Strategy's latest disclosure carries a number that most shareholders will skim past: the company's modified net asset value premium has compressed materially from its 2024 peak. The headline remains that the firm holds more Bitcoin than any corporate entity on earth. The arithmetic tells a less comfortable story. A larger BTC treasury does not mechanically translate into more crypto per share, and it certainly does not guarantee positive equity returns. The variable that governs both outcomes is a single ratio — mNAV — and it has a hard threshold at 1.0.
Let me be precise about what the disclosure does and does not say. The filing confirms continued accumulation and reiterates the BTC Yield KPI that Strategy has trained the market to watch. What it does not do is reconcile that KPI against the cost of the equity and convertible instruments used to fund the accumulation. That reconciliation is the entire ballgame, and it is where the retail narrative of "more coins equals more value" detaches from the accounting reality.
The mechanical relationship is a three-factor identity that any treasury company lives or dies by:
Price = mNAV × (B / S) × PB = Bitcoin held S = shares outstanding (fully diluted) P = spot Bitcoin price mNAV = market value of equity / net asset value ```
Log returns decompose cleanly: total return ≈ Δln(mNAV) + Δln(B/S) + Δln(P). The middle term — growth in Bitcoin per share — is the only one management actually controls. The first term, mNAV, is set by the market and management cannot will it higher. The third term is beta, and everyone exposed to the asset gets it for free. Strategy's BTC Yield metric measures the middle term alone. It is a genuine operational achievement and a dangerously incomplete measure of shareholder outcome.
Here is the accretion condition, and it is not optional. Suppose a company issues ΔS new shares at market price M and deploys all proceeds into Bitcoin at price P. New Bitcoin-per-share exceeds old Bitcoin-per-share if and only if:
(B + ΔS·M/(S·P)) / (S + ΔS) > B/S
⟺ M > B·P
⟺ mNAV > 1The implication is stark. When mNAV exceeds 1, issuing equity at a premium and buying spot Bitcoin is accretive — shareholders genuinely gain coin exposure per unit of ownership. When mNAV falls below 1, the identical operation is dilutive. Dilution is not an inherent property of share issuance; it is a function of mNAV. This is the formula that matters, and it is the one the marketing rarely surfaces.
Based on my audit work on treasury-structured vehicles through 2022 and 2023, the failure mode is not the arithmetic itself — it is the asymmetry of the flywheel. Above 1.0, the mechanism self-reinforces: accretive issuance lifts Bitcoin-per-share, which feeds sentiment, which sustains the premium. Below 1.0, the same loop runs in reverse, and management loses the equity-issuance lever precisely when it needs it most. The company must then either halt accumulation or sell coin into a market that already knows why it is selling.
There is a measurement problem the disclosure glosses over, and I would flag it before anyone treats the simple ratio as gospel. mNAV has at least three defensible definitions, and they can point in opposite directions for a leveraged treasury. Simple market-cap-to-NAV ignores debt and preferred stock and systematically overstates the premium. An enterprise-value-to-NAV calculation — market cap plus net debt plus preferred — is more conservative and produces materially different readings for a firm of Strategy's capital structure. Fully diluted EV/NAV, incorporating convertible conversions, options, and remaining ATM capacity, is the closest approximation to true per-share entitlement. Strategy funds itself heavily through converts and the STRK and STRF preferred series, and the dilution from those instruments is nonlinear and path-dependent, governed by conversion prices and cumulative dividends. Any analysis that discusses only simple share-count dilution understates the real complexity by a wide margin.
This is my central contrarian point, and it cuts against both the bulls and the headline skeptics. The consensus bear case says the premium must mean-revert because it always has. That premise is asserted far more often than it is demonstrated. If the market is pricing these vehicles as regulated, tax-efficient, custody-solved Bitcoin exposure — a structural convenience rather than a cyclical mania — then some premium may be persistent. Conversely, the bulls who wave the BTC Yield KPI as proof of value creation are measuring the one term management controls while ignoring the term that determines whether the operation was worth doing. Both camps are reading a three-variable equation and arguing about one variable each.
A second blind spot: the sample. Strategy is the most mature and most leveraged structure in the category, and inferring sector-wide conclusions from its disclosure invites selection bias. The ETH treasury cohort — Bitmine, SharpLink, and their peers — follows a visibly different mNAV path, anchored to staking yield rather than pure spot beta. Their accretion math includes a yield term that the BTC treasury model lacks entirely, and their dilution instruments are structured differently. Applying Strategy's framework to them without adjustment is an error I have watched analysts make repeatedly.
The disclosure also leaves three premises unexamined, and the warning only holds if all three survive. First, that the premium mean-reverts — the strongest and least verified assumption in the entire thesis. Second, that equity issuance is the sole funding channel; cheap debt or operating cash flow would weaken the dilution argument considerably. Third, that Bitcoin-per-share is the only value metric shareholders care about — but a shareholder may rationally accept per-share coin dilution in exchange for liquidity, optionality, or tax efficiency, in which case dilution is a priced service rather than a loss.
Code does not lie, only the architecture of intent. The intent here is legible in the capital structure: the company is optimizing for accumulation, and accumulation is funded by instruments whose cost surfaces in mNAV, not in the coin count. Hedging is not fear; it is mathematical discipline, and the discipline here is to track the premium-to-NAV ratio as the master variable rather than the treasury balance as the trophy.
What I want to see in the next disclosure is the fully diluted EV/NAV reconciliation — converts, preferred, and ATM capacity mapped against net asset value on a single page. Until that appears, the correct posture is neither accumulation nor exit but measurement. Watch the 1.0 line. Above it, the machine compounds. Below it, the machine reverses, and no quantity of Bitcoin in the treasury changes the sign of the per-share result. The question for the next quarter is not how many coins Strategy holds. It is whether the market still pays a premium for the privilege of holding them through a leveraged wrapper — and that answer lives in a ratio, not a headline. For the ETH treasuries, the same question applies with a staking-yield adjustment that most models still omit.
