Crypto & Digital Assets
A levered, perpetual-dividend-funded bitcoin holding company whose entire accretion engine — issue stock above NAV, buy BTC — has inverted to a ~0.85x discount, so it is now SELLING bitcoin to pay an ~$0.9B/yr preferred coupon; bullish only as a bitcoin call, bearish as a structure, and the discount is the tell.
Research
The verdict
A levered, perpetual-dividend-funded bitcoin holding company whose entire accretion engine — issue stock above NAV, buy BTC — has inverted to a ~0.85x discount, so it is now SELLING bitcoin to pay an ~$0.9B/yr preferred coupon; bullish only as a bitcoin call, bearish as a structure, and the discount is the tell.
Strategy Inc (Nasdaq: MSTR; renamed from MicroStrategy in 2025) is the largest corporate holder of bitcoin in the world. Two businesses sit inside one ticker:
The bitcoin treasury (the real company). As of Feb 13, 2026 it held ~717,131 BTC acquired for an aggregate $54.5B at an average ~$76,027/BTC. By the Q1 print it held 762,099 BTC at March 31, 2026 (carrying value ~$51B) and 818,334 BTC as a subsequent-event count, and 845,256 BTC by June 7, 2026. The bitcoin is described by management as "the economic backing for our equity and fixed income securities, which enables our capital markets strategy" — i.e. the BTC is collateral for the financing machine.
The legacy software business (the shell). Enterprise business-intelligence software (the old MicroStrategy product). FY2025 total revenue $477.2M, +3.0% YoY — and shrinking in mix toward subscription: subscription services $175.7M (+64.5%), product support $204.2M (−16.2%), product licenses $39.7M (−18.3%). FY2025 software gross profit was $327.8M but the segment posts an operating loss before the bitcoin line once the ~$5.4B unrealized BTC mark is set aside. Management states plainly: "We do not expect cash and cash equivalents generated by our software operations to be sufficient to satisfy our short-term or long-term liquidity needs". The software is a rounding error against a >$50B bitcoin position; it exists mostly to keep the entity classifiable as an "operating company" (see Lens 8 — index inclusion).
Contract structure / payment terms: On the software side, recurring ratable subscription + annual support (remaining performance obligation $588.0M, ~$344.2M due within 12 months ). On the capital side, the binding obligations are perpetual preferred dividends (STRF 10% fixed-cumulative, STRK 8%, STRD 10% non-cumulative, STRC variable — raised to 11.50% Feb 28 2026 ) plus convertible-note coupons. These are the "customers" that matter — the security holders the company must keep paying.
Key people: Michael Saylor — Executive Chairman & Chairman of the Board, controls a majority of Class B stock (10 votes/share) and thus a controlling share of total voting power. Phong Le — President & CEO. Auditor: KPMG LLP (McLean, VA). Incorporated in Delaware; HQ Tysons Corner, VA. Credit rating: S&P B- (Oct 27 2025, affirmed Dec 2025) — deep speculative grade.
A bitcoin-treasury vehicle has no physical supply chain; its "supply chain" is the capital-formation pipeline and the custody stack. Naming every stakeholder:
Upstream — capital sources (inputs):
Midstream — the company: Strategy issues securities → receives USD → routes through trade-execution partners affiliated with its custodians → buys BTC close-in-time to the raise to minimize market risk.
Downstream — custody (the asset sits here): Three U.S. custodians hold substantially all the BTC, deliberately split to diversify counterparty risk:
Chokepoints / single-source dependencies:
The bull's moat thesis rests on four pillars; only two are durable.
First-mover scale & index/liquidity status (durable-ish). MSTR is the original and by far the largest bitcoin treasury company, with the deepest equity liquidity, the most developed capital-markets apparatus, and an entire family of bespoke "digital credit" preferred instruments no peer has replicated at scale. Saylor invented the category; the brand is "leveraged bitcoin equity." This is a real, if narrow, moat — distribution and capital-markets sophistication.
Capital-structure engineering (genuine, but double-edged). The STRF/STRC/STRK/STRD/STRE stack is genuinely clever: it segments fixed-income demand by yield/duration/risk and raises non-dilutive capital (preferred issuance does not increase Assumed Diluted Shares). In 2025, ~$5.69B of the $25.15B raised was non-dilutive preferred. That is a real engineering edge over a naive "just sell common" vehicle.
The premium-to-NAV "moat" (NOT durable — and currently inverted). For years the core advantage was reflexive: a premium let MSTR mint accretive BTC-per-share by issuing stock above NAV, which justified the premium, etc. This is not a moat — it is a market mood, and it has reversed. mNAV went 3–4x (2024 bull) → ~2.5x (Dec 2024) → ~1.16x (spring 2026) → ~0.82–0.86x (Apr–Jun 2026). A "moat" that can flip to a discount is a beta, not a moat.
There is no operating moat. The BI software is a declining, sub-scale asset in a market owned by Power BI / Tableau / Snowflake-adjacent tooling; it is not a competitive weapon, it is a legacy annuity in slow runoff.
Bargaining power: Strategy has little over its capital providers. Preferred holders demanded — and got — a STRC rate hike to 11.50% as BTC weakened; the cost of capital rises exactly when the company can least afford it. Over bitcoin, it has zero pricing power (price-taker). Over software customers, modest pricing power inside a shrinking base.
Strategy does not report bitcoin as a revenue "segment" — the BTC P&L flows through the unrealized gain/loss on digital assets line under ASU 2023-08, not revenue. So segment economics split cleanly:
Software revenue by line (FY2025 vs FY2024):
| Line | FY2025 | FY2024 | YoY |
|---|---|---|---|
| Product licenses | $39.7M | $48.6M | −18.3% |
| Subscription services | $175.7M | $106.8M | +64.5% |
| Product support | $204.2M | $243.8M | −16.2% |
| Other services | $57.7M | $64.3M | −10.3% |
| Total revenue | $477.2M | $463.5M | +3.0% |
The story is the cloud transition: subscription up sharply, license + support in structural decline as customers migrate off perpetual licenses. Net software revenue is roughly flat and will likely stay flat-to-down. Geographically, the US dominates (Q1 2026: US $69.3M of $124.3M total), with material foreign operations (foreign pretax income was positive $52.4M in 2025 vs a $5.58B US pretax loss — because the US entity bears the BTC mark).
The bitcoin "segment" KPIs (the metrics that actually matter):
| KPI | FY2025 | FY2024 | Change |
|---|---|---|---|
| BTC Yield (BTC-per-share growth) | 22.8% | 74.3% | −51.5 pts |
| BTC Gain (BTC) | 101,873 | 140,631 | −27.6% |
| BTC $ Gain | $8.92B | $13.13B | −32.1% |
| BTC acquired | 225,030 | 258,320 | — |
| Year-end holdings | 672,500 | 447,470 | — |
The trend is sharply decelerating. BTC Yield — the company's own headline accretion metric — fell from 74.3% to 22.8% in FY2025, and to just 3.2% in Q1 2026 (vs 11%+ in Q1 2025). Cause per management: fewer BTC bought, a worse dilutive/non-dilutive mix, and rising equity issuance used to pay dividends and fund the USD Reserve rather than buy bitcoin. This is the accretion engine losing power in real time.
The headline is a catastrophe by GAAP and a non-event operationally — both true at once.
Q1 2026 income statement:
What it says: The GAAP loss is entirely a bitcoin price move (ASU 2023-08 fair-value accounting), not operational deterioration — BTC fell intra-quarter, the carry value dropped to ~$51B. The genuinely important operating fact buried in the print: preferred dividends are now a $229.5M/quarter cash obligation — ~$0.9B annualized — against a software business that generates well under that in operating cash. That gap is the company's defining stress, and it is widening as the preferred stack grows.
Balance-sheet flags: Cash $2.25B (all designated USD Reserve); BTC unencumbered; convertible debt ~$8.2B; deferred tax liability on unrealized BTC gain ~$2.42B at YE2025 — note this reverses into a deferred tax asset + valuation allowance now that BTC sits below cost basis, a future non-cash earnings risk management flagged itself.
Market reaction / what's priced in: The market has de-rated the premium entirely — MSTR ~$112.5, mkt cap ~$39.4B (June 18 2026) trading at ~0.85x bitcoin NAV. The stock is no longer rewarded for GAAP losses driven by BTC; it is being punished for the structure (leverage + perpetual coupons + dilution) on top of the BTC drawdown.
No transcripts in the research layer (transcripts=0) — sourced from filings tone + public reporting.
The narrative arc across the last ~4 quarters has shifted from offense to defense:
What they stopped saying: The triumphant "accretion" framing is muted; the new vocabulary is "asset coverage," "USD Reserve," "balance-sheet resilience," "stress scenarios." Management is now narrating downside protection, not upside compounding — the clearest tell in the whole file.
Strategy is sui generis; the right comp set is "digital-asset treasury (DAT) vehicles," not software or even miners. Multiples below are `` where sourced and n/a otherwise (never fabricated). The relevant valuation metric is mNAV (price ÷ crypto-NAV), not P/E (the company is structurally loss-making under fair-value accounting).
| Company | Ticker | Treasury asset | mNAV / key metric | Note |
|---|---|---|---|---|
| Strategy | MSTR | Bitcoin (845k BTC) | ~0.82–0.86x | Mkt cap ~$39.4B; the category benchmark, now at a discount |
| SharpLink Gaming | SBET | Ethereum | n/a | ETH-treasury vehicle |
| BitMine Immersion | BMNR | Ethereum | n/a | ETH-treasury vehicle |
| DeFi Development Corp | DFDV | Solana | n/a | SOL-treasury vehicle |
| American Bitcoin | ABTC | Bitcoin | n/a | BTC treasury/mining hybrid |
| Coinbase Global | COIN | (exchange, not treasury) | n/a | Operating comp, different model |
| Riot Platforms | RIOT | (BTC miner) | n/a | Miner, not pure treasury |
| CleanSpark | CLSK | (BTC miner) | n/a | Miner |
P/E, EV/Sales, EV/EBIT, dividend yield, 5-yr ROE — n/a — not meaningful for MSTR: the company is intentionally loss-making under ASU 2023-08, pays no common dividend, and its "earnings" are a bitcoin-price derivative. The single comp number that matters is mNAV ~0.85x. The structural read: when the category leader — the most liquid, most sophisticated, lowest-cost-of-capital DAT — trades below NAV, the smaller, higher-cost ETH/SOL copycats are almost certainly in worse shape, because their accretion math broke first and harder. Strategy is the canary and the floor for the entire DAT trade.
The pattern is unambiguous: MSTR is a leveraged bitcoin call with two idiosyncratic overlays — the premium/discount cycle and index-inclusion mechanics.
Michael Saylor — Executive Chairman, founder-archetype, the entire thesis. Track record: built MicroStrategy into a real BI company in the 1990s; survived a 2000 SEC accounting-restatement settlement (a known historical red flag); then in 2020 executed the most aggressive corporate-treasury bet in market history, converting a stagnant software firm into the world's largest corporate bitcoin holder. Quantified: turned a sub-$1B-cap software company into a vehicle that peaked above $100B in market value and amassed 845k+ BTC. He is a true believer and a once-in-a-generation capital-markets innovator — the STRF/STRC/STRK/STRD architecture is genuinely novel financial engineering.
Verdict on management: Visionary, aligned on bitcoin, structurally unaccountable to minority holders, and running a strategy that works beautifully on the way up and is untested-to-dangerous on the way down. You are not buying a management team that will pivot — you are buying Saylor's conviction that bitcoin goes up enough to outrun the cost of the leverage.
As a forensic analyst, the accounting is — surprisingly — clean; the risk is structural, not fraudulent.
Regulatory findings (required):
Conventional EPS projection is not meaningful for Strategy — GAAP "EPS" is a leveraged function of the bitcoin price (Q1 2026 LPS $(38.25) was 100% a BTC mark). Projecting it = projecting BTC, which I will not fabricate. Instead, the analytically honest forward frame is NAV-per-share and coverage under bitcoin scenarios, all `` with arithmetic shown.
Inputs: ~845,256 BTC (June 7 2026); ~$8.2B convertibles; ~$0.9B/yr preferred dividends; $2.25B USD Reserve; ~334M diluted common shares (rising); spot BTC ~$64,227 (June 21 2026); spot MSTR ~$112.5 (~0.85x mNAV).
Bitcoin NAV scenarios (gross BTC value, before debt/preferred claims):
| BTC price | Gross BTC value (845k BTC) | Note |
|---|---|---|
| $50,000 (bear) | ~$42.3B | Below aggregate cost basis ($54.5B for 717k); deep underwater, BTC sales accelerate to fund coupons |
| $64,227 (spot) | ~$54.3B | ≈ aggregate cost on the larger stack; current state |
| $90,000 (base) | ~$76.1B | Back above cost; premium may re-emerge if accretion narrative revives |
| $150,000 (bull) | ~$126.8B | The Saylor case; premium likely returns, flywheel restarts |
The decisive forward variable is not EPS — it is whether the mNAV discount persists.
No forecast.ts forecast logged — per --watchlist rules (and there is no honest standalone EPS line to commit to; the scoreable variable is BTC price, which is exogenous). A trackable proposition for later: "MSTR mNAV ≥ 1.0x by 2026-12-31" — currently I'd put p ≈ 0.35.
Bull case. Strategy is the only liquid, institutional-grade way to hold leveraged bitcoin inside an equity wrapper, with a capital-markets machine no one else has built. If bitcoin resumes its secular climb (halving cycle, sovereign/institutional adoption, debasement trade), MSTR's ~1.5–2x beta plus a returning premium produces explosive upside — the Street's $320+ targets (high $570) price exactly this. The preferred stack is non-dilutive leverage; the $2.25B USD Reserve buys ~2.5 years of runway; 23 consecutive preferred payments show the machine works. Saylor's conviction and control mean no wavering. In a bull tape, you want the most-levered, most-liquid bitcoin vehicle — and that is MSTR.
Bear case (2–3 permanent-impairment risks).
Pre-mortem (18 months out, thesis broke): Bitcoin chopped sideways-to-down through 2026–2027 in a higher-for-longer rate regime; the mNAV discount never closed; STRC's rate ratcheted higher to defend par, raising the coupon burden; the USD Reserve drained and BTC sales became routine; a convertible put date (Sept 2027 / 2028) forced either cash repayment or heavy dilution; MSTR converged toward NAV and the premium-era holders were left with a levered, decaying bitcoin position trading below the bitcoin it owns.
Are multiples too high? The stock is at a discount (0.85x NAV), so on a pure-asset basis it is "cheap." But that discount is rational: it prices the leverage, the perpetual coupon drag, the dilution, and the broken accretion engine. The asset is cheap; the structure deserves the discount.
Contrarian view (what the market refuses to see): Both sides are anchored wrong. The sell-side ($320+ targets, Strong Buy) is anchored to the old premium-flywheel model that no longer functions at a discount — those targets implicitly require the premium to return, which they don't justify. The bears calling "death spiral" underweight that the USD Reserve + 2.5-year runway + Saylor's control mean the unwind, if it comes, is slow and orderly, not a 2022-style implosion. The truth the tape is quietly telling you: MSTR has already transitioned from a premium compounding vehicle to a discounted, leveraged closed-end bitcoin fund — and it should now be valued like one (NAV ± a leverage-adjusted discount), not like a growth equity.
Dismantling the bull case.
A coal-plant bitcoin miner that re-priced itself into a $14B "AI landlord" by renting power to a venture-stage neocloud — the equity is a levered call on Fluidstack actually paying, with Google's $3.2B backstop the only thing standing between the multiple and zero.
A profitable, fast-growing digital bank mispriced as a crypto name and now mispriced again by fear — but the entire bull/bear hinges on one binary question the market cannot yet resolve: are the +$2.0B of fair-value loan marks real, or is Muddy Waters right?
A levered, sub-NAV Ethereum proxy run by the best operators in the trade — own the discount-closing buyback, not the "treasury premium" that already died; the bet is ETH plus a 0.83x→1.0x mNAV mean-reversion, fighting a 95% drawdown and a structurally bigger rival.