Fortress Infrastructure
Fortress Infrastructure $FIP appeared on our radar in early 2025 as a classic infrastructure ramp story — assets with genuine barriers to entry, an EBITDA inflection already underway, and a catalyst calendar that was unusually legible. We built a position and rode the re-rating through the first half of the year. We exited on the day management announced the Wheeling acquisition and the Ares refinancing, fast enough to analyse the terms of the preferred structure and exit at the opening. The stock traded up sharply on the announcement before the market had absorbed the details of the preferred structure. Within days, most of those gains reversed.
Since October 2025, we have had FIP on the shelf. The reasons were simple: the capital structure introduced by the M&A was, in our view, deeply unfavourable to common shareholders, the management team’s communication track record was poor, and there was no obvious near-term catalyst to change either of those facts. Nothing has changed structurally. But with the post-transaction results now available and the Wheeling integration underway, it is a reasonable moment to revisit the question of whether the underlying asset quality — which we have always thought was real, particularly in the railroad and Long Ridge — creates a monitoring opportunity. This analysis is that exercise: an honest assessment of where the business stands right now.
$FIP FY25 Results Overview
Management describes Q4 as a record quarter, and on the EBITDA line that is technically correct: $80.2M of adjusted EBITDA was the highest quarterly figure in the company’s history. The problem is that EBITDA is not what matters here — cash is. And on a cash basis, Q4 was not a good quarter. Interest expense came in at $90.3M, meaning the business did not cover its consolidated interest bill even in its best-ever quarter.
The gap between EBITDA and interest expense widened from -$2.4M in Q3 to -$10.1M in Q4 — despite EBITDA growing $9.3M — because the Wheeling bridge loan was reflected for a full quarter for the first time (it only contributed five weeks in Q3). Total cash fell $28M. Total obligations — debt plus preferred — grew $77M, largely from PIK accrual on the RailCo Pref A. The capital structure gained more ground on the business in Q4 than in any prior quarter.
For the full year 2025, adjusted EBITDA was $232.3M against $127.6M in 2024. Management extrapolates from Q4 to claim an annualised run rate of “just over $320M”. Unlike much of what management says, this number is actually defensible — and if anything conservative. Q4 included a planned 8.5-day maintenance outage at Long Ridge in October and an unplanned 19-day steam turbine repair in December that management estimates cost $3.5M of EBITDA. Annualising a quarter that absorbed two outages likely understates the clean run rate. We are comfortable using $320M as the working baseline for 2026, with the caveat that it assumes the Wheeling integration proceeds and Jefferson’s ammonia contract contributes for a full year. The harder truth is that even taking $320M at face value, the interest bill on the post-refinancing capital structure runs to approximately $350–360M annually. The arithmetic does not close without the Long Ridge sale.