependence.

Backlog is visibility, not revenue

Record backlog is a genuine asset. It is also a slow asset. Management expects only 30%-35% of quarter-end backlog to convert within 12 months: “We currently expect approximately 30%-35% of quarter-end backlog to convert to revenue within the next 12 months” — Chris Thome, Chief Financial Officer · 2026-11-05, another 20%-25% within one to two years, and the remainder beyond two years. That is the right structure for multi-decade shipbuilding programs, but it explains why the stock does not trade on backlog alone. The company reaffirmed, rather than raised, FY27 revenue guidance, and the midpoint implies about 18% growth—good, but not the kind of upside that snaps a 40% drawdown. The order mix is also lumpy. Q2 book-to-bill was 1.1x, and first-half book-to-bill was 1.2x, but prior calls have repeatedly warned that quarterly orders can swing widely. In June, Thome said “our business as a whole is susceptible to volatility and lumpiness with regards to orders.” — Christopher Thome, Chief Financial Officer · 2026-06-08 That remains true. The current quarter's Aftermarket orders were up 91% to $18.3 million, which is a useful counterweight, but aftermarket sales were still down 3% year over year at $8.6 million. Orders lead sales; the conversion is not instant. The weaker spot is energy and process, where revenue fell 17% as large capital projects remained soft. That is not a new problem—it has been the drag for several quarters—but it means the consolidated growth rate is doing more work than the segment detail suggests. Add in the addition of FlackTek, which contributed $9.5 million of Q2 revenue, and the 6% consolidated growth is clearly acquisition-inclusive. The legacy business is growing in defense and space but shrinking in energy; the mix is improving, but it is not a clean organic acceleration.

The balance sheet is the quiet bull case

What the bear case cannot easily dismiss is the cash and the balance sheet. Graham generated $34.1 million of operating cash flow in the first half, ended the quarter with $71 million of cash, no debt, and $75 million available on its revolver. That gives management room to keep investing in capacity—the new Arvada facility, the cryogenic test facility in Florida, automated welding—without financial strain. It also supports the M&A pipeline that Malone says remains full. The risk is that the market is no longer rewarding capacity investment at any price. The stock's drawdown says investors want to see that the new capacity converts into revenue and margin, not just into a larger asset base. Graham's seventh consecutive record backlog is a strong leading indicator. But leading indicators are only valuable if the lag is tolerable. At 102x operating income, the lag is the entire thesis.

What to watch next

The next quarter needs to answer three questions. First, does gross margin normalize toward the mid-20s, or does the Q2 level hold? Second, do power-dense orders in radar, microreactors, and space convert from development to production at the pace management implies? Third, does the stock stabilize after a 40% drawdown, or is the market telling us that the June peak was the top of a cycle in defense/space small caps? Graham is not a broken story. It is a company with a record backlog, a net-cash balance sheet, and a genuinely differentiated technical position in power-dense rotating equipment. But the report also shows why the stock has struggled: the headline growth is acquisition-assisted, the margin beat has a timing asterisk, and the backlog converts slowly. The next leg depends on proving that the new narrative is operational, not just rhetorical.