Time Horizon

12-24 months. DEMAND type thesis with 365-day half-life. Alpha decay ≈0.11%/day. No binary catalyst — convergence is incremental, driven by quarterly earnings disclosures, Class D redemption filings, and the eventual resumption of coverage. The forcing function is not one event but the accumulation of evidence into a vacuum where no one is currently modeling the business.

Key dates within horizon: FY2026 earnings May 28, 2026 (first full-year report of the acceleration). Class D preferred put right activates July 2, 2027. Coverage initiation (40% probability by Dec 2026).

Base Rate

Reference class: small-cap MLPs with zero sell-side coverage undergoing fundamental business transformation (segment simplification + deleveraging + volume acceleration).

This is a thin reference class. The closest analogs are midstream MLPs that simplified from multi-segment businesses into single-asset pure-plays during 2018-2023 (e.g., Crestwood Equity, Summit Midstream, MPLX restructuring). Typical outcome: multiple expansion of 1.5-2.5x over 18-36 months as coverage initiates and complexity discount fades, contingent on EBITDA delivery.

Base rate: Zero-coverage MLP with accelerating EBITDA → re-rating within 24 months
Prior: ≈55% (limited sample, but structural simplification + coverage gap is a known pattern)
Prior odds: 1.22

The base rate is uncertain because the reference class is small. What strengthens it: NGL has a direct comp (ARIS takeout at 10x, Oct 2025) showing the market DOES pay higher multiples for produced water assets. What weakens it: MLP structure + complex preferred stack + IDRs actively deter institutional ownership, which could make the information vacuum self-reinforcing.

Alpha vs Beta

Factor regression (250 daily observations, NGL ~ SPY + XLE + AMLP + IWM):

Total trailing 1Y return:           +169%
  Market beta (SPY, β=0.26):          ~+5%     (not significant, p=0.52)
  Energy sector (XLE, β=0.25):        ~+3%     (not significant, p=0.29)
  Midstream MLP (AMLP, β=0.80):       ~+8%     (significant, p=0.034)
  Small cap (IWM, β=0.24):            ~+2%     (not significant, p=0.46)
  Idiosyncratic alpha:               ~+151%    (the actual move)

R-squared = 16.5%. Idiosyncratic variance = 83.5%, well above the 75% target. The only statistically significant factor is AMLP (midstream MLP index) at 8% of variance. This stock moves on its own fundamentals, not sector or market dynamics.

Forward decomposition (estimated):

Total 12-24mo return forecast:  +40-90% (8.3x → 10-12x multiple expansion + EBITDA growth)
  Market beta:                   ~+4-8%  (SPY long-run, β=0.26)
  Midstream beta:                ~+3-6%  (AMLP, β=0.80)
  Idiosyncratic alpha:          +30-76%  (the thesis)

The thesis is almost entirely idiosyncratic. You're betting on NGL-specific dynamics: volume acceleration, margin expansion, Class D resolution, coverage initiation, distribution reinstatement. Factor exposure is negligible.

B -- Business Model

NGL Energy Partners is the largest independent produced water disposal company in the United States. It operates a toll-road business in the Permian Basin: producers pay NGL approximately $0.60 per barrel to collect, transport, and inject their produced water underground.

Revenue decomposition (Water Solutions, 83.5% of EBITDA):

The unit economics: $0.60/bbl fee, $0.18/bbl operating cost, $0.42/bbl EBITDA margin (74.1%). At 47.6% utilization of 6.46M bbl/day permitted capacity, incremental barrels arrive at near-zero marginal cost. This is infrastructure leverage in its purest form.

Two other segments exist -- Crude Oil Logistics and Liquids Logistics -- but they're in structural decline from divestitures and lower commodity volumes. NGL is converging toward a pure-play water company. The market still sees a diversified midstream MLP; the filings show a single-asset story with captive producer customers.

CF structure: Fee-based (79% of Water Solutions), partially commodity-exposed (14% skim oil, linked to WTI). MVC contracts provide a volume floor (≈1.5M bbl/day minimum). RPO increased 54% in 9 months ($385M to $595M, 10-Q). Deficiency payments declining because producers are exceeding their minimums -- a positive signal masked as a negative in the fee/bbl trend.

Sources: 10-K pp.56-57 segment tables (q=0.95), 10-Q pp.42-43 (q=0.95)

Phi -- Financial Trajectory

The Acceleration

Water Solutions volumes: +4.5% (FY2024) to +8.6% (FY2025) to +17% (Q3 FY2026). Delaware Basin specifically: +19.2% YoY in Q3 FY2026. All-time daily disposal record of 3.5M bbl/day set January 16, 2026. This is not a one-quarter spike -- it's three years of accelerating growth driven by the structural water-oil ratio tailwind (mature Permian wells produce 8-12 barrels of water per barrel of oil, vs 3-4 for new wells).

EBITDA trajectory: $594M (FY2023) to $623M (FY2025) to $660M (FY2026E) to at least $700M (FY2027 guidance). Guidance raised twice during FY2026 (from $615-625M to $650-660M). Q3 run-rate annualizes to $690M.

Water Solutions EBITDA margin: 66.4% to 69.6% to 71.7% to 74.1% (Q3 FY2026). Expanding approximately 200 basis points per year on operating leverage. OpEx/bbl declining from $0.24 to $0.18 as fixed-cost infrastructure absorbs volume.

FCF Conversion: Temporarily Depressed

FCF/EBITDA dropped from 61% (FY2023) to 37% (FY2025) to ≈42% (9M FY2026). The cause: expansion capex nearly tripled, from $79M to $200M+, against initial guidance of $105M. CFO Cooper on Q2 call: "Began fiscal year modest growth expectations as initial growth CapEx guidance about $60 million. Contract volume requires additional $100 million growth CapEx." This is demand-driven capex -- new customer contracts materializing faster than expected. The right problem.

Maintenance FCF yield to common equity: 16.4% ($255M on $1.55B equity market cap).

The Capital Structure Labyrinth

$3.2B pro forma debt + $818M preferred ahead of $1.55B common equity. The waterfall to common unitholders:

If management distributes 50% of maintenance FCF: approximately $1.10/unit/year, or an 8.8% yield at current price. This is gated by the 4.75x leverage covenant and Class D retirement.

ROIC vs WACC: The Inflection

ROIC is crossing WACC right now. Water Solutions standalone ROIC is already 11-12%, dragged down by the declining legacy segments. As those segments shrink further, consolidated ROIC converges toward Water Solutions' economics.

Sources: 10-K pp.56-78 (q=0.95), 10-Q pp.4-59 (q=0.95), 8-K 2026-03-12 term loan (q=0.95), Q2/Q3 earnings transcripts (q=0.85)

K -- Competitive Position

Moat Stack

Regulatory franchise (K_reg). The Texas Railroad Commission indefinitely suspended deep produced water injection permits in Culberson and Reeves counties (January 2024). NGL lost only 10,000 bbl/day (0.15% of capacity) but gained an effective permitting freeze on new competition. NGL holds 6.46M bbl/day of existing permitted capacity -- 3.4M bbl/day of unused headroom that competitors cannot replicate through new permits. 10-K: "With our unique positioning outside of the affected areas, we have the ability to grow our asset base." The moat is geographically bounded to these counties, not basin-wide, but it covers NGL's core Delaware operations.

Infrastructure scale (K_scale). 800+ miles of large-diameter pipeline, 194 injection wells, 90 facilities. 98% of Delaware Basin water arrives by pipeline, not truck. Replacement cost estimated at $3.3-6.8B versus $5.3B enterprise value -- NGL trades at or below the cost of rebuilding its physical network. OpEx/bbl at $0.18 and declining. Near-zero marginal cost per incremental barrel at 47.6% utilization.

Switching cost (K_switch, phi=0.7-0.85). 765,000 dedicated acres under multi-decade contracts. Producers switching from pipeline ($0.60/bbl) to truck ($3-5/bbl) face a 5-8x cost increase. RPO growth of 54% in nine months ($385M to $595M) demonstrates new long-term contracts being signed. MVC deficiency payments declining because actual volumes exceed minimums.

Customer Concentration

73% of Water Solutions revenue from top 10 unnamed customers, described as "large, investment grade" E&Ps and "super majors." The LEX II pipeline expansion was underwritten by a single investment-grade producer with a 200,000 bbl/day MVC and multi-decade acreage dedication -- most likely ExxonMobil (post-Pioneer), ConocoPhillips, or Chevron based on Delaware Basin geography. NGL does not name customers in any filing.

Pore Space: The Unquantified Tail Risk

10-K: "The amount of subsurface pore space that is capable of permanently storing injected produced water is finite and requires constant replenishment."

No reserve life is disclosed anywhere -- not in the 10-K, not in the 10-Q, not in the DEF 14A. This is the single biggest unquantifiable risk in the thesis.

Partial quantification from transcripts (EVP Doug White, Q2 FY2026 call): approximately 4M bbl/day from legacy injection wells + 1M bbl/day from Andrews County expansion + 800K bbl/day from a pending TPDES surface discharge permit = ≈5.8M bbl/day total capacity. Current utilization: 3.07M bbl/day. Headroom: ≈2.7M bbl/day. At ≈500K bbl/day annual growth: approximately 5.5 years of headroom. Adequate for the thesis timeline but unverified by any independent geological assessment. Management claim, q=0.85.

Competitive Landscape

Aris Water Solutions was acquired by Western Midstream at approximately 10x EV/EBITDA in October 2025, removing the direct public comparable. NGL handles 3x ARIS's volume (3.07M vs ≈1.0M bbl/day) and trades at a 20% discount (8.3x vs 10x). Remaining competitors: WaterBridge (private, Five Point/LandBridge/FANG ecosystem), Select Water Solutions (WTTR, smaller, less Delaware-focused), and E&P captive operations.

Sources: 10-K pp.570-629, lines 1381-1387, 2167-2175, 2879-2883, 4525-4529 (q=0.95). 10-Q pp.42-50 (q=0.95). Q2/Q3 transcripts (q=0.85). WES/ARIS acquisition filing (q=0.95).

G -- Governance

Structure. Standard MLP. GP controls with 0.1% economic interest plus IDRs. Board of 6 (3 independent). LP voting capped at 20%, no director elections, no annual meetings. GP resolves conflicts under "good faith" standard (below fiduciary). This is typical for midstream MLPs but below C-corp governance standards. IDRs take 48.1% of marginal distributions above $2.03/year -- dormant at current zero distribution, punitive at maturity.

Compensation. CEO Krimbill total: $2.79M (salary $835K + discretionary bonus $2.0M). Pay ratio 26:1. No formulaic performance metrics (governance weakness). Philosophy explicitly tied to distribution reinstatement (alignment positive). No severance, no golden parachutes. LTIP expired May 2021 -- management had zero equity compensation for five years. New 2025 LTIP (10M units, 7.95% dilution) approved February 2026, no awards granted yet.

Insider signal. Net buying of $1,042,715 over the past 12 months, zero sales. CEO Krimbill bought 100,000 units at $4.63 (February 2025, now +170%). Director Collingsworth bought 100,000 units at $5.80 (September 2025, now +116%). Both open-market purchases with personal capital. Neither has sold a single unit.

Officer stability. No departures. Team stable through the entire transformation: CEO Krimbill (founder, 15+ years), CFO Cooper (since 2020), EVP Water Solutions White (3+ years).

Class D preferred holders. EIG Neptune (energy PE), FS Energy and Power Fund (institutional credit), GCM Grosvenor (alt asset manager). These are sophisticated institutional investors who understand how to exercise the July 2027 put right. The obligation is real, not decorative.

Governance flags. Aviation VIE: two $8.1M airplanes (total $16.2M), NGL owns 90%, management owns 10% with a put option forcing NGL to buy their stake. Dollar-immaterial ($424K NCI) but structurally unfavorable alignment. Prairie Operating: management purchased $9.9M of stock, mostly sold. Neither is thesis-altering.

Sources: DEF 14A 2025-12-29 (q=0.95). 10-K governance provisions (q=0.95). Form 4 insider transactions (q=0.95). 8-K 2026-02-09 LTIP (q=0.95).

Beta -- Factor Profile

Orthogonal Sharpe of 1.65 flags above the 1.5 threshold for potential misattribution. Assessment: genuine, not misattributed. Zero coverage + fundamental business transformation = real idiosyncratic alpha source. The stock went from $4 to $12.52 on business improvement, not sector rotation. All pairwise correlations are low (0.32-0.38 vs SPY, XLE, AMLP, IWM).

Options chain. Ghost town. Total put OI across all four expirations: 655 contracts. No institutional hedging because no institutional ownership. No meaningful mechanical floors or ceilings. The chain confirms the information vacuum -- nobody is here.

Thesis type: DEMAND. 12-24 month half-life. Alpha decay approximately 0.11%/day. No urgency. The thesis is durable and can tolerate patient entry.

Delta -- Expectations Gap

What the Price Implies

At $12.52 (EV/EBITDA 8.0x), the market prices NGL as a mature, zero-growth midstream MLP. Gordon Growth decomposition implies g* of approximately 0%, d* of 0 years above WACC, and flat margins at current consolidated levels. The forward P/E of 14.73 is built on a single stale EPS estimate from Wells Fargo, which had a $4 price target in June 2023. The stock has tripled since. No analyst has asked a question on an NGL earnings call. Management guides FY2027 EBITDA of at least $700M into an empty room.

There is no consensus to disagree with. The information channel is off.

Gap Table

All 10 gaps are positive. Every dimension is mispriced in the same direction because the same root cause -- the information vacuum -- affects all of them. When zero analysts cover a company, the market defaults to a generic multiple. All company-specific dynamics (volume acceleration, margin expansion, deleveraging, buyback, distribution path) go unpriced because they are unknown to the marginal buyer.

The three independently actionable gaps are volume growth (Tier 1 evidence, forcing function at May 28 earnings), Class D resolution (Tier 1 8-K evidence, mechanical quarterly filings), and the multiple gap (public ARIS comp, requires coverage initiation to close). Everything else is dependent on these three or probability-gated.

Steelman Bear Case

The strongest argument against this thesis: the information vacuum is self-reinforcing, and the Class D put right creates a capital crunch that could dilute common unitholders.

Here's the bear logic, engaged honestly:

NGL trades at 8x because it deserves 8x -- not because nobody's looking, but because the structural barriers to institutional ownership are real and durable. Consider the stack of deterrents: MLP structure (K-1 tax complexity excludes most mutual funds and many institutional mandates), IDRs (48.1% marginal take above $2.03/yr caps long-run unitholder returns), complex preferred stack ($818M ahead of common), zero distribution (excludes income-seeking MLP investors, which is the majority of the MLP buyer base), and low liquidity (ADV ≈$2.5M). Every potential buyer has a reason not to buy. The absence of coverage isn't a mispricing -- it's rational avoidance.

Now add the Class D put right: July 2, 2027, the remaining ≈511,000 Class D units ($752M at face) can be put to NGL. The $950M term loan provides ≈$262M in net proceeds for redemption. That leaves a ≈$490M gap that must come from (a) FCF generation between now and then (≈$250M of maintenance FCF over 15 months), (b) additional refinancing at what could be punitive rates (existing debt is already at 8-8.4%), or (c) some form of equity issuance that dilutes common unitholders. If the put is exercised in full and NGL can't cover it, the equity value compresses.

And pore space: management says 5.5 years of headroom. There is no independent verification. The 10-K explicitly says pore space "is finite and requires constant replenishment." If geological capacity constrains growth before the thesis plays out, NGL is a 5-7x EBITDA asset (declining MLP with a capped growth trajectory), not a 10-12x asset.

How we engage with this:

The bear case is partially right on the MLP structural deterrents -- they ARE real barriers to ownership, and they likely explain a 1-2x discount to C-corp midstream peers. But the bear case conflates "structural discount" (which might be 1-2 turns) with "total information failure" (which accounts for the other 2-3 turns of gap vs ARIS takeout). An 8x multiple doesn't just price in K-1 complexity; it prices in zero growth, zero coverage, zero distribution, and perpetual value destruction. That's too much discount for the structural issues alone.

On Class D: the bear assumes full exercise. But the holders are EIG, FS Energy, and GCM Grosvenor -- sophisticated PE firms who understand that exercising a put at par on an asset generating 16%+ FCF yield could trigger a liquidity event that destroys the value of their remaining stakes. The game theory favors negotiated retirement at a premium to market, not a hostile exercise that forces a fire sale. Management has been retiring $50M tranches every few months. The mechanism is working. The risk is real but the bear's scenario assumes adversarial holders, which contradicts the observed cooperative redemption pattern.

On pore space: we concede this point partially. 5.5 years of headroom is a management claim, not an audited reserve. The tail risk is unquantifiable. But the thesis timeline is 12-24 months, well within the claimed headroom. And NGL is actively diversifying its disposal geography (LEX II to Andrews County, TPDES surface discharge permit) precisely to reduce pore space concentration risk.

The bear case explains a discount. It does not explain THIS discount. A 20% haircut to ARIS's 10x (for MLP structure + cap complexity) gives 8x. But NGL handles 3x ARIS's volume, has higher margins, a regulatory permit moat ARIS didn't have, and is growing faster. The appropriate MLP-adjusted comp is 9-10x, not 8x. The gap is the information vacuum, not the structure.

Kill Criteria

Thesis dies if:
- FY2026 EBITDA < $620M (vs $650-660M guide) --> exit
  Signals: volume deceleration or margin compression
  Resolves: May 28, 2026 earnings

- Class D put exercised in full AND NGL cannot refinance within 6 months --> exit
  Signals: 8-K disclosure of put exercise, no concurrent financing announcement
  Resolves: July-December 2027

- Pore space capacity constraint disclosed (any TRRC enforcement or 10-Q risk factor addition) --> cut 50%
  Signals: new risk factor language, TRRC order, production curtailment
  Resolves: ongoing, monitor quarterly filings

- Pro forma leverage > 5.5x for two consecutive quarters --> cut 50%
  Signals: EBITDA miss + capex overshoot + debt increase
  Resolves: quarterly

- Major customer loss (>10% of Water Solutions revenue) --> exit
  Signals: volume drop >15% QoQ outside seasonal, 8-K disclosure
  Resolves: quarterly

Thesis strengthened if:
- FY2026 EBITDA > $680M (above high end of guide) --> add
- Sell-side coverage initiation with price target --> add (information vacuum collapsing)
- Class D retirement pace > $75M/quarter (accelerated timeline) --> add
- Common distribution declared (any amount > $0) --> add aggressively
- M&A approach disclosed (NGL as target) --> hold/evaluate

What to Watch

May 28, 2026: FY2026 earnings. The single most important data point in the near term. Resolves whether EBITDA hits $660M (our prediction at 72%), updates FY2027 guidance specificity (currently "at least $700M"), and discloses Q4 volume trajectory. Also the first chance to see cumulative Class D redemption progress post-$950M term loan.

Quarterly 8-K filings. Each Class D tranche redemption is a filing event. Track cadence: $127M redeemed in 9M FY2026. With $950M TL proceeds available since March 2026, the pace should accelerate. If it doesn't -- flag.

Coverage initiation. Our prediction: 40% by December 2026. The triggers that might prompt it: $700M EBITDA makes NGL impossible to ignore for midstream analysts, the $950M term loan was a capital markets event, and the unit price now exceeds typical institutional minimums ($10-15). If an initiation happens, it collapses the entire expectations gap table. Watch for: RBC, Barclays, Mizuho (active midstream coverage).

Permian rig count and WTI price. NGL's volume growth depends on continued Permian drilling. Water-oil ratio provides a structural floor (volumes grow even if drilling is flat), but a sustained oil price collapse below $50 would slow the growth engine. Monitor Baker Hughes rig count and forward WTI curve.

TPDES surface discharge permit. Currently pending. NGL has guided "early this year" (Q3 transcript, February 2026). If granted, adds ≈800,000 bbl/day of disposal capacity without pore space consumption -- transformative for the pore space risk narrative. If denied or delayed, the pore space overhang persists.

LR Signal

LR = 2.0 -- Mild-to-strong bullish. Market underpricing fundamentals, but convergence timeline is uncertain.

The evidence quality is high (Tier 1 filings, q=0.95 for the core volume/margin/Class D claims). The divergence from market pricing is large (all 10 dimensions mispriced in the same direction). But the convergence mechanism is slow and uncertain -- it requires either coverage initiation (40% probability) or incremental earnings accumulation into an information vacuum.

Why not higher: the meta-gap (coverage absence) could persist for years. MLP structural barriers are real. The Class D put right is a genuine financial risk. And pore space remains unquantified. A bullish thesis with an uncertain timeline and one unquantifiable tail risk caps the LR below 3.0.

Why not lower: the evidence is overwhelmingly Tier 1 (SEC filings). The volume acceleration is not a forecast -- it's a measurement in the 10-Q. The ARIS takeout comp is a public transaction. The insider buying is real money, not options grants. When every dimension aligns at high credibility, the prior should move.

Evidence