Every thesis on this site so far has been bullish, which is embarrassing for a bear market survivor, so here is the other kind. By March 2025 I was telling anyone who would listen that the most crowded trade in crypto — MicroStrategy, by then renamed Strategy, the corporate bitcoin-treasury vehicle trading at a fat premium to the coins it held — was going down. Not because of anything in its filings. Because of megawatts.
The idea, which I started calling the time-to-power multiple, goes like this. The AI buildout is bottlenecked not by chips but by energized grid interconnections — the lead time to get new power hooked up has stretched from months to years. Companies already sitting on high-watt connections are therefore being bum-rushed by AI companies who want their capacity, and one cohort holds a disproportionate amount of it: bitcoin miners. That single fact, I thought, was quietly catastrophic for bitcoin itself. First, a miner that pivots to AI has to fund the datacenter conversion, and its treasury is bitcoin — structural sell pressure from the very companies that used to be structural buyers. Second, every megawatt of hashrate reallocated to inference is hashrate that no longer secures the network, concentrating what remains and eroding the decentralization the whole story rests on. Third, somewhere out past the horizon, quantum computers break the cryptography outright. And fourth, nearest to the ground: after the 2024 election everyone and their mother had already bought — the president was personally championing the coin, and the tape was priced for applause. Overbought, plus three slow leaks.
If bitcoin was headed down, MSTR was the instrument to express it: a company whose assets were bitcoin, levered with convertible debt, trading well above the value of its own coins. Leverage on the way up is leverage on the way down. The thesis said: short the proxy, not the coin.
What I never did was open a position. The thesis lived in conversations and valuation notes — March 2025 texts about shorting the thing, per-watt math on the miners — and stayed there. Which is exactly what makes it testable now, honestly: no cherry-picked entries to defend, no exits to rationalize. Just the question that actually matters: suppose you had the bear thesis in March 2025 and expressed it the obvious retail way, with puts. What happens to you?
PART I — THE TAPEWhat MSTR actually did
Here is the punchline the hero chart already gave away: the thesis was right. From $250.92 on March 3, 2025, Strategy finished at $100.77 on July 2, 2026 — down 60%, with a low print of $82.31 a week before this article. Bitcoin itself (measured by the IBIT ETF, which is what a retail account actually sees) fell about 29% over the same window, so the leverage worked exactly as advertised: MSTR fell twice as hard as the coin, and the premium-to-NAV compressed on top.
But look at the shape of the year, because the shape is the whole article. The first thing the stock did was go up 82%. From March through mid-July 2025 MSTR ran from $251 to $455.90 — bitcoin was rallying toward its October peak, and the proxy front-ran it with full leverage. The top in MSTR came on July 16, 2025, nearly three months before bitcoin’s own top in early October: the premium started deflating while the coin was still climbing, which is what it looks like when the marginal buyer of a levered proxy is exhausted. Then fourteen months of stairs down: $322 by September, $177 by November, $152 at New Year’s, $82 by late June 2026.
So a bear who started buying puts on March 1, 2025 spent four and a half months being wrong by half an account before being right by a career. Whether you survived the being-wrong part — and what the being-right part cost to hold — is not a detail. On a name this volatile, it’s the entire game.
PART II — THE MACHINEEight traders, one brain
Same setup as article No. 2, pointed the other direction. The first article on this site reconstructed my real 2021 options ledger — 378 round trips — into a behavioral fingerprint: the patterns that made money (Monday-midday entries, near-the-money strikes around 0.45 delta, multi-day holds harvested into strength) and the patterns that lost it (Tuesday-morning chases, strikes one rung too far out with three days to live, averaging down, riding carcasses to expiry). In 2021 those habits traded calls. Here, each is transposed to the put side and pointed at MSTR.
Every strategy starts with $20,000, sizes each ticket at the 2021 median (10% of equity in premium), buys at the ask, sells at the bid, and pays 2021 TD Direct commissions: $9.99 plus $1.25 a contract, each way. The quotes are the real NBBO, pulled hourly for 2,500+ MSTR put contracts across all 74 expiries from March 2025 to July 2026. Rules are mechanical; nothing sees the future inside the run. Three benchmarks trade no personality at all: a plain 100-share short, a monthly at-the-money put roll, and a monthly put spread — buy the ATM put, sell the 15%-out put against it — the textbook way to cheapen a bearish bet on an expensive-vol name.
| Strategy | Enters | Exits |
|---|---|---|
| Best self the 2021 winning patterns, mirrored to puts | Monday midday, only into strength (≥2% above the prior 5-day closing low — sell rallies, don’t chase dumps); puts at ~0.45 |delta|, 10–24 days out | Harvest into weakness at 1.75× cost or once deep in the money (|delta| ≥ 0.92); 2-session time-stop if never green; out the day before expiry; never averages down |
| As I traded the modal 2021 behavior, frequencies from the ledger | ~3 entries/week, tilted Thu–Fri; median ~3 days to expiry; at-the-money to 12% out, per the 2021 moneyness ladder | Morning-heavy; quick harvests near 1.5×; quick punts when red; 11% of lots ride to expiry; occasional averaging down. Run 51 times; median shown |
| Worst self the 2021 losing patterns, mirrored to puts | Tuesday 10:30 a.m., chasing any −1.5% prior day — buying puts after the down move; ~10% out of the money, ≤3 days to expiry | Averages down once when 40% underwater; sells only on capitulation (bid ≤ 30% of cost) or rides to expiry |
| Best in, worst out | Best self’s entries | Worst self’s exits |
| Worst in, best out | Worst self’s entries | Best self’s exits |
| Short the stock | Short 100 shares on day one | Never — unless the margin desk decides otherwise (30% maintenance, forced cover on breach) |
| Monthly ATM put the simplest options expression | First trading day of each month: one at-the-money put ~30–60 days out, 10% of equity | Sold at month-end; rolled |
| Monthly put spread the cheapened expression | First trading day of each month: buy the ATM put, sell the −15% put, same expiry, 10% of equity on the net debit | Both legs closed at month-end; rolled |
PART III — THE RACESixteen months, eight accounts
Same thesis. Same tape. Same starting stake. Same commissions. The only variable is behavior — and this time the tape spends its first four months punishing everyone.
EIGHT WAYS TO TRADE A THESIS THAT WAS RIGHT
account equity, daily, log scale · marked to the bid · click legend to toggle · hover to inspect
FIG. 1 — The race.
PART IV — THE LEDGEREvery strategy, by the numbers
FIG. 2 — Final, maximum and minimum values are daily closes, marked at the bid. Fees are commissions only, at 2021 TD Direct rates. “As I traded” shows the median of 51 randomized runs.
PART V — THE PRICE OF BEARISHNESSWhat puts on a 70-vol name cost
Every strategy above had to pay the same toll at the door, so let’s price the door. MSTR options in this window traded at a median 68% implied volatility — several times what a boring large-cap costs — because the market (correctly) expected violence in both directions. Here is what one month of bearish conviction cost, sampled on the first trading day of every month:
ONE MONTH OF CONVICTION, PRICED MONTHLY
~30–60 DTE ATM put premium as % of the stock price · solid = naked put, hatched = ATM/−15% spread · hover to inspect
THE VOL YOU PAID vs THE VOL YOU GOT
implied vol at entry vs realized vol over the option’s life · hover to inspect
FIG. 3 & 4 — The toll.
PART VI½ — THE BLOTTEREvery trade, named
Every position, named: exact put contract (expiry and strike), entry, exit, and the rule that closed it. The strikes range from at-the-money down to 20% out depending on the strategy — filter to compare how each one walked the chain.
TRADE BLOTTER — exact contract, entry, exit
PART VI — ENTRY vs EXITWhich half of the behavior mattered
Article No. 2 found that on a runaway bull tape, exits explained most of the gap between my best and worst selves. The crossbreeds ask the same question on a bear tape — and get a different answer.
THE FOUR CROSSES: FINAL VALUE OF $20,000
same tape, entries × exits · log scale
PART VII — THE HONEST PARTWhat this does and doesn’t prove
Every backtest is a flattering mirror. This one has all of article No. 2’s distortions plus two of its own:
There is no trade ticket. The thesis dates to March 2025, but talk is not a timestamped position, and a skeptic is entitled to discount any backtest whose start date the author chose himself. The defense is narrow but real: every strategy here is committed before seeing the option data, runs mechanically from day one — including the four and a half months where the trade gets destroyed — and the interesting results are mostly negative anyway. It’s hard to accuse a backtest of hindsight flattery when its headline is that the thesis-haver usually loses.
The strategies were built in hindsight, from a different market. The rules are distilled from my real 2021 call-buying ledger and mirrored to the put side by hand. “Best self” is best-in-2021, transposed — not best-possible, and not necessarily good here.
One stock, one regime, one of the great round trips of the cycle. MSTR went +82% then −78%. That specific shape — rip first, collapse later — is maximally hostile to early bears and maximally kind to late ones, and every conclusion below is conditional on it. On the tape where MSTR rips to $600 and stays, the disciplined strategies bleed out slower than the reckless ones, and that’s the whole difference.
The fills are hypothetical and polite. Hourly NBBO snapshots; buy at ask, sell at bid; no slippage beyond the spread; no assignment; positions up to roughly a hundred contracts assumed to fill without moving quotes. MSTR weeklies were liquid, but not infinitely so. The short-stock line ignores borrow fees and assumes a clean forced cover at the close on the day maintenance breaks — a real margin desk would have been messier and probably earlier.
The randomized trader is a distribution, not a line. “As I traded” makes random choices with the 2021 ledger’s frequencies; I ran 51 seeds and show the median, with the 5th–95th percentile range in the table. The others are deterministic.
No taxes, no FX, no interest on cash, no early assignment. Same as always. And the 2021 account’s punitive currency-conversion fees are left out of both sides of every comparison.