Research Library
The charts and tables that carry the argument — drawn from 21squared's research briefings and presentations. Every exhibit is sourced and dated. No hype; just the data that made us build a firm around one asset.
All content on this page is for informational and educational purposes only and does not constitute investment advice, an offer, or a solicitation. Sources and as-of dates are stated per exhibit. Past performance is not indicative of future results.
I — The Macro Case
The case for a hard, non-sovereign asset starts with the arithmetic of the system it hedges.
The debt has never shrunk year-over-year in this window. Every policy path — growth, inflation, or repression — debases the unit it's denominated in.
Interest is the fastest-growing federal budget line. Rates can't rise without breaking the budget; inflation becomes the release valve.
The red line is the benchmark that matters. Anything that can't outrun debasement is a slow guaranteed loss — the question is only which hard asset does it best.
II — The Asset
Four numbers from BlackRock's own research desk, then the two charts that reframe Bitcoin's risk.
The reflex sell-off is liquidity, not verdict. Over the window that matters, the market has repriced Bitcoin upward after every shock since 2020.
Held for a day, Bitcoin is a coin flip. Past three years, the loss probability collapses below 1% — and no ten-year window in the sample has ever lost. Time horizon is the risk-management tool.
III — The Portfolio Math
Fidelity's ten-year study of a 60/40 portfolio with a Bitcoin sleeve, rebalanced annually — the single most useful table in institutional Bitcoin research.
| BTC alloc. | Ann. return | Ann. volatility | Sharpe | Sortino | Max drawdown |
|---|---|---|---|---|---|
| 0% | 9.44% | 10.26% | 0.72 | 1.09 | −20.64% |
| 1% | 11.25% | 10.65% | 0.85 | 1.34 | −21.02% |
| 3% ★ | 14.56% | 12.04% | 1.01 | 1.76 | −21.79% |
| 5% | 17.55% | 13.80% | 1.09 | 2.09 | −23.21% |
| 7% | 20.30% | 15.65% | 1.12 | 2.34 | −24.64% |
| 10% | 24.09% | 18.41% | 1.15 | 2.62 | −26.72% |
3% of the portfolio lifted returns by +5.1 points per year while the worst drawdown deepened by barely one point. ★ Sharpe and Sortino improved most in the 1% → 3% step — Fidelity's own headline observation — and kept rising through 10% (Sharpe 1.15).
| Institution | Recommended BTC allocation | Methodology | Published |
|---|---|---|---|
| BlackRock | 1–2% | Risk-contribution budgeting — reaffirmed to advisors Jun 2026; IBIT added to BlackRock's own model portfolios | Dec 2024 · Jun 2026 |
| Bank of America | 1–4% | CIO guidance; 15,000+ Merrill & Private Bank advisers may proactively recommend BTC ETFs | Jan 2026 |
| Morgan Stanley | 0–4% | Equal-weight model | Q1 2026 |
| Bitwise | 5% | Mean-variance optimization | 2025 |
| Fidelity | 9.4% (max Sharpe) | Kelly criterion + mean-variance | Mar 2026 |
1–3% is mainstream institutional consensus. Below 1% is too small to matter; the mathematical optimum sits far above what most committees can stomach. The honest counterweight: distribution opened all year — Vanguard's platform (Dec 2025), BofA's advisers (Jan 2026), BlackRock's models (Jun 2026) — yet US spot-ETF flows still ran ≈ −$5.8B net year-to-date through July 13. Access is normalizing; the flow confirmation hasn't arrived.
IV — The Cycle
Our own analysis — Bitcoin against its two most reliable long-term floors. Not a prediction; a map of where price sits relative to its history.
Closes below the 200-week MA have occurred only at the 2015, 2018, 2020 and 2022 cycle lows — every one of them was, in hindsight, an accumulation window.
Loading the pipeline read…
An outside check on the same question. Glassnode — an institutional on-chain data provider with no connection to us — publishes its own market composite built from a largely different set of inputs, weighted toward macro and capital-flow measures we deliberately exclude. On 7 July 2026 it read 22 out of 100, its second-lowest band. Our pipeline read 18.4–18.9 a week later. Two independently constructed indices, built by different people from different data, landing within about four points. That is evidence about the level, not proof that either set of weights is right — and their index subsequently fell back a band for roughly three weeks before recovering, which is a fair reminder that agreement at a point says nothing about the path.
| Check | Reading | What it says |
|---|---|---|
| Drawdown | −48.6% · 308 days | Off the October 2025 all-time high — the shallowest major Bitcoin drawdown on record. Prior cycle bottoms printed −77% to −85%. |
| Galaxy bear-bottom scorecard | 4 of 13 triggered | Only price-relative indicators (Mayer, 200-week MA, Fear & Greed, hash ribbons) have fired. Every valuation and capitulation metric — MVRV family, NUPL, Reserve Risk, Puell — has not reached historical bottom zones. |
| The reclaim level | ~$67,400 | Short-term holder cost basis, computed in-house from the full UTXO distribution. Bitcoin has traded below it for over ten months. Note the level falls as recent buyers' cost basis resets — the bar comes down to meet price as much as price rises to meet it. |
| Miner production cost | ~$75,000 | Spot trades below the estimated marginal cost of production — historically an accumulation-zone condition, not a timing signal. |
| The largest corporate buyer now sells | Two-way since Jul 2026 | Strategy Inc. formally added bitcoin sales to its capital policy — "sell BTC when advantageous to the Company" — and has executed three, including one at a $203m realised loss to fund preferred dividends. Roughly 76% of its stack is underwater. The marginal corporate bid can no longer be assumed one-directional. |
| Self-custody carries a real tail | ~1,400–1,800 BTC | A five-year seed-generation flaw in a leading hardware wallet was exploited in July 2026; forensic estimates put confirmed losses at 1,367–1,816 BTC across more than 4,500 addresses, and roughly $2.3bn of coin moved as owners migrated. Custody architecture is a live risk to be designed around, not a solved problem. |
| The debasement trade, so far | Metals are winning | Since January 2025: silver +107%, copper +66%, gold +60% — against Bitcoin −35%. If the thesis is monetary debasement, the market has been expressing it through metals. This is the single most uncomfortable number on this page, and it is why Box 1 holds a deliberate gold sub-sleeve. |
Our composite says deep value; the capitulation checklist says the classic bottom signature has not printed. Both can be true — which is why the mandate sizes positions by rules instead of calling bottoms.
V — The Income Layer
Bitcoin digital credit — exchange-listed preferred instruments whose coupons are backed by large Bitcoin treasuries. The asset class our mandate is designed to use as its income base.
| Instrument | Coupon | Issuer & backing | Structure notes |
|---|---|---|---|
| STRC | 12.00% | Strategy Inc. — 842,138 BTC held (~4% of supply), the largest corporate Bitcoin treasury. $10.5bn notional outstanding; 71% retail-held. | Perpetual preferred, monthly dividends; variable rate managed toward par; deepest liquidity of any preferred globally. |
| SATA | 13.00% | Strive Inc. — ~20,000 BTC, fully debt-free balance sheet. | Perpetual preferred; ~1.58× asset coverage of senior claims; ~10-year dividend reserve policy. |
Double-digit coupons are the compensation for issuer and structure risk — these are credit instruments, not savings accounts. STRC traded down to $71.63 in the June stress episode and has since recovered ~30% to $94.60 — still below its ~$99–100 par objective — while SATA closed at $99.47, about 1% from par; the asset class traded $13.0B par-normalized in June alone: a real secondary market, with real drawdowns. Two caveats belong with that recovery: the issuer has been funding preferred dividends and buybacks partly by selling bitcoin, and at a flat bitcoin price Strategy's own disclosure puts STRC's probability of falling below 1.0× coverage over its duration at ~25%. Position sizing and entry price do the risk management.
Charts don't manage a sleeve — sizing, entries, and the patience to hold do. That's the mandate we're building.
Start the Conversation Back to Main Site