Babcock & Wilcox Enterprises (BW)

The sleeper story is that existing coal plants at 50% utilization could drive a massive parts/services supercycle without any new builds, but management is mesmerized by shiny AI data center projects. The $2.5B booking headline is almost entirely one project that won't generate meaningful revenue until 2027-2028, creating a revenue recognition cliff nobody's discussing. The binding constraint isn't demand but execution and timing: they can't convert pipeline to revenue fast enough to justify the valuation. Most revealing quote: 'We kept the guidance where we are' despite claiming transformational momentum. My view changes if they book a second major AI project this quarter with 2026 revenue recognition.