RAIL — FreightCar America, Inc. · 2026-09-07 · Verdict: WATCH · Conviction 2

Price $7.10 (screen row close 2026-09-06; no live quotes available) · Mkt cap $233M (32,775,760 shares) · EV ~$286M (fully diluted for the 2023 Warrant, debt at face) · EV/FY25 EBIT 8.4x, EV/FY26E adj. EBITDA 6.5-7.9x · FCF yield 13.5% on FY2025, ~6% annualizing H1 2026 · Net debt ~$42M · ADV $1.3M Sources read: 10-K 2026-03-09 (Items 1 and 7; Item 1A was not in the bundle), 10-Q 2026-08-03, DEF 14A 2026-03-19, 8-Ks of 2026-05-04, 06-16, 07-02, 08-03 and 08-05, Form 4s (12 in trailing 12 months). No transcript or prepared remarks were filed as an exhibit, so section 5 rests on written guidance versus results.

1. What the business actually does

FreightCar America builds freight railcars, rebuilds and converts idled cars back into service, and sells parts through an Aftermarket segment (10-K 2025, Item 1). All production is at one AAR-certified plant in Castaños, Mexico, where 1,916 of its 1,986 employees work. Customers are financial institutions (78% of 2025 sales) and shippers (16%), and the top five were 75% of 2025 revenue, up from 48% in 2024. FY2025 revenue was $501.0M, gross profit $73.2M (14.6%), operating income $33.9M (Item 7).

2. Why it is mispriced, and the limit of the edge

The clean distortion is mechanical. On June 26, 2026 PIMCO's vehicle OC III LFE II LP net-exercised its 2020, 2021 and 2022 warrants and the company issued 13,619,377 shares, taking shares outstanding from 19,062,155 to 32,775,760 (8-K 2026-07-02; DEF 14A footnote 2). Screens read that as a 71.4% issuance, and universe_under2b.csv flags exactly that. It was not dilution: these were effectively penny warrants already inside the reported count, since weighted basic shares were 31,793,746 in Q2 2025, before any exercise (8-K 2026-08-03). Economics unchanged, optics changed. The exercise also flipped equity from negative $107.4M to positive $36.2M and cut the warrant liability from $168.5M to $14.0M, ending years of GAAP dominated by mark-to-market noise (a $32.2M loss in 2025 alone, Item 7).

The dirty distortion is that the stock is 52% off its high because the operating year fell apart. That is a miss, not a misunderstanding, and it caps this at WATCH.

3. Unit economics and what broke

Q2 2026 revenue was $113.1M on 927 deliveries versus $118.6M on 939, but gross margin fell from 15.0% to 5.5% and Manufacturing from 13.4% to 3.2% (10-Q). Adjusted EBITDA was $1.2M against $9.3M, and only $2.2M of the gross profit gap is one-time realignment cost. Management cites mix plus a ramp that "began later than planned due to customer delivery timing, reducing fixed-cost absorption," pushing deliveries into early 2027 (8-K 2026-08-03, CEO).

Backlog is the bull's exhibit: 3,972 units worth $344M at June 30 against 1,926 units worth $137M at December 31, on roughly 45% of industry new-railcar orders, with value per unit up from $71k to $87k as mix shifted from rebuilds to new cars. The filings never disclose that backlog's margin, and winning 45% of orders in a soft market at a 3.2% manufacturing margin is at least as consistent with buying share as winning it.

Aftermarket is the piece worth owning and is not yet compounding profit. Revenue went $18.2M (2024) to $27.1M (2025) to $20.2M in H1 2026 (+44.5%), yet segment operating income was $7.2M in both 2024 and 2025 and $4.49M in H1 2026 against $4.10M (Item 7; 8-K 2026-08-03 segment tables). Gross margin fell from 47% to 32% as the Carly Railcar Components distribution business was added. Growth so far is bought, not earned.

4. Balance sheet and capital allocation

Cash $63.0M against debt carried at $100.7M, roughly $105M face on the $115M Blue Torch Term Loan due December 2028 at 9.7%, plus a $35M ABL with $24.5M available (10-Q). Net debt of $42M is 1.0x guided EBITDA, which is fine; the coupon costs about $12.5M a year against $40M of EBITDA, which is not. An excess cash flow sweep applies, and $7.3M of the $7.8M H1 financing outflow was term loan repayment: free cash flow belongs to the lender, not to buybacks. There is no asset floor either, since the plant is leased and $52.1M of the $276.1M asset base is a deferred tax asset.

OC III LFE affiliates held about 49.2% of outstanding stock at December 31, 2025 (DEF 14A related-party note); one warrant for 1,636,313 shares remains. The Castaños plant is leased from director Jesús Gil and his brothers, who also supply materials, trucking and a warehouse, with $3.36M of related-party payables at year end. The board renewed a 15% poison pill on August 5, 2026 running to July 5, 2027 and added Bradley Pickard, a rail and logistics investment banker, as a director on June 10, 2026 (8-Ks 2026-08-05, 2026-06-16).

5. Management: said versus did

On May 4, 2026 the company reaffirmed 4,000-4,500 deliveries, $500-550M revenue and $41-50M adjusted EBITDA, and the CEO called Q1's 16.8% gross margin "our highest quarterly gross margin in over a decade." Thirteen weeks later: 3,500-3,900 deliveries, $410-460M revenue, $36-44M EBITDA (8-Ks 2026-05-04, 2026-08-03). The reaffirmation did not survive a quarter.

The incentive design predicted the shape of the year. For 2025 the bonus plan paid adjusted EBITDA at 84% of target and backlog at 0%, and the Chief Commercial Officer carries backlog at 45% weighting against 25% each for EBITDA and cash flow (DEF 14A). In 2026 backlog rose 121% and manufacturing margin collapsed. All 12 Form 4s in twelve months are grants (code A); no insider has bought a share with the stock down 52%.

6. Valuation

Fully diluted shares about 34.4M (32.78M plus the 1.64M-share 2023 Warrant); net debt about $42M; EV about $286M.

Base (50%): FY2027 revenue $520M as backlog converts, $12M of Castaños savings run full year, aftermarket $50M, adjusted EBITDA $55M at 6.5x with net debt $35M, giving $9.30. Bear (30%): backlog converts at Q2 margins and the order cycle rolls over, adjusted EBITDA $28M at 5.5x with net debt $45M, giving $3.20. Bull (20%): deliveries above 4,500 with gross margin back to 15%, adjusted EBITDA $70M at 7.5x with net debt $25M, giving $14.50. Probability-weighted value is about $8.50 against $7.10, roughly 20% upside for a business that earned a 3.2% manufacturing gross margin last quarter.

Reverse DCF: at $7.10 the $286M enterprise value requires about $28.6M of after-tax operating profit held flat forever at a 10% cost of capital, which is FY2025's $33.9M of EBIT taxed at 15% and never growing again.

The decisive number is the implied second half. Guidance of $36-44M against $4.3M booked in H1 requires $31.7-39.7M on H2 revenue of $233-283M, a 13.8% adjusted EBITDA margin at the midpoint against 8.2% for all of FY2025 and 7.8% in the best recent quarter. The guide asks for the best half-year margin in the company's modern history immediately after its worst quarter.

7. Catalysts and timeline

Q3 2026 results, early November: the first period carrying the $12M of savings and the ramp, and the only real test of the guide. Exercise of the last 1,636,313-share warrant retires the $14.0M liability. A PIMCO monetization would clear a 46% overhang and is the largest potential re-rating event.

8. Risks and pre-registered kill criteria

Structural risks: one plant, in Mexico, leased from a director's family, exposed to tariff changes the company flags itself (10-Q); five customers at 75% of revenue; no tangible asset floor; 9.7% money against a cyclical order book. Kill criteria, all observable in filings: 1. Q3 2026 consolidated gross margin below 11%, or FY2026 adjusted EBITDA guidance cut again below $36M. 2. Backlog units decline sequentially at Q3 or Q4 2026, or backlog value per unit falls below $80,000. 3. Aftermarket segment operating income below $2.3M in each of two consecutive quarters despite revenue growth. 4. Net debt above $75M at any quarter end, or disclosure of a Term Loan covenant waiver or amendment.

9. Verdict and one-paragraph summary

WATCH, conviction 2. FreightCar America looks like a 71% share issuance on top of a broken cyclical, and only the second half is true: the 13,619,377 shares PIMCO received in June were penny warrants already inside the 31.8M weighted basic count a year earlier, so nothing was diluted, while the exercise cut the warrant liability from $168.5M to $14.0M and took equity from negative $107.4M to positive $36.2M. What is genuinely broken is the plant. Q2 gross margin fell from 15.0% to 5.5% and manufacturing to 3.2% on a late ramp, management reaffirmed guidance in May and cut it in August, and the remaining guide of $36-44M against $4.3M earned in H1 requires a 13.8% second-half EBITDA margin from a company that has never beaten 8.2% for a year. The backlog is real at 3,972 units and $344M, up 121%, but its margin is undisclosed and a 45% order share won at a 3.2% manufacturing margin reads as much like price as like product. Probability-weighted value is about $8.50 against $7.10, no insider has bought a share in twelve months, and free cash flow is swept to a 9.7% term loan. The setup is interesting for who owns it, a 46% PIMCO stake, a renewed 15% pill and a rail M&A banker just added to the board, but the price is not low enough to pay for a guide this back-end loaded. The November Q3 print settles it.

Research for discussion, not investment advice. Positions and sizing are the reader's decision.