Sierra Turbine: A Backlog That Does Not Mean What It Looks Like
Sierra Turbine Systems finished FY2024 with revenue of $921 million and adjusted EBITDA of $138 million. Year-end backlog was $520 million: $180 million is a non-binding framework agreement with a utility covering 2026-2029 deliveries with no minimum purchase obligation, and $340 million is firm purchase orders, of which management says 'roughly half' ships within twelve months, consistent with a historical 50-60% conversion rate. Aftermarket - short-cycle, book-to-ship under 90 days - was 38% of FY2024 revenue. Both plants ran at 88% and 92% of practical capacity in Q4; a third plant adds 25% of capacity in July 2026. A 6% list price increase took effect January 2025. Nickel-based superalloy costs are up 9% year over year and are 31% of COGS. Guidance is FY2025 revenue of $1.05 billion (+14%) and a 15.5% EBITDA margin.
1. Build the FY2025 revenue bridge. What can you actually count from backlog, and what has to be forecast from scratch?
Read the worked answer
Step 1 - strip the framework agreement. The $180M utility agreement has no minimum purchase obligation and a delivery window running to 2029. It is an option, not an order. Firm backlog = $340M. Counting the full $520M makes backlog coverage look like 520 x 0.55 / 1,050 = 27% of guided revenue; the correct figure is 340 x 0.55 / 1,050 = 18%. A 9-point difference in 'backlog-supported' revenue is the entire margin of safety in the story.
Step 2 - apply a defensible conversion rate. Management says 'roughly half,' history says 50-60%. Use 55% as a midpoint and show 50% as the downside: 340 x 0.55 = $187M from firm backlog inside FY2025.
Step 3 - the rest has to be earned. Revenue of $1,050M less $187M = $863M (82%) must come from short-cycle aftermarket and orders not yet booked. Aftermarket FY2024 = 38% x 921 = $350M; at price plus modest volume that is roughly $375-390M. That leaves ~$475M from OEM book-and-ship that has not been ordered yet - on revenue that also faces a capacity ceiling. The forecast is therefore almost entirely an order-intake and capacity call, not a backlog call.
Step 4 - check the price/backlog interaction. Read the backlog footnote: if firm orders are stated at contract prices with escalation clauses, some of the 6% increase is already inside the $340M and applying price on top double counts. If backlog is stated at the price in effect when the order was taken, price must be layered on. This single question moves FY2025 revenue by up to 6%.
2. Is 14% revenue growth achievable given the capacity position? Show the arithmetic.
Read the worked answer
Capacity is a constraint in hours, not dollars - that distinction is the whole answer.
Hour headroom: blended utilization is roughly 90% (88% and 92%, weighted by plant size). Practical ceiling is about 95%; beyond that, overtime, scrap and maintenance degrade the plant. Hour headroom = 95/90 - 1 = +5.6%.
Dollar headroom at constant hours: a 6% price increase raises revenue per hour by 6% without consuming a single additional machine hour. Revenue capacity = 1.056 x 1.06 - 1 = +11.9%.
Against a 14% guide, that leaves a ~2-point gap requiring overtime, debottlenecking, or pulling the third plant forward from July 2026. Feasible, but it is an execution assumption, not a structural given.
Sensitivities:
| Price realization | Capacity-implied revenue growth | Gap to 14% guide |
|---|---|---|
| 6.0% (full list) | +11.9% | -2.1 pts |
| 4.0% | +9.8% | -4.2 pts |
| 2.0% | +7.7% | -6.3 pts |
The key check is realized versus list price. List increases are routinely eroded by competitive discounting, fixed-price long-term service agreements, and customer-specific rebates. Ask for realized price per unit or price per machine hour, not the list change. If realized is 3-4% rather than 6%, the guide requires volume growth of roughly 10% into a plant with 5.6% hour headroom.
Model 10-12% revenue, not 14%, and show the capacity constraint explicitly. Note also that the 25% capacity addition lands mid-2026, so it contributes nothing to FY2025 other than relieving the second half.
3. Build the price/cost bridge on margin. Does the 15.5% EBITDA guide make sense?
Read the worked answer
Assumptions: COGS at ~76% of revenue (consistent with a 15% EBITDA margin and ~4% D&A); superalloy at 31% of COGS = 23.6% of revenue; other COGS 52.4% of revenue; SG&A ~9% of revenue. All effects stated in points of revenue.
| Driver | Effect (pts of revenue) |
|---|---|
| Price +6.0%, ~100% drop-through | +6.0 |
| Superalloy +9% on 23.6% of revenue | -2.1 |
| Other COGS inflation ~3% on 52.4% | -1.6 |
| SG&A inflation ~3% on 9% | -0.3 |
| Net | +2.0 |
FY2024 margin was 138 / 921 = 15.0%. A clean price/cost bridge implies ~17.0%, before any operating leverage on volume. Guidance is 15.5%.
A 1.5-point gap (~$16M on $1.05B) means one of three things, and they have very different implications:
- Price is not sticking. Management knows the 6% list will realize at 3-4%. That is a competitive-erosion signal and it also breaks the revenue guide - the two shortfalls compound.
- Cost inflation is worse than superalloy. Energy, labor, freight, or supplier pass-through not visible in the input disclosure. Ask for the FY2025 standard-cost update.
- Mix is diluting. OEM original equipment units carry lower margin than aftermarket parts; if OEM outgrows aftermarket, blended margin falls even with favorable price/cost. Test the aftermarket share against the 38% base - every point of mix shift toward OEM costs roughly 30-40 bp of blended margin.
Most likely reading: a modest guide against a favorable bridge is a signal, not conservatism. Management guides to what it can see. Either take revenue down to 11-12% and hold margin near 15.5-16.0%, or hold revenue and take margin to 14.5%. Taking both the 14% top line and the 15.5% margin requires the price to stick at list and the plants to run above practical capacity - two favorable assumptions in the same model.