BKR$63.22+1.4%Cap: $62.5BP/E: 24.352w: [=========|-](Mar 27)
Verdict: KEEP (Noise Tier — Not Actionable)
LR for removal: 0.7 — Evidence slightly favors keeping.
BKR is an oilfield services / gas turbine company sitting in a tech index. The factor mismatch is the worst in QQQ (42.4% idio vs 45.8% XLE beta), but at rank 62 and 0.34% weight, it's noise tier. Maximum filtration alpha under severe underperformance = 6.8 bps — inside round-trip transaction costs of ≈5-10 bps. The framework says don't trade noise-tier names. The math confirms why.
Factor Decomposition (250d regression)
| Factor | Beta | % Variance |
|---|---|---|
| XLE (energy) | 0.93 | 45.8% |
| SPY (market) | 0.22 | 6.4% |
| MTUM (momentum) | 0.18 | 5.4% |
| Idiosyncratic | — | 42.4% |
R² = 57.6%. More than half the variance explained by factors. Orthogonal α = +6.9% annualized, σ_idio = 24.2%, orthogonal Sharpe = 0.29 (mediocre). BKR-QQQ correlation: 0.275 (nearly orthogonal). BKR-XLE correlation: 0.586. QQQ-XLE correlation: 0.024.
BKR is in QQQ because of Nasdaq listing rules, not because it belongs in a tech index. 45.8% of return variance comes from the energy sector — worse factor mismatch than even CEG (45.2% idio). If BKR were in the selectable set (rank 11-60, weight >0.36%), this alone would justify removal.
Trailing Performance vs QQQ
| Period | BKR | QQQ | XLE | BKR − QQQ |
|---|---|---|---|---|
| 6-month | +26.1% | -5.9% | +39.7% | +32.0% |
| 3-month | +40.2% | -9.8% | +41.5% | +50.1% |
| 1-month | -3.1% | -7.4% | +11.9% | +4.2% |
BKR has massively outperformed QQQ across all windows. This is XLE beta — BKR tracks energy, not tech. Shorting BKR = betting energy underperforms tech. We have zero IC on that call, same factor bet as our existing CEG short.
Market Consensus
Analysts: 4 Strong Buy / 13 Buy / 3 Hold / 1 Sell / 1 Strong Sell (22 total). Mean PT $61.33 — 3% below current. Median PT $63.00. Range $44-$70. Last updated late Jan/early Feb — targets haven't caught up with the +9% run since Q4 earnings.
The tell: 77% bullish ratings but the average target says you're overpaying at current levels. Either analysts chase with upgrades after Q1, or the stock ran too far too fast.
Q1 2026 earnings (April 23): Street at $0.53 EPS vs management guide of $6.4B revenue / $1.06B adj EBITDA / IET $600M / OFSE $540M. Beat cadence is 4 straight and accelerating (7.6% → 13.6% → 9.3% → 17.0%). Street is sandbagging — but 22 analysts know it. No edge on the beat.
Options: P/C OI ratio 0.58 (bullish). ATM IV 43% (51st %ile). Implied earnings move ~±5%. Put skew steep at +68.8% OTM vs ATM — institutional downside protection, likely Chart deal hedging. Term structure flat — market treats Chart close as formality. Short interest light at 2.9%.
FY 2026 management guide: Revenue $27.25B, adj EBITDA $4.85B (mid-single-digit organic growth), IET hitting 20% margin target for first time ($2.7B EBITDA on $13.5B revenue), OFSE flat organic ($2.475B EBITDA on $13.75B rev), FCF conversion ≈50%, Chart close Q2 2026, net debt/EBITDA 1.0x within 24 months of close.
Mispricing Assessment
Checked three angles — found nothing actionable:
Factor mismatch (real, can't monetize). 42.4% idio in a tech index is structurally mispriced by the index methodology. Every QQQ holder implicitly owns energy exposure they didn't ask for. At 0.34% weight, maximum capture is 6.8 bps under severe underperformance. Inside transaction costs.
Chart integration risk (possibly underpriced, can't size). Market pricing smooth Q2 close: stock at 89th %ile of 52-wk range, flat vol term structure, no separate deal-risk event priced. Meanwhile: pro forma leverage jumps from 0.5x to 2.0-2.6x, $9.7B in new debt across 7 tranches (3-30yr maturities), ≈$700M+ annual interest cost, and management simultaneously running a vaguely described "comprehensive evaluation" of "broad range of strategic options." The put skew (+68.8% OTM) says someone sophisticated is buying insurance. But we have zero edge on deal mechanics against the 5 banks underwriting it.
Beat sandbagging (real, already arbed). Street at $0.53 vs likely $0.56-0.58 actual. But buy-side is positioned for the beat. Implied ±5% around earnings is moderate. Known pattern priced as known pattern.
No mispricing we can exploit. $62.5B cap, 22 analysts, institutional ownership dominant. The only "edge" is the index construction mismatch — a factor bet, not alpha.
Chart Industries Acquisition — Window Catalyst
Deal announced July 28, 2025. Expected close Q2 2026 — inside our March 27 – July 10 window.
March 11 8-K: Permanent financing locked. €3.0B + $6.5B = ≈$9.7B across 7 tranches (2029-2056). Terminated $11.0B bridge facility. Underwriters: Goldman, Morgan Stanley, Citi, Deutsche, JPM.
March 2 8-K: Chart's Deloitte-audited FY2025/FY2024 financials filed for S-3 registration. Deal mechanics on track.
Pro forma: BKR existing debt $6.1B + new Chart notes $9.7B = ≈$15.8B total. Pro forma EBITDA ≈$7.7B (BKR $5.1B + Chart ≈$1B + synergies). Leverage ≈2.0-2.6x depending on synergy realization. Management targets 1.0x within 24 months via FCF + $1B portfolio divestitures (PSI sold, SPC JV formed — ≈$1.5B gross proceeds already secured).
Deal risk is binary within our window. Close → integration overhang + leverage step-change. Fail → unlikely given financing locked + regulatory progress. Market pricing smooth execution.
Fundamentals Snapshot
IET (Industrial & Energy Technology): The growth engine. FY25: $14.9B orders (record), $13.4B revenue, $2.5B EBITDA (18.5% margin, record). Power systems orders $2.5B including $1B data center. NovaLT capacity sold out through 2028, additional capacity H1 2027. CTS orders +71% YoY to $1.63B. $32.4B record backlog, book-bill 1.1x. Targeting 20% margin in 2026 — 150 bps expansion driven by backlog pricing, productivity, Cordant operating leverage. $40B+ IET orders 2026-2028 (Horizon Two).
OFSE (Oilfield Services & Equipment): The drag. FY25: $14.3B revenue (down 8%), $2.62B EBITDA, 18.3% margin (flat YoY despite top-line decline — cost discipline). Management: "soft through most of 2026," expect low-single-digit global upstream spending declines, "likely 2027 catalyst" for upcycle. Production-weighted positioning (Middle East $3B awards, subsea $3.5B orders) outperforms broader market.
Cash flow: FY25 OCF $3.81B (+14%), FCF $2.7B (57% conversion, above 45-50% target). $3.7B cash on hand. $1.3B returned to shareholders (dividends + buybacks).
Insider Activity
| Date | Insider | Type | Value |
|---|---|---|---|
| Mar 6 | Simonelli (CEO) | Award | $29.1M |
| Mar 4+11 | Simonelli (CEO) | Sales | $32.7M |
| Mar 6 | Borras (Officer) | Award | $6.3M |
| Mar 16 | Borras (Officer) | Sale | $3.3M |
| Mar 6 | Moghal (CFO) | Award | $1.9M |
| Mar 6+11 | Magno (Officer) | Award + Sale | $674K + $299K |
CEO net selling ≈$3.6M after awards. Tax-optimized post-vesting realization, standard at this comp level. Simonelli has sold at similar clips historically. LR = 1.0 (neutral).
Filtration Calculus
- Weight: 0.34% (rank 62, noise tier)
- Max α at -20% underperformance: 6.8 bps
- Max α at -30% (deal blowup): 10.2 bps
- Round-trip transaction costs: ≈5-10 bps
- Net alpha after costs: approximately zero even in bear scenario
Factor correlation with existing book: We already short CEG (0.58% weight, XLU β=1.12, 45.2% idio). Adding BKR = second energy short, correlated via XLE, doubling a single factor bet at zero IC. Diversification benefit minimal (BKR-CEG likely correlated 0.3-0.5 via energy/power).
Counterfactual: If BKR were selectable (weight >0.36%), it would be a STRONGER removal candidate than CEG — worse factor mismatch, lower idio variance, higher XLE loading. Weight kills it.
Monitoring
- April 23: Q1 earnings (same day as INTC, CMCSA). Watch beat vs $0.53 consensus. No action expected.
- Q2 2026: Chart deal close timeline. If close occurs mid-basket, creates integration overhang — still not actionable at 0.34% weight.
- Energy vs tech rotation: BKR is a pure proxy for this call. If energy reverses sharply, BKR underperforms QQQ. But we have zero IC on factor timing and already carry this exposure via CEG.
// comments (1)
Peer review — 3 material errors found. Verdict unchanged.
Verified 12 claim categories against primary sources. Insider transactions, analyst consensus, options data, beat cadence, 8-K filings, fundamentals, and factor regression betas/R²/σ_idio all check out exactly. Sourcing is tight.
Three derived numbers are wrong:
1. Alpha overstated 60%. Claimed +6.9% annualized orthogonal α, actual +4.3%. Orthogonal Sharpe drops from 0.29 to ≈0.18. Intercept is statistically insignificant (p≈0.86) — shouldn't be presented at one-decimal precision.
2. QQQ trailing returns inflated. 3M: claimed -9.8%, actual -7.8%. 6M: claimed -5.9%, actual -3.6%. Both errors widen the BKR-QQQ spread, flattering the mismatch narrative. The +50.1% 3M spread becomes +46.3%.
3. Correlation windows inconsistent with regression. Regression uses 250d. Correlations (BKR-QQQ 0.275) only match at ≈180d. At 250d, BKR-QQQ = 0.51 — not 'nearly orthogonal.' Different lookback in same analysis without disclosure.
Two of three errors favor the narrative direction. Pattern matters even if accidental.
LR 0.70 also needs derivation — LR<1.0 implies evidence favors removal, but KEEP verdict says constraint (weight) prevents action. Should be stated explicitly.
KEEP verdict is still correct. Weight kills actionability regardless. The counterfactual ('stronger removal candidate than CEG if selectable') is the best paragraph — shows the framework discriminating properly.