Greenbrier heads into its fiscal fourth-quarter report with a split personality. On one side sits a railcar builder that has squeezed better margins out of trough volumes and quietly built a much larger leasing business. On the other sits a backlog that has fallen every quarter to its lowest level in more than a decade, and a full-year earnings forecast that management has cut three times. The October 26 release, due after the close, will show which of those stories carries more weight as the company closes out fiscal 2026.
Wall Street expects earnings of $0.89 per share, down about 29% from the $1.26 in core earnings posted a year ago. That is a clear contraction, though not a surprise given how far deliveries have fallen. The more useful comparison is with management's own math. Greenbrier narrowed its fiscal 2026 outlook to $3.00 to $3.15 per share, down from $3.00 to $3.50, after starting the year at $3.75 to $4.75. With roughly $2.21 already booked through the first three quarters, the guidance implies fourth-quarter earnings of about $0.79 to $0.94. The consensus sits in the upper half of that window, so analysts are not asking for heroics, but they are assuming management lands closer to the top of its range than the bottom. Revenue guidance of $2.40 billion to $2.50 billion was held steady, which matters because further delivery slippage into fiscal 2027 is the most obvious way the quarter could come up short.
Margins are where the prior narrative will be validated or challenged. Aggregate gross margin bottomed at 11.8% in the second quarter and rebounded to 14.1% in the third, which management credited to insourcing in Mexico, labor efficiency and cost discipline. Executives framed these improvements as structural, describing a pattern of higher lows. A fourth-quarter margin that holds near or above the third-quarter level on still-thin volume would support that claim. A slide back toward 12% would suggest the recovery owed more to timing and mix than to lasting change. Gains on asset sales, which flattered the first half, have largely faded, so the quality of operating earnings deserves extra scrutiny.
Leasing is the bright spot that needs to keep shining. The owned fleet grew from about 17,000 cars to 20,600, utilization reached 99%, renewals carried double-digit rate increases, and the roughly 4,400-car secondary-market purchase supports a goal of doubling recurring revenue by 2028. Evidence that those acquired cars are producing revenue, and that lease originations remain a large share of orders, would show the business is becoming less dependent on the new-build cycle. Investors should also look for any update on two newer overhangs: a possible Section 232 tariff on Mexico-built tank cars, with unclear retroactivity, and a customs ruling on couplers now under appeal. Management played down both, citing tariff pass-through clauses, but any quantified exposure would change the conversation.
The market has not given Greenbrier the benefit of the doubt. The stock has fallen about 18% since the last report while the S&P 500 gained roughly 2%, and at $39.11 it sits just above its post-earnings low of $37.61 and well beneath its 200-day moving average near $48.69. The entire prior-quarter trading range sat above today's price. Sentiment leans firmly bearish. That pessimism lowers the bar somewhat, though it also signals deep skepticism about the industry rebound management has promised.
That rebound is the crux. Management expects North American deliveries to recover from roughly 24,000 to 25,000 in calendar 2026, the weakest since 2010, to more than 34,000 in 2027, helped by tighter trucking capacity, slower rail velocity and emerging demand for heavy-duty cars tied to data-center construction. As the fiscal year closes, the report is a natural moment to show whether that view is turning into orders. If the backlog stabilizes or grows and margins hold, the higher-lows thesis survives. Another backlog decline alongside soft commentary on fiscal 2027 would suggest the trough is lasting longer than management has acknowledged.