Frequency Electronics: The Backlog, Not the Beat, Is the Story

Frequency Electronics reported a record quarter on Sept. 10, and the headline numbers are easy to be impressed by: revenue of $23.5 million, up 70% from a year earlier, and GAAP earnings of $0.41 a share against roughly $0.20 the street had penciled in. Big beats usually get the headline. They are not the point here.

The point is the piece of paper that made the beat possible — a funded backlog of $129 million, up 82% year over year — and what that backlog does to the company's financial story. For years Frequency Electronics was the kind of small-cap you avoided for structural reasons: lumpy revenue from a handful of government contracts, thin margins, and a balance sheet so light it could barely fund its own growth. Not long ago its cash had drained to about $1.6 million. It was a company priced like a hope.

Why the backlog outranks the beat

The demand story is a technology cycle with obvious timing: the proliferated satellite constellations now being launched in the thousands each need high-precision clocks and timing systems, and Frequency Electronics is a long-standing leader in that niche. The economics come from follow-on production. Once engineering is paid for, each additional satellite order rides on sunk development cost — which is why management says the follow-on work carries above-average margins, and why the numbers actually show it.

Look at the operating trajectory rather than the beat itself. In fiscal 2026 the company did $63.2 million of revenue and still lost about $3.0 million at the operating level — that was the old story, a business that couldn't quite cover its costs. This quarter it produced $5.2 million of operating income with gross margin near 46%, and backed that with $3.0 million of cash generated from operations against a $1.2 million use of cash a year earlier. The switch from cash-burn to cash-generative is the proof point, and it is early: on a trailing twelve-month basis free cash flow is still slightly negative, an honest signal that the conversion has not fully arrived yet.

There is a second, quieter change that matters just as much. Frequency raised roughly $73 million from a secondary offering, lifting cash on hand from $1.6 million to about $61 million while remaining debt-free. The old bear case leaned on the balance sheet — a thin, fragile capital structure that could not finance the ramp it was promising. That objection is gone. A company that could not afford to scale now has the cash to do it.

The complicated part: the price and the proof

That is the case, and it is a real one. The honest tension is that the market has not been asleep. The stock is up more than 120% over the past year to about $62, off an $80.60 high, and it trades near 29 times forward earnings and roughly 11 times sales. This is not a beaten-down, expectations-reset entry hiding in the corner. The rerating is partly priced, and the edge now depends on whether the backlog converts to cash at something like the margins the last quarter showed.

So the bear case deserves a straight answer. Revenue still leans on a small number of large contracts, a few primes can move the year, and the niche faces larger competitors with deeper pockets. The single most important thing to watch is the conversion itself. If the funded backlog keeps growing and quarterly operating cash flow holds positive through the ramp, the high multiple slowly earns itself. If instead the backlog stalls — or production mix shifts enough to crack that 22% operating margin — then a stock priced for a finished turnaround has room to fall hard, because nothing in the cash path yet justifies the valuation on its own.

The judgment

What makes Frequency Electronics worth attention is not the EPS beat. It is that the structural reasons to avoid the company have been dismantled at the same time the operating numbers inflected — a record funded backlog, a cash-rich debt-free balance sheet, and evidence the cash conversion is beginning. The market is increasingly pricing the new story, so the asymmetry is thinner than it was a year ago; the reward now comes from the backlog actually turning into free cash flow, and the tripwire is if that stops before the multiple is earned. I can be wrong again, but the setup is now a funded growth story whose proof point is observable quarter by quarter: watch the funded backlog and the sign of operating cash flow.