Coffee Holding (JVA): The Margin Spike Is Timing, Not a New Earnings Base

Coffee Holding reported a fiscal third quarter that looks like a clean turnaround: $1.99 million of net income, or $0.35 a share, swinging from a $1.2 million loss a year earlier, and the shares jumped about 11% on the news. Wall Street's reflex is to read that as proof the little Staten Island coffee roaster has turned a corner. The numbers underneath tell a thinner story — revenue actually shrank, and most of the profit came from sources that will not repeat at anything close to this size.

The headline profit arrived despite a shrinking top line. Net sales fell to $21.69 million, down 9.3% from a year earlier. The swing came entirely from gross margin, which exploded to 25.0% of sales from just 9.4%. A roaster's gross margin is the spread between what it pays for green beans and what it charges customers, and that spread widened for three distinct reasons, each with a different shelf life.

The first is inventory timing. Green coffee prices collapsed starting in late January after the world swung from physical tightness to a forecast record 2026/27 crop. Coffee Holding bought coffee at those lower prices during the first half of this year, then sold it through in the third quarter while its customers were still being repriced — a one-time windfall you sell down, not a margin level you maintain.

The second is the government refunding tariffs the company paid in 2025. Coffee Holding says these tariff refunds added roughly $0.06 a share to the quarter, and management expects more over the next two quarters — after which, by definition, the pool is exhausted. A refund is cash coming back for money already spent, not new earning power.

The third is trading. The quarter included a net gain from trading coffee futures and options, against a net loss in the year-ago period. But the company's own stated policy in the prior quarter was to scale back hedging and short-term trading to a small capacity. Management is deliberately shrinking the exact activity that helped the profit.

That is the whole tension in the stock, which looked cheap even after the pop: roughly $20 million in market value against $3.9 million of net income in the first nine months of fiscal 2026, up from just $592,000 a year earlier. Nine-month revenue, by contrast, reflects a business being repriced downward as coffee falls. A wholesale roaster's top line is a pass-through of the commodity — when beans get cheaper, so does the revenue, because the savings go to customers. That is why the sales decline and the margin jump are two sides of the same coin, and why the $0.35 quarter is best read as a snapshot of a volatile quarter rather than the start of a new earnings run-rate.

None of this makes the quarter bad. Management's decision to absorb tariffs in 2025 kept its roasted-customer base intact, and cheaper beans plus tariff refunds genuinely fattened the period. Real cash came in. The problem is the market paying up for it.

Management says it expects to grow revenue, gross margins, and profitability from the end of this calendar year into 2027. That is the test, and it is a falsifiable one. For the margin to be real, it has to survive what built it: the cheap-inventory windfall works through, the tariff refunds stop after the next two quarters, and the trading gains fade by design. What should remain — and what would justify today's move — is a margin that holds with revenue stabilizing rather than shrinking further.

Right now the evidence supports neither celebration nor alarm, and it definitely does not support treating a one-off margin spike as a new base. The honest call is to wait: the business is genuinely improving from a weak point, but a low multiple on peak, refund-and-timing-heavy earnings is a classic "cheap for a reason" setup, not a bargain. The refunds end in two quarters. Whether the 25% margin was real will be answerable then.