Back to Portfolio
SEI

Solaris Energy Infrastructure, Inc.

$65.62 USD 6.25B market cap April 15, 2026 (Refreshed from March 27, 2026)
Solaris Energy Infrastructure, Inc. SEI BUFFETT / MUNGER / KLARMAN SUMMARY
1 SNAPSHOT
Price$65.62
Market CapUSD 6.25B
EVUSD 5.50B
Net DebtUSD 715M (pre-Genco), ~$950M post-Genco
Shares58.1M (basic), ~70M fully diluted
2 BUSINESS

Solaris provides behind-the-meter, turnkey power-as-a-service to AI data centers and industrial customers. Owns and operates natural gas turbines (primarily Solar/Caterpillar 16.5 MW and 38 MW units) co-located at customer sites under long-term contracts (7-15 years). Also operates legacy oilfield logistics business (sand silo systems) generating $80M+ annual FCF. Key customers: xAI (Colossus campus, Mississippi) and undisclosed investment-grade tech company (Hatchbo LLC). Scaling from 780 MW operated to 3,100 MW by 2029.

Revenue: USD 622M Organic Growth: 99%
3 MOAT NARROW

First-mover in large-scale BTM gas generation for hyperscaler data centers. 2+ years operational track record at xAI Colossus. Switching costs from integrated power systems tied to air permits (7-15 year contracts with take-or-pay). Supply chain lock-in with secured OEM turbine delivery slots (2-3 year lead times). Vertical integration growing via HVMVLV (distribution), in-house SCR manufacturing, and Solaris Pulse monitoring software. Moat is real but narrow -- turbines are commoditized hardware, competitors (Halliburton, Liberty, Williams) entering market. Moat is temporal: strongest while demand massively exceeds supply.

4 MANAGEMENT
CEO: Bill Zartler & Amanda Brock (Co-CEOs, Zartler since founding 2014, Brock since Q3 2025)

Aggressive growth strategy: $647M CapEx in FY2025, $935M+ committed over next 3.5 years. Maintained $0.48/share dividend through transformation. Bold M&A (MER $200M, HVMVLV, Genco $620M). Two convertible bond issuances ($155M + $748M). Insider ownership improved to 10.5%. Strategy is high-conviction growth-at-all-costs.

5 ECONOMICS
21.8% Op Margin
~8% (distorted by growth phase; unit economics suggest 25-33% unlevered) ROIC
USD -438M (heavy growth CapEx; operating CF $209M) FCF
3.3x (pre-Genco) Debt/EBITDA
6 VALUATION
FCF/ShareUSD -6.26 (fully diluted, growth phase)
FCF Yield-9.5%
DCF RangeUSD 33 - 52 (fully diluted)

Pro forma EBITDA $800-900M by 2029 at 3,100 MW. Growth 30% years 1-5, 5% years 6-10, 2% terminal. WACC 10.5%. Maintenance CapEx 8% of revenue. ~70M fully diluted shares. At $65.62, stock trades 26-99% above fair value under optimistic assumptions.

7 MUNGER INVERSION -61.5%
Kill Event Severity P() E[Loss]
Valuation compression on any execution miss -35% 45% -15.8%
xAI customer concentration blowup (~37% of 3.1 GW target) -55% 18% -9.9%
Convertible dilution depresses per-share economics -15% 60% -9.0%
Regulatory/environmental shutdown of BTM gas generation -50% 15% -7.5%
AI CapEx winter / hyperscaler pullback -35% 20% -7.0%
Competitor entry compresses returns -25% 25% -6.3%
Debt/leverage crisis if growth stalls -40% 15% -6.0%

Tail Risk: Correlated downside: AI spending winter + environmental regulation + xAI financial distress could all hit simultaneously. In this scenario, Solaris is left with $1.5B+ debt on stranded assets generating $200-300M EBITDA. Convertibles convert at depressed prices, diluting remaining equity. Stock could decline 75-85%.

8 KLARMAN LENS
Downside Case

AI CapEx slows 30-40% in 2027-2028. xAI renegotiates JV terms lower. Environmental regs require costly upgrades. Competition from Halliburton/Liberty compresses margins. $1.5B+ debt on $300M EBITDA instead of $600M+. Convertibles dilute 15-20%. Stock $12-18 (5-7x distressed EBITDA on fully diluted basis).

Why Market Wrong

Market is pricing in perfect execution of 3,100 MW growth plan with no hiccups. Stock has risen 10x+ in 18 months on a nascent business model. Since March review, price rose another 8.5% without new earnings data. Trailing PE now 102x, EV/EBITDA 25x. Pro forma multiples look reasonable but require $2B+ in capital deployment to materialize. $903M convertibles will dilute 15-20%. No room for setbacks.

Why Market Right

Data center power demand is genuinely unprecedented and accelerating. Hyperscaler CapEx guidance is $600B+ for 2026. Grid constraints will persist for years. Solaris has first-mover advantage, secured turbine slots, and signed contracts. Customer diversification improving (xAI down from 67% to 37% of target capacity). BTM power model may be the only way to meet near-term compute power needs.

Catalysts

Positive: Third hyperscaler customer, Q1 2026 EBITDA beat, capacity beyond 3,100 MW, re-rating to power utility multiples, successful Hatchbo contract execution. Negative: xAI issues, environmental enforcement, AI spending slowdown, execution miss, convertible dilution event.

9 VERDICT WAIT
B T3 Adaptable
Strong Buy$30
Buy$37
Sell$80

Solaris is a genuinely innovative company at the intersection of AI infrastructure and power scarcity. Management has executed a remarkable pivot, and the pro forma economics are attractive (3-4 year paybacks, $600M+ EBITDA run rate). Customer diversification is improving (xAI down from 67% to 37% of 3.1 GW target). However, at $65.62 the stock prices in perfect execution with no margin of safety. 102x trailing PE, 25x EV/EBITDA, $1B+ debt, $903M convertible overhang, and -$438M FCF create significant downside risk. DCF range $33-52 fully diluted. Wait for $30-37 to accumulate with adequate margin of safety.

🧠 ULTRATHINK Deep Philosophical Analysis

SEI - Ultrathink Analysis

Updated: April 15, 2026

The Real Question

The real question is not whether AI data centers need more power -- they obviously do. The real question is: does owning portable gas turbines and renting them to hyperscalers create durable economic value, or is this simply a very well-timed equipment leasing business that will mean-revert once supply catches up with demand?

Every generation, there is an infrastructure buildout that creates fortunes for early movers -- railroads, telephone networks, fiber optics, cell towers. The ones that created lasting wealth (cell towers, pipelines) shared a common trait: they occupied irreplaceable positions in a network. The ones that destroyed capital (fiber optics in 2001, shale drilling in 2015) shared a different trait: the assets were commoditized and overbuilt.

The question for Solaris is which category behind-the-meter gas generation belongs in. And the honest answer, even with three more weeks of data and reflection since our March analysis, is: we still do not know. The company has been in the power business for roughly 20 months. That is not enough time to distinguish a secular shift from a cyclical peak. The stock, however, has moved from $60.45 to $65.62 in that time -- the market growing more confident while the evidence base remains unchanged.

Hidden Assumptions

The market assumes:

  1. Hyperscaler AI CapEx will continue at $500-600B+ annually for the foreseeable future. This is the single largest assumption. If AI spending disappoints -- not crashes, just disappoints -- the entire demand curve shifts.
  2. Grid buildout will remain slow enough that behind-the-meter gas is needed for 10-15+ years. But what if grid modernization accelerates? What if policy changes make utility-scale solutions more attractive?
  3. Solaris will achieve 3-4 year paybacks on $2B+ in deployed capital. This has been demonstrated at small scale with one primary customer. Replicating it across multiple sites, customers, and geographies is assumed but unproven.
  4. xAI will remain a financially healthy, contract-honoring counterparty for 15 years. xAI is private, likely unprofitable, and controlled by a mercurial founder.
  5. The $903M in convertible notes will convert at prices that do not materially impair per-share economics. At current prices, conversion adds ~12M shares (15-20% dilution). The market seems to be ignoring this.

We assume: 6. That 102x trailing PE is obviously too expensive. But what if we are wrong? What if the pro forma $600M+ EBITDA materializes by 2027 and the stock is actually trading at 7-8x forward EBITDA? The market may be smarter than we think -- pricing in what is becoming visible but not yet reported.

This is the uncomfortable truth for value investors: sometimes the expensive stock is actually the cheap one, if the growth is real and imminent. The question is whether we have the analytical conviction to distinguish genuine growth from promoted growth. And the April price action -- up 8.5% on no new information -- makes us more cautious, not less.

The Contrarian View

For the bears to be completely right, these things would need to be true:

  1. The AI buildout decelerates sharply. Not impossible. Every technology investment cycle has a trough. Cloud computing spending decelerated in 2022-2023 before re-accelerating. If AI models hit diminishing returns on compute, or if a recession forces capital discipline, the hyperscalers will defer data center projects.

  2. Behind-the-meter gas generation proves to be a temporary bridge, not a permanent fixture. The bears argue that grid operators will eventually build enough capacity, that small modular reactors will arrive, that the cost of gas generation plus emissions controls will exceed grid power. Solaris management implicitly acknowledges this by calling their solution a "bridge" -- but the stock is priced as if the bridge lasts 15+ years.

  3. The equipment leasing analogy holds. Turbines depreciate. Natural gas prices fluctuate. Operating costs escalate. In a competitive market with Halliburton and Liberty entering, returns on capital compress toward the cost of capital. What looks like a 25-33% unlevered return becomes 10-12% -- insufficient for the risk.

  4. The capital structure cracks under weight. $1.2B+ in debt post-Genco, $903M in convertibles, $935M+ in committed CapEx. If EBITDA growth stalls even temporarily, the debt service becomes crushing. The company has never been tested in a downturn with this balance sheet.

The bear case is not crazy. It is entirely plausible. And at $65.62, the stock price offers even less compensation for this possibility than it did three weeks ago at $60.45.

What Has Improved

Credit where due -- two developments since the MER acquisition meaningfully strengthen the thesis:

  1. Customer diversification. The Hatchbo contract (500+ MW, 10-year) and Genco acquisition reduce xAI concentration from ~67% to ~37% of the 3,100 MW target. This is genuine progress toward the single most important de-risking catalyst.

  2. Insider ownership. Climbing from 6.2% to 10.5% signals management conviction. They are eating their own cooking more enthusiastically.

  3. EBITDA cadence. Four consecutive quarters of EBITDA growth ($47M, $61M, $68M, $69M) with guidance for continued acceleration ($72-84M in Q1-Q2 2026). The ramp is real, not theoretical.

These improvements are meaningful. They would justify a modest increase in our entry prices -- which we have done, moving Strong Buy from $28 to $30 and Accumulate from $35 to $37. But they do not justify buying at $65.62.

Simplest Thesis

Solaris is a well-timed equipment leasing business renting gas turbines to AI hyperscalers at peak demand; improving on diversification and execution but still priced for perfection at 102x earnings with $1.2B+ in debt and $903M in convertible overhang.

Why This Opportunity Exists (As a Future Entry)

This is not a typical value investing opportunity. The stock is expensive, not cheap. The "opportunity" -- if one exists -- is a future opportunity that requires patience.

  1. Narrative stocks overshoot in both directions. SEI has risen 10x+ in 20 months on a compelling story. Stories attract momentum investors who exit violently on any disappointment. A single earnings miss, an xAI headline, or an environmental enforcement action could send the stock down 30-50% in days. That is the entry point.

  2. Execution risk is underappreciated at these multiples. Scaling from 780 MW to 3,100 MW requires deploying $2B+ in capital across multiple sites and customers in 3 years. This is an extraordinary operational challenge. Q4 2025 EBITDA growth decelerated to just 1.5% sequentially -- the first hint that scaling is not a straight line.

  3. The convertible bond overhang is real. $903M in convertible notes will eventually convert to equity, diluting existing shareholders by 15-20%. The market has not fully priced this. At $65.62, fully diluted EV/EBITDA is higher than headline numbers suggest.

  4. Cyclicality is structural. AI CapEx spending will have peaks and troughs. When the trough comes -- and it will -- power infrastructure stocks will be punished disproportionately. That is when value investors buy.

The behavioral factor creating this edge is narrative extrapolation -- the market projects the current growth rate indefinitely and pays a peak multiple for it. History teaches that growth rates mean-revert and peak multiples compress. The April price appreciation on zero new information is a textbook example.

What Would Change My Mind

I would move from WAIT to BUY if:

  1. The stock falls to $30-37 (8-11x pro forma EV/EBITDA) while the fundamental thesis remains intact.
  2. Solaris announces a third and fourth hyperscaler customer, reducing xAI concentration below 30% of the orderbook.
  3. Q1-Q2 2026 EBITDA materially beats guidance ($77M+ and $84M+ respectively), confirming the ramp trajectory.

I would move from WAIT to REJECT if:

  1. xAI shows signs of financial distress or Musk signals reduced commitment to Colossus.
  2. Net debt exceeds 4.5x EBITDA without corresponding growth.
  3. Pro forma EBITDA guidance is cut below $500M.
  4. A major environmental enforcement action forces shutdown at an operating site.

The Soul of This Business

Solaris exists because of a profound mismatch between the speed at which AI companies want compute and the speed at which the American electrical grid can deliver power. That mismatch is real, measurable, and likely to persist for years.

But the soul of this business is not power generation. It is speed. The ability to deploy 500 MW of reliable, clean-burning power to a data center in 6-12 months versus the 3-5 years required for grid interconnection. Speed is what xAI is paying for. Speed is what the second hyperscaler is paying for. Speed is the moat.

The fragility, however, is also in speed. When the grid catches up -- and it will, eventually -- the speed premium erodes. When competitors with deeper pockets (Halliburton with $20B in revenue, Williams with $80B in assets) commit to the market, the supply constraint that gives Solaris pricing power dissolves.

The soul of this business, then, is temporal: it exists because of a specific moment in history when demand for AI compute is growing faster than the infrastructure to support it. Solaris is the bridge. Bridges are valuable. But investors should not pay permanent infrastructure prices for temporary bridge assets.

The key insight is this: if Solaris can convert its first-mover advantage into switching costs, long-term contracts, and operational know-how that competitors cannot easily replicate, the bridge becomes a tollway. If it remains simply a fleet of rented turbines, the bridge becomes obsolete.

At 102x earnings and $65.62, the market is betting heavily on the tollway. At $30-37, we would be paying for the bridge and getting the tollway for free. That is the margin of safety we require. Three weeks have not changed this calculus. If anything, the additional 8.5% price appreciation on no new information has strengthened our conviction that patience will be rewarded.

SEI - Solaris Energy Infrastructure, Inc.

Executive Summary

3-Sentence Thesis

Solaris Energy Infrastructure has pivoted from oilfield logistics into behind-the-meter power generation for AI data centers, deploying natural gas turbines at customer sites under long-term contracts in what may be the most capital-intensive infrastructure buildout in modern history. The company operates ~2,200 MW of gas turbine capacity (scaling to 3,100 MW by 2029) under 7-15 year contracts with hyperscalers including xAI and a second undisclosed investment-grade tech company, generating a pro forma EBITDA run rate exceeding $600M. At $65.62, the stock now trades at 102x trailing earnings and 25x EV/EBITDA -- even more stretched than our March review -- making this a compelling business at a price that demands extraordinary patience from a value investor.

Key Metrics Dashboard (Updated April 2026)

Metric Value Assessment
Revenue (FY2025) $622M +99% YoY
Adj. EBITDA (FY2025) $244M +137% YoY
Net Income (FY2025) $30M Suppressed by interest/D&A/NCI
EPS (Diluted) $0.66 102x PE
Total Debt $1,069M D/E: 1.89x
Net Debt $715M 3.3x EBITDA
FCF (FY2025) -$438M Heavy growth CapEx
Dividend $0.48/share 0.73% yield
Pro Forma EBITDA (run-rate) >$600M When 2,200 MW deployed
Total Capacity Target 3,100 MW By 2029
Current Price $65.62 Up 8.5% since March review
52-Week Range $16.81 - $70.17 Near highs
Insider Ownership 10.5% Improved from 6.2%
Beta 1.06 Moderate

Verdict: WAIT - Excellent Business, Valuation Even More Stretched

Since our March review, SEI has appreciated another 8.5% to $65.62 without any material change in fundamentals (Q1 2026 earnings not yet reported). The trailing PE has expanded from 93x to 102x, and EV/EBITDA from 22x to 25x. The thesis remains unchanged: this is a genuinely innovative company at the intersection of AI infrastructure and power scarcity, but the current price offers zero margin of safety. We continue to wait for $30-37 to accumulate.

Phase 0: Business Understanding

What Does Solaris Do?

Solaris Energy Infrastructure provides behind-the-meter, turnkey power-as-a-service to data centers and industrial customers. Two business segments:

  1. Power Solutions (~70% of EBITDA, heading to 90%): Owns and operates natural gas turbines (primarily Solar Turbines/Caterpillar 16.5 MW and 38 MW units) co-located at customer sites. Provides "molecule to electron" service: gas procurement, generation, distribution, voltage regulation, and emissions control. Revenue is a fixed capacity payment (lease-like) with variable fuel pass-through. Contracts run 7-15 years.

  2. Logistics Solutions (~30%, declining share): Legacy oilfield business providing sand silo systems and top-fill equipment for hydraulic fracturing. Cash cow generating $80M+ annual FCF. 93 fully utilized systems with utilization in the mid-90s%.

Corporate History and Key Milestones

  • 2014: Founded as Solaris Oilfield Infrastructure
  • 2017: IPO on NYSE
  • 2018-2023: Oilfield logistics company, stock $5-15 range
  • Sept 2024: Acquired Mobile Energy Rentals (MER) for ~$200M -- gained 600 MW turbine fleet and xAI relationship
  • Sept 2024: Rebranded to Solaris Energy Infrastructure
  • 2025: Scaled from 150 MW to 780 MW operated. Formed Stateline Power JV (50.1/49.9) with xAI for 900 MW. Revenue doubled to $622M.
  • Feb 2026: Announced second hyperscaler (500+ MW, 10-year contract with Hatchbo LLC affiliate)
  • March 2026: Acquired Genco Power Solutions (400 MW, ~$620M) + 500 MW turbine slots = 3,100 MW total by 2029

Key Customers

  1. xAI (Elon Musk) -- Colossus data center, Southaven, Mississippi. 1,140 MW committed (52% of 2,200 MW orderbook, ~37% of 3,100 MW target). Stateline Power LLC JV, 15-year term.
  2. Undisclosed investment-grade global tech company (Hatchbo LLC affiliate) -- 500+ MW, 10-year contract starting Q1 2027.
  3. Various industrial -- Microgrids, gas processing, utility resiliency. Smaller but diversifying.

Aschenbrenner Position Context

Leopold Aschenbrenner's Situational Awareness LP holds SEI at 2.0% portfolio weight (~$86M). His fund is built around the thesis that AI scaling will continue and the infrastructure buildout will be enormous. SEI fits his power-for-compute thesis directly.

Phase 1: Risk Analysis (Inversion)

Risk Register

# Risk Event Probability Severity Expected Impact
1 xAI customer concentration blowup 18% -55% -9.9%
2 Regulatory/environmental shutdown of BTM gas gen 15% -50% -7.5%
3 AI CapEx winter / hyperscaler pullback 20% -35% -7.0%
4 Valuation compression on any execution miss 45% -35% -15.8%
5 Debt/leverage crisis if growth disappoints 15% -40% -6.0%
6 Competitor entry compresses returns 25% -25% -6.3%
7 Grid buildout faster than expected, BTM obsolete 12% -40% -4.8%
8 Management execution failure at scale 15% -30% -4.5%
9 Convertible dilution depresses per-share economics 60% -15% -9.0%
10 Elon Musk reputational/political risk to xAI 10% -25% -2.5%
Total Expected Downside -73.3%

Deep Risk Analysis

1. Customer Concentration (CRITICAL -- Improving) xAI concentration has improved from ~67% of the 2,200 MW orderbook to ~37% of the 3,100 MW target. The Hatchbo contract (500+ MW) and Genco acquisition meaningfully diversify. However, on current revenue, xAI likely still represents 60%+ of power segment earnings. True diversification requires executing the Hatchbo contract starting Q1 2027 and securing additional customers.

2. Valuation Risk (ELEVATED -- Worsened Since March) At $65.62, SEI trades at 102x trailing PE (up from 93x in March), 25x EV/EBITDA (up from 22x), and 10x P/S. The stock has rallied 8.5% since our March review without new earnings data. The market is pricing in not just the 2,200 MW deployment but the full 3,100 MW target with near-perfect execution. Any disappointment triggers violent repricing.

3. Regulatory Risk BTM gas generation faces environmental scrutiny. The Colossus campus generated local opposition. EPA quad K amendments provide 24-month temporary operation (up from 12), but permanent air permits remain uncertain. If a new EPA regime tightens rules, compliance costs escalate or operations face curtailment.

4. AI CapEx Cyclicality The thesis depends on hyperscalers spending $500-600B+ annually on data center infrastructure. History shows every tech investment cycle has peaks and troughs. If AI models hit diminishing returns on compute, or a recession forces capital discipline, data center projects get deferred.

5. Financial Leverage Net debt of $715M (3.3x trailing EBITDA) with $935M+ in committed CapEx. Post-Genco, total debt is approximately $1.2-1.3B. Two convertible bond issues ($155M at 4.75%, $748M at 0.25%) will dilute shareholders 15-20% on conversion.

6. Convertible Dilution (NEW EMPHASIS) $903M in convertible notes will eventually convert to equity, diluting existing shareholders by 15-20%. The forward PE of 385x (per AlphaVantage) partially reflects this dilution impact. Investors buying today are buying into a capital structure that will expand significantly.

Bear Case

AI CapEx slows 30-40% in 2027-2028. xAI restructures and renegotiates the Stateline JV. Environmental regulations tighten. Competition from Halliburton, Liberty, Williams compresses margins. Solaris left with $1.5B+ debt on assets generating $300M EBITDA instead of $600M+. Convertibles convert at depressed prices, further diluting per-share economics. Stock retreats to $12-18 (5-7x distressed EBITDA).

Phase 2: Financial Analysis

Revenue Trajectory

Year Revenue YoY Growth
2020 $103M COVID trough
2021 $159M +55%
2022 $320M +101%
2023 $293M -8%
2024 $313M +7% (MER acquired Sept)
2025 $622M +99% (first full year with power)
2026E $1.0-1.2B +60-90% (Hatchbo ramp + Genco)
2027E $1.5-1.8B +50-60% (full Hatchbo + additional)

Profitability Analysis

Metric FY2023 FY2024 FY2025 Pro Forma
Gross Margin 26.9% 25.9% 45.9% 50%+
Operating Margin 17.0% 16.9% 21.8% 30%+
EBITDA Margin 29.4% 31.9% 39.2% 45%+
Net Margin 8.3% 5.0% 4.8% 15-20%

Gross margins expanded dramatically in 2025 as the higher-margin power business scaled. EBITDA margin trajectory is strong but net income remains suppressed by interest expense (~$28M), depreciation ($84M), and non-controlling interest (JV partner share).

Quarterly Momentum (Key Indicator)

Quarter Revenue Adj. EBITDA Sequential Growth
Q1 2025 $126M $47M --
Q2 2025 $149M $61M +30% EBITDA
Q3 2025 $167M $68M +11% EBITDA
Q4 2025 $180M $69M +1.5% EBITDA
Q1 2026E -- $72-77M +4-12% EBITDA
Q2 2026E -- $76-84M +6-9% EBITDA

Q4 2025 EBITDA growth decelerated to just 1.5% sequentially. This is the first signal that the pace of MW deployment is leveling off before Hatchbo and Genco capacity comes online. Guidance for Q1-Q2 2026 shows re-acceleration, which will be the key metric to watch.

ROE / ROIC Analysis

  • ROE (TTM): 7.8% -- Poor, but distorted by heavy growth investment and NCI
  • ROIC: ~8% on trailing invested capital, distorted by growth phase
  • Target ROIC on deployed generation: 3-4 year payback implies 25-33% unlevered returns per turbine
  • Pro forma ROE (at $600M+ EBITDA): Would be 20%+ once growth phase concludes

Balance Sheet Assessment

Metric FY2023 FY2024 FY2025 Post-Genco Est.
Total Assets $468M $1,123M $2,143M ~$2.8B
Total Debt $34M $320M $1,069M ~$1.25B
Net Debt $28M $205M $715M ~$950M
D/E Ratio 0.16x 0.90x 1.89x ~2.2x
Net Debt/EBITDA 0.3x 2.0x 3.3x ~2.7x (on 2026E)

The balance sheet has transformed from conservative to leveraged in 18 months. Post-Genco, total debt is approximately $1.2-1.3B. On pro forma 2026 EBITDA of ~$350M+, net debt/EBITDA should improve to ~2.5-3.0x. Manageable but leaves no room for disappointment.

Owner Earnings Calculation

Component FY2025
Net Income $30M
+ D&A $84M
+ SBC $23M
- Maintenance CapEx (~10% of fleet value) -$45M
= Owner Earnings (growth-adjusted) ~$92M
Per share (58M basic, ~70M fully diluted) ~$1.31
P/Owner Earnings (fully diluted) 50x

DCF Valuation (Updated)

Assumptions:

  • Pro forma EBITDA at 3,100 MW: $800-900M by 2029
  • Growth rate years 1-5: 30% (from current $244M base)
  • Growth rate years 6-10: 5%
  • Terminal growth: 2%
  • WACC: 10.5% (reflecting leverage and execution risk)
  • Maintenance CapEx: 8% of revenue
  • Fully diluted share count: ~70M (post-convertible dilution)

DCF Range: $33 - $52 per share (fully diluted)

At $65.62, the stock trades 26-99% above fair value even under optimistic DCF assumptions.

Relative Valuation

Metric SEI Power Infrastructure Peers
EV/EBITDA (TTM) 25.0x 10-14x
P/E (TTM) 102x 15-25x
P/S (TTM) 10.1x 2-4x
EV/EBITDA (pro forma ~$600M) ~8x Reasonable if achieved

On pro forma numbers, SEI looks reasonable at ~8x EV/EBITDA. On trailing numbers, it is extremely expensive. The market is pricing in flawless execution of the 3,100 MW growth plan with no dilution impact.

Phase 3: Moat Analysis

Moat Assessment: NARROW (Widening)

Moat Sources:

  1. First-Mover Advantage (Moderate): Solaris was first to deploy large-scale BTM gas generation for hyperscaler data centers. 2+ years of operational track record at xAI Colossus gives them reference customers and proprietary know-how that cannot be replicated overnight.

  2. Switching Costs (Moderate-High): Once turbines are installed on a Title V air permit at a data center site, switching providers is expensive and disruptive. Equipment is deeply integrated into the site's electrical architecture. Contracts are 7-15 years with take-or-pay provisions.

  3. Supply Chain Lock-In (Moderate): Secured turbine delivery slots from Solar Turbines/Caterpillar and GE with 2-3 year lead times. OEM capacity is tight. Having equipment on order is itself a competitive advantage -- competitors cannot simply buy turbines on the spot market.

  4. Operational Know-How (Moderate): Proprietary Solaris Pulse monitoring software, in-house SCR manufacturing for emissions control, integrated molecule-to-electron capability. Managing multi-hundred-MW sites is complex and requires specialized expertise.

  5. Vertical Integration (Growing): HVMVLV acquisition (voltage distribution), in-house SCR manufacturing, gas handling, balance-of-plant engineering. Moving from "turbine rental" to "complete power plant operator" creates stickier customer relationships.

Moat Weaknesses:

  • Turbines are commoditized hardware -- anyone with capital can buy them
  • Competitors (Halliburton, Liberty, Williams) have deeper pockets and existing customer relationships
  • Technology risk: grid buildout, SMRs, or renewables+storage could erode BTM gas advantage
  • The moat is ultimately temporal -- strongest while demand massively exceeds supply

Moat Duration: 5-8 years. First-mover advantage is strongest in the current supply-constrained environment. As the market matures and more capacity comes online, structural advantages will narrow unless Solaris converts speed advantage into permanent switching costs.

Phase 4: Decision Synthesis

Management Assessment

Co-CEOs: Bill Zartler (Chairman) and Amanda Brock

  • Zartler founded Solaris in 2014, has been CEO throughout. Made the pivotal MER acquisition decision.
  • Brock joined as Co-CEO Q3 2025. Former CEO of Aris Water Solutions. Deep infrastructure expertise.
  • President: Kyle Ramachandran -- Operational leader, global generation experience.
  • New CFO: Steve Tompsett -- Capital markets specialist, focused on optimizing cost of capital.

Insider Ownership: 10.5% (improved from 6.2% in prior filing) -- Meaningful skin in the game. Capital Allocation: Aggressive growth. Maintained $0.48/share dividend through transformation. Bold M&A (MER, HVMVLV, Genco). Two convertible bond issues. Strategy is high-conviction growth.

Assessment: Management has executed exceptionally well on a bold strategy. The pivot from oilfield logistics to power infrastructure was visionary. However, the pace of capital deployment ($2B+ over 3 years) is unprecedented for a company of this size, and execution risk at scale is real.

Position Sizing

Given the risk profile (high growth, high leverage, customer concentration, stretched valuation), maximum allocation would be 2-3% of portfolio at an attractive entry price.

Entry Prices (Updated for $65.62)

Level Price Implied EV/EBITDA (Pro Forma) Gap to Current Reasoning
Strong Buy $30 ~8-9x -54% Severe market dislocation + thesis intact
Accumulate $37 ~10-11x -44% Reasonable risk/reward for growth
Fair Value $45-55 ~12-15x -16% to -31% Pro forma execution priced in
Current $65.62 ~18-19x -- Premium valuation, no margin of safety
Sell $80+ ~22x+ +22%+ Excessive optimism

What Has Changed Since March 2026

Factor March 27 April 15 Direction
Price $60.45 $65.62 Worse (more expensive)
PE ratio 93x 102x Worse
EV/EBITDA 22x 25x Worse
Market Cap $3.21B $6.25B Reflects dilution
Insider ownership 6.2% 10.5% Better
52-week low $14.09 $16.81 Updated
Fundamental thesis Strong Strong Unchanged
Q1 2026 earnings Pending Still pending No new data

Monitoring Metrics

Metric Trigger Action
Q1 2026 EBITDA Beat $77M guidance high Confirms ramp trajectory
Q1 2026 EBITDA Miss $72M guidance low Review thesis
Customer #3 announcement Contract signed Positive -- re-evaluate entry
Net Debt/EBITDA Exceeds 4.0x Concern
EBITDA margin Falls below 30% Review pricing power
xAI financial health Any distress signals Immediate review
Convertible conversion Large dilution event Update per-share math

Conclusion

Solaris Energy Infrastructure remains one of the most interesting power infrastructure stories in the market. Management has executed a remarkable pivot, the demand tailwind is enormous, and the pro forma economics are attractive. The company's narrow moat is widening through vertical integration and long-term contracts.

However, since our March review, the stock has appreciated 8.5% to $65.62 without any new earnings data, pushing the trailing PE to 102x and EV/EBITDA to 25x. The valuation leaves even less margin of safety than before. The DCF range of $33-52 per share (fully diluted) suggests the stock is meaningfully overvalued at current prices.

The key risk that has improved is customer diversification -- the Hatchbo contract and Genco acquisition reduce xAI concentration from ~67% to ~37% of the 3,100 MW target. The key risk that has worsened is valuation -- the stock continues to rally without corresponding fundamental catalysts.

For a value investor, this remains a WAIT. The business deserves ownership; the price does not yet deserve our capital.

Recommendation: WAIT -- Accumulate below $37, Strong Buy below $30

Analysis based on: SEC 10-K (FY2024), AlphaVantage financial statements and company overview, earnings call transcripts Q1-Q4 2025, investor presentations, and company press releases. Refreshed April 15, 2026 with live price data. No analyst reports used as inputs.