Xylem [XYL]
The memo issues NO position verdict. It scores Criteria and hands the book an analysis.
| criterion | verdict | basis |
|---|---|---|
| Quality | PASS | 14 consecutive profitable years; gross margin in a 36.9–39.6% band for fourteen years; Q2'26 operating margin 16.7%, +340bp YoY; four segments, all profitable; no metric retired or redefined across 12 annual reports and 10 transcripts |
| Accounting quality | PASS | same-quarter DSO +0.4 days (noise); DPO SHORTENED 0.5 days — the inverse of the CLS pattern; supply-chain financing disclosed and quantified at $269m (2.9% of revenue) and rising only $25m in six months |
| Growth durability | FAIL | organic revenue +1.3% (Q2), +0.5% (six months); Measurement & Control organic revenue −1.4% and orders −1.6%; total backlog fell 9.0% (FY2024 $5,070m → FY2025 $4,615m) while revenue rose 5.5% |
| Valuation (implied-path test) | FAIL | required 19.3% vs 1.3% demonstrated organic; −18.0pp; fails at every exit multiple in range including the screen's own 23.5x |
| Catalyst | PASS | Q3'26 results; FY2026 10-K backlog disclosure resolves the orders question |
| Capital allocation | INDETERMINATE | reported ROIC 8.80% vs an 8.5% WACC — marginal. Ex-goodwill ROIC is 63.8%, so the operating business is excellent and the $7.5bn Evoqua price is what drags the consolidated figure. Two defensible readings; a missing basis for choosing is INDETERMINATE, not FAIL |
Equity. Default under the brief; needs no argument. Realised 252-day volatility is 26.2% — the lowest of the three names in this cluster. No LEAP is proposed because the IV−RV spread was not measured this run, and the brief requires it to be measured, not assumed.
No entry is written to trade_recommendations.jsonl — the valuation output is a FAIL and the
ledger records specific recommendations.
| Thesis | a multiple re-rating, not a growth story. Essentially all of the +31.2% 12-month target is the move from the 6th to the 25th percentile of its own six-year multiple history. NTM revenue is only +3.7% on TTM |
| Horizon | 12 months |
| Size | 26.2% realised vol and 0.88x net leverage support a full-weight slot on risk grounds; the FAIL on the five-year test is the reason not to, not the volatility |
| Invalidation, hard | FY2026 total backlog below $4,615m in the FY2026 10-K. If backlog falls a second consecutive year while Xylem reports a 1.32 book-to-bill quarter, the orders metric is not measuring what it appears to |
| Invalidation, soft | organic revenue growth negative in any quarter; operating margin back below 14% |
| Downside case, named cause | The Q2'26 orders spike does not convert, or converts at Water Solutions and Services margins. $850m of the $879m organic orders increase came from WSS, which ran a 14.1% operating margin in Q2'26 and 12.1% for the half — Xylem's second-lowest. If the multiple instead reverts toward the six-year minimum of 2.93x on the same NTM base, the price is $111.65, −4.3% to spot. That is the floor the own-history series supports. Type: MEASURED — logged and scored, does not reject the name |
Quantified data-centre revenue. Applied Water grew organic orders +9.2% on data-centre projects, and "data center" mentions went from zero across three consecutive quarters to 19 in Q2'26 (13 of them in Q&A). Xylem has never put a dollar figure on it. If Applied Water's data-centre exposure is disclosed and is material, the 0.5–1.3% consolidated organic rate stops being the right numerator and this becomes a different analysis. Until then it is not, and the FAIL stands.