Strategy (MSTR): What Are You Really Buying?

A leveraged bitcoin holding company, valued honestly. What the stock is, what it is worth, and a fair hearing of the questions critics and the public keep asking.

TON618 Capital Research · Prices as of 26 Jun 2026 (MSTR $82.31) and 28 Jun 2026 (BTC ~$59,517) · Download PDF · Valuation model (XLSX)

Update, 31 Jul 2026 — Q2 2026 results: addendum — The Drain, Made Visible (scorecard; fair value restated to ≈$94).

Key questions, answered

The questions investors, critics, and the press keep asking — adjudicated as fairly as we can, before the full analysis below.

1What you are really betting on

Start with the asset, because it dictates the valuation method. Bitcoin is a commodity, and commodities have no intrinsic value — no coupon, no dividend, no cash flow to discount. A barrel of oil, an ounce of gold, a bitcoin: each is worth what the marginal buyer will pay, set by supply and demand, not by a stream of future cash flows. That is the opposite of an equity, whose intrinsic value is the present value of its cash flows. You cannot run a discounted-cash-flow model on a thing that produces no cash. The honest fair value of a bitcoin is its market price; the honest fair value of a pile of bitcoin is its net asset value (NAV) — mark-to-market, full stop.

This is why a conventional “stock valuation” of MSTR is mostly a category error, and why the bulls and bears so often talk past each other. ~95% of MSTR’s asset value is bitcoin, whose fair value is simply NAV. The only part of MSTR that can be valued like a normal equity is the small software business. So the correct framework is sum-of-the-parts: value the software on comparable-company multiples, value the bitcoin at NAV, subtract the debt and preferred that sit ahead of common, and you have the fair value of the equity. We do exactly that in §5.

What remains after that arithmetic is the real question: what are you actually betting on by owning the stock rather than the coin? Three things, in order of importance:

• Bitcoin’s direction, geared. With $22B of senior claims sitting beneath $50B of bitcoin, the common equity is a ~1.7× levered, convex claim on BTC — it outperforms in up-markets and is destroyed faster in down-markets. This is the dominant driver by far.
• The premium (mNAV). Whether the market pays above or below NAV for that levered exposure. This was the whole game in 2021–24; it is now ~1.0× and is the swing factor between “fairly valued” and “value trap.”
• The financing carry and idiosyncratic risk. A permanent ~$1.55B/yr preferred coupon (~5% of NAV/yr), refinancing risk at convert maturities, key-man/governance, custody, and tax.

2The premium that was the whole story — and its collapse

For four years, the interesting thing about MSTR was never the bitcoin; anyone could buy bitcoin. It was that the market paid a premium to NAV for MSTR’s bitcoin — peaking near 2.5× in December 2024. That premium did real work: it made above-NAV equity issuance accretive (the “BTC Yield” flywheel), and it was the single most contested number in the stock.

Daily price since June 2024 (MSTR split-adjusted, left axis; bitcoin, right axis). MSTR peaked ~11 months before bitcoin and has fallen far harder — the premium unwind, in prices.

Multiple-to-NAV (enterprise basis) over time. The MSTR-specific overvaluation that defined the stock is essentially gone.

mNAV plotted against bitcoin's drawdown from its all-time high. Richest near the top; gone by a deep drawdown.

Price sources: bitcoin — CoinMetrics (daily, keyless); MSTR — Alpha Vantage, Schwab fallback. Drawdowns are actual; mNAV anchors are illustrative from reported figures.

This second view answers a question worth being honest about: how much of the premium was ever real, versus a bull-market mood? The answer is “mostly mood, plus some structural decay.” The premium peaked at ~2.5× with bitcoin at its highs, then bled lower through 2025 even while bitcoin was strong — because the spot-ETF launches quietly killed the “only way to own bitcoin in a brokerage account” rationale — and the 52% bitcoin drawdown into 2026 finished the job. A premium that evaporates in a drawdown was never balance-sheet value; it was sentiment. That is the single most important thing to internalize before valuing the stock.

The Louis Vuitton problem. Paying 2.5× NAV for MSTR was like paying full retail for a Louis Vuitton bag when you could buy the identical bag direct from the contract manufacturer. The underlying “leather” — bitcoin — is the same; the markup is brand, access, and convenience (Saylor’s conviction, index inclusion, a stock your 401(k) can hold). That markup was defensible when bitcoin was hard to buy. But the spot-bitcoin ETFs (IBIT and peers, ~0.25% fees) opened the factory door: anyone can now buy the bag direct, cheaply. With the premium back to ~1.0×, the LV markup has largely evaporated — you are now buying close to “at cost,” which is the bull’s consolation and the bear’s vindication at once.

Jim Chanos built the canonical trade around exactly this: long bitcoin, short MSTR, harvesting the spread as the premium normalized. He entered around 2.5× and covered around 1.15× in November 2025 — the thesis worked, and he said so. The premium has since drifted to ~1.0×. The relevant point for a buyer today is that this trade is largely done: the easy premium-compression has happened. What is left is the levered bitcoin exposure and the structural carry — which is a very different proposition from the one the bears were short.

3The fortress: capital structure designed to survive a siege

Whatever one thinks of the strategy, the balance sheet is engineered with real skill — a fortress built to hold out as long as possible. The design goal is explicit: never be a forced seller. Three features deliver that.

• Cheap, unsecured, long-dated debt. ~$6.7B of convertible notes at 0–2.25% coupons, no bitcoin pledged as collateral, laddered 2027–2032. No margin, no maintenance test — a falling BTC price triggers nothing.
• Perpetual preferred. ~$15.5B across five series (STRK 8%, STRF 10%, STRD 10%, STRE 10%, STRC 11.5%) that never mature and whose dividends can be deferred in extremis. Permanent capital that cannot be called.
• An equity ATM that can issue into strength. Together these mean the company can refinance and wait, rather than liquidate, through a drawdown.

All of MSTR’s capital obligations on one scale. The perpetual preferred (gold) is by far the largest and never matures; the convertible maturities (teal) are the smaller, dated cash walls.

But a fortress under siege still has a garrison to feed. The preferred’s strength — permanence — is also its cost: a ~$1.55B/yr cash dividend, forever, against an asset that yields nothing. That bill has grown from ~$0.3B/yr in early 2025 toward ~$1.5–1.6B as the preferred stack scaled. It is funded only two ways: issue more securities (which needs market access and, ideally, a premium) or sell bitcoin. In June 2026 the company crossed that Rubicon for the first time, selling 32 BTC to fund a distribution — small and symbolic, but a tell. The fortress is sound; the question is the cost of the siege over time.

4Business quality: moat, earnings quality, and governance

A valuation needs a view on business quality — the things the balance sheet doesn’t show. For MSTR three matter: whether the model has any durable edge, whether the reported earnings mean anything, and who controls the capital-allocation lever.

Moat — a financing edge, not a structural one

The original moat was access: for years MSTR was the easiest way to own bitcoin inside a brokerage account or a mandate that banned spot crypto. The spot-bitcoin ETFs ended that in 2024. What remains is real but narrower: scale and cost of capital — MSTR runs the largest, most-liquid bitcoin balance sheet and the deepest access to the convertible and preferred markets (no peer has issued ~$15B of bitcoin-backed preferred), wrapped in index membership, a deep options/ETF ecosystem, and the Saylor brand. The catch: every one of those advantages is contingent on the premium. The cost-of-capital edge exists only while markets fund it at or above NAV; at a discount it inverts. So MSTR’s moat is a pro-cyclical financing moat — it widens in bull markets and evaporates in bear markets — not a structural one. That is a materially lower-quality moat than a network effect or switching costs.

Earnings quality — ignore GAAP; watch coverage

Fair-value accounting (ASU 2023-08) runs unrealized bitcoin marks through the income statement, so GAAP net income now swings violently with the bitcoin price — roughly +$10B in a strong 2025 quarter, ~−$12.5B in Q1 2026. GAAP EPS is noise: it measures the quarter’s bitcoin move, not operating performance, and MSTR should never be valued on an earnings multiple. The operating business throws off little cash (software ~breakeven, revenue ~$450M and declining), so there is essentially no free cash flow to cover the ~$1.5–1.6B/yr preferred dividend — it is met by issuance or bitcoin sales (the first 32-BTC sale came in June 2026). And the headline KPI, “BTC Yield,” is a self-defined non-GAAP metric that can flatter value-destructive issuance (NYDIG’s critique). The figures that actually carry information are BTC-per-share, NAV, and preferred coverage — not earnings.

Governance — concentrated, founder-controlled

The binding letter in “ESG” here is G. A dual-class structure gives Michael Saylor super-voting Class B control, so the strategy is inseparable from one person’s conviction — acute key-man risk. Capital allocation is aggressive and almost entirely discretionary (issue equity at will, scale the preferred, now sell bitcoin); common holders are along for the ride. The preferred carries its own governance backstop — deferring its dividends can hand preferred holders board representation, which protects them but signals stress and ranks ahead of common. Net: a founder-controlled vehicle whose governance quality rests on continued trust in a single capital allocator.

5The real risk factors

Stripping away both the hype and the “Ponzi” noise, here is what actually threatens permanent capital loss for the common, roughly in order of how live each is.

1. Bitcoin drawdown (the master risk). Everything keys off one variable. At ~1.7× leverage, a 50% BTC fall is a far larger fall in MSTR’s equity value, and the convexity turns vicious near the breakeven.
2. Discount-to-NAV / reflexivity reversal. Below 1.0× mNAV, the accretion machine inverts: issuance shrinks BTC-per-share, which justifies a deeper discount — a slow self-reinforcing spiral that destroys equity value without any default.
3. The preferred funding drain. ~$1.55B/yr of cash dividends with no operating cash flow to pay them. In a prolonged bear with markets shut, this forces dilutive issuance or bitcoin sales — the opposite of the accumulation thesis.
4. Refinancing walls. Convertibles must be repaid in cash at maturity if out-of-the-money (they all are). The walls are laddered and modest (~$2B in 2027–28 vs. $50B of BTC), so manageable unless capital markets are closed at the wrong moment.
5. Index exclusion. A future MSCI/S&P decision to treat bitcoin-treasury companies as ineligible would force billions in passive selling (JPMorgan: $2.8–8.8B). MSCI deferred in Jan 2026 — the risk is postponed, not gone.
6. Tax (CAMT). The Corporate Alternative Minimum Tax can levy cash tax on unrealized bitcoin gains under fair-value accounting — a bizarre risk that could force BTC sales in an up-year. Policy-dependent and bidirectional.
7. Governance / key-man. Dual-class super-voting stock concentrates control in Michael Saylor; the strategy is inseparable from one person’s conviction.
8. Custody & regulatory. Single-asset custody risk, and the contingent — though currently remote — ‘40 Act reclassification risk (only if bitcoin were deemed a security).

6Valuation — sum-of-the-parts, then the bet

We value MSTR the only way that respects what it is: software on comparable multiples, bitcoin at NAV, less the claims that rank ahead of common.

Bitcoin at mark-to-market NAV; software at comps; less convertible debt and preferred. Equity NAV vs. current market cap.

At ~$86/share fair value versus ~$82 traded, MSTR is roughly fairly valued to NAV — arguably marginally cheap. This is the single most important shift in the story: the large idiosyncratic overvaluation is gone. Note the SOTP confirms the “not really a business” point — the software contributes ~$1.5B of a ~$53B asset base; this is a bitcoin balance sheet with a software footnote.

What you actually own: a levered, convex BTC claim

Fair value per share as bitcoin moves, vs. holding bitcoin directly. Above today, leverage amplifies upside; below, it underperforms and the equity is wiped near $23k.

Because $22B of senior claims sit beneath the bitcoin, the common behaves like a call option on bitcoin struck around $26k: convex and geared to the upside, with total loss in a deep enough decline. That convexity — not a premium — is now the core of the bull case. It also means any single “price target” is really a bet on bitcoin; the table below makes the conditionality explicit.

Per-share figures hold share count constant and ignore further accretion/dilution; they are directional, not point forecasts. Probabilities are deliberately omitted because the expected value is almost entirely a function of one’s bitcoin view — which is the honest summary of the whole security.

Relative value — the whole treasury cohort has de-rated

MSTR is no longer the premium outlier. As the model was copied across ~150+ listed crypto-treasury companies, the premium compressed sector-wide: by mid-2026 roughly one in three of them trade below 1.0× mNAV, and MSTR — once ~2.5× — now sits around ~1.0×, mid-pack rather than richest. On relative value the “pay up for treasury access” trade is broadly over; MSTR screens roughly in line with peers and with the cost of simply owning bitcoin. Its remaining differentiator is structural — scale, financing depth, and no-margin-call leverage — not a valuation discount.

MSTR vs. a levered bitcoin ETF at the same risk

If the goal is geared bitcoin exposure, MSTR is one option among several now. Held at risk parity against, say, a 2× bitcoin ETF, the trade-offs are:

• MSTR advantages: term leverage with no margin calls and no daily-reset volatility decay. A 2× daily ETF (e.g., MSTU/MSTX-style products) bleeds value in choppy markets through rebalancing; MSTR’s leverage is long-dated and refinanced, not rebalanced, so it has no path-decay and cannot be liquidated on a wick.
• MSTR disadvantages: a ~5%/yr preferred carry ($1.55B on ~$31B NAV), tracking error to bitcoin via a premium/discount that an ETF does not have (MSTR can de-rate to a discount; an ETF tracks NAV by construction), plus idiosyncratic key-man, refinancing, index, and tax risk.

The clean way to say it: MSTR is crash-resistant leverage with a carry and a wildcard premium; a levered ETF is clean BTC beta with decay and a margin/borrow cost. Which is “better” depends on horizon (MSTR’s no-decay edge compounds over long holds) and on your read of the premium (now near 1.0×, the historical tailwind/headwind is roughly neutral).

7The polarizing part: vision, ambassadorship, and the critics

Few stocks split serious people as cleanly as MSTR, because two true things sit uncomfortably together: the strategy is intellectually coherent and structurally fragile. A fair note gives both sides their best voice.

The bull case & Saylor as bitcoin’s ambassador

Michael Saylor is, by common consent on both sides, the most effective corporate evangelist bitcoin has had — free education (Saylor Academy), an open-sourced “treasury company” playbook now copied worldwide, and a vocabulary that gave holders conviction through 50%+ drawdowns. The underlying thesis is serious: cash is a melting asset, bitcoin is durable “digital capital,” and intelligent, non-callable leverage lets a public company accumulate it across decades.

“Bitcoin is a swarm of cyber hornets serving the goddess of wisdom, feeding on the fire of truth, exponentially growing ever smarter, faster, and stronger behind a wall of encrypted energy.”Michael Saylor, 2020

“#Bitcoin is the Apex Property of the Human Race.” · “There is no second best.”Michael Saylor, 2024 (keynote theme)

“I’m going to be buying the top forever. Bitcoin is the exit strategy.”Michael Saylor, Bloomberg, Feb 2024

“If you’re a short seller and your thesis is the company’s got to sell equity in order to fund the dividends, I would like nothing better than to… rip your wings off.”Michael Saylor, Q1 2026 earnings call

A fair-disclosure footnote to the bull case: the “never sell” promise is no longer literally true. After a ~$12.5B Q1 2026 loss, Saylor said Strategy would “probably sell some bitcoin” to fund dividends and executed a small (32 BTC) sale in June 2026, which he framed as a tactical brushback at short-sellers. The thesis is intact; the absolutism is not.

Tom Lee’s Exxon precedent. Fundstrat’s Tom Lee argues there is S&P precedent for valuing a company on a commodity it holds — Exxon was a top company for a generation, valued on its oil. Examined, the analogy actually cuts toward NAV, not premium: Exxon is an operating business that finds, extracts, refines, and sells oil — converting a commodity into cash flow. MSTR does none of that; it holds. The honest comparable is a closed-end commodity fund, which trades at NAV (often a discount), not a premium — which is exactly where MSTR has landed.

Wall Street’s structured-product validation. Bulls note that Goldman, Morgan Stanley, and JPMorgan now issue bitcoin-linked structured notes, and that MSTR single-stock ETFs (MSTY ~$4B, MSTU ~$1.6B, MSTX ~$1.2B) have multiplied — evidence that institutions are happy to build levered/income products on bitcoin and on MSTR. True, and a real demand signal. The double edge: it also confirms MSTR’s “unique access” moat is gone (you can get geared bitcoin a dozen ways now), and the reflexive ETF flows add volatility.

The bear case, in the critics’ own words

“We’re selling MicroStrategy stock and buying bitcoin… basically buying something for $1, selling it for $2.50.”Jim Chanos, May 2025 (trade closed ~1.15× in Nov 2025)

“None of the reasons cited for the stock’s relative attractiveness justify paying well over double for the same coin.”Kerrisdale Capital short report, 2024

“If you buy $100 of Bitcoin and put it in a pot, you can slice the pot into shares and sell them for $200… a perpetual motion machine.”Matt Levine, Bloomberg (mechanism, not fraud)

“A company that believes it is issuing equity at a premium when it is not can easily end up making value-destructive financing decisions.”NYDIG Research, on the “BTC Yield” metric, 2025

The bears’ strongest point is not “fraud” — it is that the accretion engine is premium-dependent and reverses below NAV, and that “BTC Yield” is a self-defined, methodology-sensitive metric that can flatter value-destructive issuance. That critique has largely played out in the de-rating to ~1.0×. The sharpest forward-looking worry is the preferred drain meeting a sustained discount — the reflexive reversal — not a sudden liquidation. (Note on attribution: the widely-shared “leveraged Ponzi” line often pinned to David Gerard was in fact a commenter’s; Chanos’s own word was “gibberish,” not “Ponzi.”)

8The MSCI episode — and the “incumbent attack” question

No controversy around MSTR generates more heat than the late-2025 push to throw bitcoin-treasury companies out of the major equity indices, which much of the bitcoin community read as a coordinated, existential attack by Wall Street incumbents on MSTR, bitcoin, and the crypto industry. Because that belief is widespread, it deserves to be documented and adjudicated on the evidence rather than waved away.

What actually happened

• 10 Oct 2025 — the proposal. MSCI opened a consultation to exclude “Digital Asset Treasury Companies” (DATCOs) — firms with ≥50% of total assets in digital assets — from its Global Investable Market Indexes. The preliminary list named 39 companies (~$113B float-adjusted cap; 18 existing constituents to be removed, 21 blocked from entry), MSTR among them. It landed the same day as the largest crypto liquidation on record (~$19B) — which was itself triggered by a White House post announcing 100% tariffs on Chinese imports, a documented macro cause. The shared date became Exhibit A for the conspiracy reading.
• 20 Nov 2025 — the JPMorgan note. JPMorgan (Panigirtzoglou) estimated exclusion could force ~$2.8B of passive selling from MSCI alone, up to ~$8.8B if other providers followed; ~$9B of MSTR’s then-$59B cap sat in passive funds. By 3 Dec JPMorgan added the risk was “already priced in.”
• 10 Dec 2025 — Strategy fights back. Saylor and President Phong Le sent a 12-page letter calling the 50% threshold “discriminatory” for singling out digital assets while leaving oil-, real-estate-, gold-, and bond-heavy firms untouched: “Strategy is not a fund, not a trust, and not a holding company. We’re a publicly traded operating company with a $500 million software business and a unique treasury strategy.”
• 6 Jan 2026 — MSCI declines. MSCI determined not to exclude DATCOs in its February 2026 review, deferring to a broader future consultation on non-operating companies generally. MSTR jumped ~6–7%. The feared $2.8–8.8B of forced selling never occurred.

The mundane explanation (which is also the strongest one)

Strip away intent and a decades-old, asset-neutral rule remains: broad equity indices track operating companies and exclude investment vehicles. In its own words, MSCI’s indices “seek to measure the performance of operating companies and exclude entities whose primary activities are investment-oriented,” and some DATCOs “exhibit characteristics similar to investment funds, which are not eligible for inclusion.” S&P’s methodology already excludes closed-end funds, BDCs, and royalty/investment trusts — categories a 90%-bitcoin balance sheet plainly resembles. As OMFIF put it, inclusion “could compel passive investment funds to acquire billions of dollars in bitcoin exposure regardless of their investment mandates” — unintended crypto beta forced on investors who never asked for it. That is precisely the closed-end-fund framing this note uses to value MSTR; the index question is the same point in governance form.

The attack narrative, scored against the evidence

The honest verdict: suspicious timing is real; proven coordination is absent. There are no leaked communications, no filings, no regulatory findings — and where the loudest specific claims could be checked against primary documents, they failed. Antitrust law has a name for the rest: conscious parallelism — competitors independently making similar decisions is legal and expected, especially when each has an obvious independent commercial reason (an index tightening its operating-company rule; banks selling their own bitcoin wrappers). The strongest credible critique — that MSCI singled out digital assets while ignoring >50% cash/gold/real-estate firms — is an allegation of single-provider bias, not multi-party collusion, and it is a fair one.

Loose ends still open (as of this refresh)

• MSCI’s deferral is not a resolution. Its promised broader consultation on “non-operating companies” appears not yet opened as of late June 2026; an interim freeze on DATCO weight/share increases remains, and the question can resurface at any review.
• S&P 500: eligible on paper, still excluded. After fair-value accounting let MSTR post a large Q2 2025 GAAP profit, it met the index’s earnings test — yet the S&P committee declined to add it (Sept 2025), at its discretion. There is no S&P DAT-exclusion policy; the obstacle is committee judgment.
• Nasdaq-100: retained, but one reclassification from trouble. MSTR kept its seat in the Dec 2025 reconstitution. The live tail risk is an ICB sector reclassification from “technology” to “financials” — the Nasdaq-100 excludes financials — which would cost eligibility (~$1.6B of forced selling, per Jefferies). The early-2026 review window passed with no confirmed change, but the overhang stands.
• FTSE Russell is accommodative — no DAT-exclusion rule, and it added ETH- and SOL-treasury names in its June 2026 reconstitution.

Net read. The acute index threat has been deferred, not killed — worth monitoring, but a routine (if contestable) index-governance question that got mythologized, not a proven coordinated assault. The most important tell: MSTR’s troubles since the MSCI relief have been overwhelmingly fundamental — the premium collapse and the preferred-dividend drain documented above — not index-driven. The genuinely existential variables for MSTR remain bitcoin’s price and its own reflexive funding model, not Wall Street’s index committees.

9Conclusion — what we think it is worth, and the open questions

Fair value ~$86/share at spot bitcoin, versus ~$82 traded — roughly fairly valued, arguably marginally cheap to NAV. The MSTR-specific overvaluation that defined the stock for four years has corrected: the premium is back to ~1.0×, the trade the bears made is largely complete, and what remains is a clean-ish, ~1.7× levered, convex claim on bitcoin carrying a permanent ~$1.55B/yr preferred coupon and refinancing/reflexivity risk.

What you are really buying: not “bitcoin at a discount” and not “free alpha,” but geared bitcoin without margin-call risk or daily-reset decay, minus a carry, minus the optionality of the premium (now neutral), plus idiosyncratic risk. Over a long horizon and a constructive bitcoin view, the no-liquidation leverage is genuinely valuable and the stock should outperform spot BTC. Over a flat-to-down bitcoin path, the preferred drain and a potential slide to a discount make it a poor substitute for simply holding the coin (or a spot ETF). It is, in the cleanest terms, a leveraged bet on bitcoin’s direction with extra ways to be right and extra ways to be wrong.

Is it a fraud, a Ponzi, about to be liquidated? No, no, and no — on the evidence. It is a real, fully-disclosed, over-collateralized balance sheet with a fortress liability structure. But it is also not a perpetual motion machine, and the parts of the bear case that were structurally true (premium-dependence, the preferred drain, the “not really a business” point) are real and now visibly in play.

Catalysts & open questions to watch

• Bitcoin’s price & halving-cycle position — the master variable; the geared, convex payoff means everything keys off it.
• 2027–2028 convertible maturities / investor puts — the first hard cash walls (~$2B); the refinancing terms struck in the prevailing tape will be the tell on market access.
• Quarterly prints — preferred-dividend coverage, the ATM issuance pace, and any further bitcoin sales beyond the first 32-BTC sale.
• Does mNAV stabilize at ~1.0×, or break into a persistent discount that inverts the model?
• How is the ~$1.55B/yr preferred coupon funded through a bear — issuance, or accelerating bitcoin sales (the June 2026 sale being the first)?
• The deferred MSCI/S&P index question — exclusion would force passive selling on a vulnerable float.
• CAMT cash-tax treatment of unrealized gains — still policy-dependent.
• Succession/governance: how much of the premium (and the conviction bid) is Saylor-specific?