How to Value a Bitcoin Treasury Company

01

The Wrong Maps

Every analyst who encounters a BTCTC reaches for the nearest familiar framework. Each one fails. Not slightly off, but structurally wrong.

Framework

Finance / NAV
P/Book · Closed-End Fund · NAV discount

Measures the chest. Tells you what you're paying for BTC held today. Completely blind to the engine filling the chest. A static fund at 1x NAV and an active BTCTC at 1x mNAV look identical. They are not.

Ignores the engine

Framework

Tech / Growth
P/E · EV/EBITDA · Price/Sales

Measures the engine but assumes an operating business underneath: revenue, customers, margins, competitive moats. BTCTCs have none of that. There is no EBITDA. The framework has no slot for BTC/share growth as earnings.

No slot for accumulation

Framework

Mining / Royalty
Reserve NAV · Stream growth

Closest in spirit. But royalty streams are contractual and counterparty-dependent. BTCTC accumulation is capital markets-dependent and self-directed. The analogy breaks on first contact with a convertible offering.

Wrong risk architecture

The reason every framework fails is the same: none were built for an entity whose only product is BTC/share growth, whose only operating model is capital markets arbitrage, and whose only relevant asset is provably finite. That combination has never existed before.

02

What BTCTCs Actually Are

Three structural characteristics make Bitcoin Treasury Companies a new asset class with no prior equivalent. They share one characteristic with finance companies, one with tech, and one with neither.

I.

The treasure is provably finite

BTC supply is capped at 21 million. Accumulation isn't extracting from an unlimited resource. It's capturing a larger slice of a fixed pie. This makes BTC/share growth fundamentally different from earnings growth. The underlying asset cannot be diluted by its issuer. No equity, bond, or commodity shares this property.

II.

The "earnings" are ownership slices, not cash flows

There is no revenue. No customer. No product margin. The earnings are the BTC/share growth: the increasing claim each share has on the treasury. Structurally closer to a land trust than a business. Value accrues silently, not through distributions, but through the appreciation of a fixed underlying claim.

III.

The operating model is pure capital markets arbitrage

The engine works when cost of capital is below the BTC appreciation rate. That is the entire business model. No supply chain. No R&D. No talent retention risk. No competitive moat to defend. The risk profile is simpler and more legible than any operating company, which is exactly why standard operating company metrics produce nonsense when applied here.

03

Defining the Earnings Metric

For any P/E comparison to be honest, the earnings proxy must be accurate and dilution-complete. BPS yield is not. CEBE sats yield is.

Definition

CEBE Sats Yield

CEBE Sats Yield = (CEBE_end ÷ CEBE_start) ^ (4 ÷ quarters) − 1

The annualized rate at which BTC/share, net of all senior claims, grew over a measured period. Denominated in satoshis. No BTC price effect. No USD translation. Pure engine output.

What it includesDebt · Preferred · Converts · Warrants · Dilution

What it excludesBTC price movement (sats are sats regardless of price)

Why not BPS yieldBPS ignores senior claims. Phantom growth is real and frequent.

Why not USD yieldUSD yield conflates engine output with BTC price. Strips out what you're actually measuring.

Empirical validation CEBE directional accuracy: 62% · BPS: 43.5% · p<0.01 (22Q, McNemar's)

BPS yield cannot serve as the earnings proxy because it ignores the cost of every instrument used to acquire BTC. A quarter showing BPS growth can simultaneously show CEBE contraction: phantom growth. Over 22 quarters of Strategy data, this happened 9 times. CEBE accounts for the true cost of every instrument. That integrity is what makes the P/E comparison honest rather than flattering.

04

CEBE mNAV as the Valuation Anchor

Plain mNAV tells you what you're paying for the chest. CEBE sats yield tells you how fast the engine is filling it. CEBE mNAV combines them into the only question that fully values a BTCTC.

The CEBE mNAV Framework

CEBE mNAV = mNAV ÷ CEBE Sats Yield

To show the arithmetic: if the engine grows sats/share at 40.6% annually and you paid 2x NAV to own it, you paid 2 divided by 0.406 -- 4.9x the annual output. Same logic as paying $49 for a stock that earns $10 per share. That's your multiple on the machine.

mNAV What are you paying for the BTC held today?

CEBE Sats Yield How efficiently is the engine accumulating, net of all instrument costs?

CEBE mNAV implied P/E What multiple are you paying on the accumulation engine?

mNAV is not inherently cheap or expensive. It is a direct multiplier on your cost of the engine. A company at 3x mNAV with a 40% CEBE yield is cheaper than a company at 1x mNAV with a 5% yield. You need both numbers. Neither alone is sufficient.

05

The Math: Real Data

Strategy (MSTR) is the only BTCTC with sufficient history to produce a defensible steady-state CEBE yield. Six verified quarters. Full capital structure complexity. All numbers sourced from SEC filings.

Strategy (MSTR): Verified Engine Output Q3 2024 → Q1 2026 · 6 quarters · SEC filings

Inputs

CEBE start (Q3 2024)91,593 sats/share

CEBE end (Q1 2026)152,637 sats/share

Quarters measured6

Capital structure includedConverts · STRK · STRF · STRD · STRC · Cash offset

Engine Output

(152,637 ÷ 91,593) ^ (4 ÷ 6) − 1 = +40.6% annualized

Annualized CEBE Sats Yield +40.6%

Sats-denominated. If BTC doubles and CEBE in sats stays flat, dollar yield looks great but the engine produced nothing. Sats yield shows zero. That's the honest number.

This is a trailing yield across six quarters of verified accumulation history -- not a forward projection. The sats yield is unaffected by BTC price movement: the engine kept accumulating from 91,593 to 152,637 sats even as BTC fell from $82K to $68K between Q1 2025 and Q1 2026. That separation is precisely why sats yield is the right metric. USD-denominated yield conflates engine performance with BTC price and produces misleading signals in both directions.

06

The Implied P/E: Across the mNAV Spectrum

Using Strategy's verified 40.6% annualized CEBE sats yield as the earnings proxy. The only variable is the mNAV you pay at entry.

Entry mNAV	CEBE Sats Yield	Implied P/E	vs Nvidia (~22-35x)
1.0x	40.6%	2.5x	−93%
1.5x	40.6%	3.7x	−89%
2.0x	40.6%	4.9x	−86%
2.5x	40.6%	6.2x	−82%
3.0x	40.6%	7.4x	−79%
4.0x	40.6%	9.9x	−72%
5.0x	40.6%	12.3x	−65%
Nvidia (reference)	--	~22-35x	baseline

Source: MSTR CEBE computed from SEC filings Q3 2024–Q1 2026. Sats yield annualized over 6 quarters. mNAV is the entry multiple at time of investment. Nvidia forward P/E shown as approximate range (22-35x) as of Q1 2026 depending on estimate vintage. Not a buy/sell recommendation.

↗ This table is a standalone tweet panel. Share on X →

The Forward P/E Distinction

Nvidia's 22-35x is a forward P/E -- projected future earnings built from revenue models, margin assumptions, and competitive positioning that can and do miss. The CEBE sats yield in this table is a trailing yield, not a forward projection. It represents what the engine produced during six quarters of verified accumulation history. This is not a P/E in the traditional sense because BTCTCs don't generate cash flow -- they generate sats accumulation. It's the structurally equivalent metric for an asset class that operates on a different earnings model. The capital structure is public. The BTC/share trajectory is mechanically traceable. The primary unknown is whether BTC appreciates faster than the weighted cost of capital -- one variable instead of fifty.

07

What Each Framework Can and Cannot See

Framework	Values the chest?	Values the engine?	Full instrument cost?	Equity-comparable?
Finance / P/NAV	YES	NO	NO	NO
Tech / P/E	NO	PARTIAL	NO	YES
BPS Yield	YES	PARTIAL	NO	NO
mNAV alone	YES	NO	NO	NO
CEBE mNAV ↗	YES	YES	YES	YES

08

When the Engine Works, and When It Doesn't

CEBE sats yield is not static. It is a direct function of the spread between BTC appreciation and the weighted cost of all capital instruments. CEBE mNAV surfaces that spread in real time.

Positive Spread

BTC > Cost

BTC appreciation exceeds weighted cost of preferreds, converts, and dilution. CEBE sats yield is positive. The engine creates value. mNAV is a meaningful multiple on a productive asset.

Negative Spread

BTC < Cost

Preferred yields and dilution outrun BTC appreciation. CEBE sats yield turns negative. Drag expands. The implied P/E on the engine becomes undefined. You are paying for an engine generating negative returns.

The Complete Valuation Statement

Full BTCTC fair value = f(BTC price, mNAV, CEBE sats yield, capital efficiency spread).

The capital efficiency spread is BTC appreciation rate minus weighted instrument cost. When the spread compresses (as drag compression research documents in detail), CEBE yield falls, the implied P/E rises, and the valuation case weakens regardless of BTC price direction. CEBE mNAV is the only metric that makes this condition observable and comparable to equities.

09

Data Quality Standards

CEBE mNAV implied P/E is only as reliable as the yield history behind it. Not all companies have enough data to produce a meaningful steady-state yield figure.

Anchor-Ready

6+ positive CEBE quarters
No negative CEBE interruptions
Established capital structure

Currently: MSTR only. Metaplanet has 6 quarters but carries startup-phase distortion from a near-zero base. As the tracker matures and more companies build history, this list will grow.

Not Yet Comparable

Fewer than 6 valid quarters
Near-zero starting CEBE
Volatile or incomplete periods

Currently: all other tracked companies. Their CEBE and drag figures are real and useful. Their annualized yields are mathematically correct but not yet stable enough to anchor a P/E argument.

A P/E ratio on two quarters of earnings history is not meaningful for any asset class. The same standard applies here. CEBE mNAV implied P/E requires enough history to distinguish the engine's structural output from startup-phase noise. Send Bobby a verified data set and we will add it.

10

The Positioning Statement

The Question the Industry Has Not Answered

"How do you value a Bitcoin Treasury Company?"

Not like a bank: they don't yield cash.
Not like a tech company: they don't have customers.
Not like a fund: their engine is actively managed, not static.

They are a new asset class: treasury accumulation vehicles.

The correct valuation anchor is CEBE mNAV: the implied multiple you're paying on the accumulation engine, measured with full instrument integrity across debt, preferred, converts, and dilution.

cebetracker.io defined the standard for computing it.