Few investors have Coffee HoldingJVA-- on their screens. At a market value just above $20 million, the Staten Island wholesale roaster is the kind of name retail scanners skip. But its fiscal third quarter, reported this morning, deserves a look. Sales fell 9%, yet the business swung from a $1.2 million loss a year ago to $1.99 million of net income — $0.35 a share. The stock jumped nearly 15%.
That combination — falling revenue and a record profit — is the whole story, and it turns on how a coffee roaster actually makes money. This is a pass-through business. Green coffee beans are by far the largest cost, and coffee bean prices swung wildly through 2025 and into 2026: arabica hit a record near $4.40 a pound in 2025, then crashed starting in late January 2026 as a record Brazilian harvest reached the market, trading around $2.90 a pound by mid-year. A roaster passes those swings along to customers. When the bean price falls, its revenue falls too — which is why sales dropped to $21.7 million from $23.9 million.
But revenue is not profit. What matters is what the company paid for the beans sitting in its warehouse versus what it sold them for. Coffee Holding bought its inventory during the first half of fiscal 2026, when prices had already collapsed, then sold it through a quarter of stable-to-firm shelf prices. The result was a gross margin of 25.0% of sales, versus 9.4% a year earlier — the arithmetic behind the swing to the black. Gross profit more than doubled to $5.4 million.
Quick Backtesting Tool
Now the part that should keep a skeptical investor calm. That margin is flattered by things that will not repeat at the same size. Tariff refunds added roughly $0.06 a share in the quarter, and there was more of the same promised over the next two quarters. The company also booked a net gain on coffee hedging and trading activity, versus a loss a year ago. And the biggest line — the cheap inventory itself — is a one-time harvest. Once those low-cost beans are sold through and repriced at current market levels, gross margin tends to drift back toward the low-teens history this company has posted most years.
That is why the easy math is a trap. If you lazily annualize $0.35 a quarter, you get a stock that looks almost absurdly cheap against its ~$22 million market value. The honest question is how much of the $0.35 is a windfall and how much is the new, durable floor. Management itself frames the quarter as a volatility play, calling it one of the most volatile periods in coffee market history — precisely the sort of conditions that produce a one-off inventory windfall rather than a permanent margin reset.
The forward case, to be fair, is not nothing. The CEO argues the company kept 100% of its roasted customer base by absorbing tariffs in 2025 rather than passing them along, and that it has since won significant new business with major national brands that have shown no sign of cutting prices. If that is true, the lower pass-through revenue has a floor, and incremental volume drops straight to a healthier gross profit. Management expects profit growth to continue through calendar 2026 and into 2027.
The whole thesis reduces to one testable number. Watch the gross margin once the inventory bargain, the tariff refunds, and the hedge gains roll off. If it holds meaningfully above the historical mid-teens, Coffee Holding has genuinely re-rated from a marginal, chronically break-even roaster into something with real earning power. If it snaps back toward the low teens, this quarter was a well-timed commodity trade, not a new business model — and the stock's jump would look like a gift to sellers, not a signal. Revenue holding steady is the second tell. Expectations here are close to zero, so a resettled market that surprises to the upside has room to run; a margin that normalizes with volume still shrinking is the condition that breaks the re-rate.
The quarter proves the model can print money when the commodity finally moves its way. The live question — windfall or floor — gets answered in the next two to three quarters, as the cheap beans run out and the refunds are booked. Sit on your hands until then, and let the margin tell you which version of this company you actually own.
