Bitcoin is down 14% since Strategy's peak. Strategy is down 69%.

Introduction

Strategy — the company formerly called MicroStrategy — is a software business that spent six years turning itself into a way to own bitcoin through the stock market. So the fair test of the shares is not whether they went up, but whether they tracked the thing they hold. Over the past two years they have not: bitcoin is down about a seventh from the day MSTR peaked, and MSTR is down more than two thirds.

Start here: what you are actually looking at

Strategy is a bitcoin holding company with a software business attached. Its revenue — growing 7.8% a year — is a rounding error next to its bitcoin position, which was carried at $73.2bn at the last balance sheet date we could pull. That single fact invalidates most of the ratios you would normally reach for, and it is why this post leans on the price comparison instead.

Its reported profit is mostly an accounting mark. Under the rules in force since 2025, companies carry bitcoin at fair value and run the change through the income statement each quarter. So “earnings” for Strategy is largely a restatement of what bitcoin did. In the first half of 2026 that produced a $22.77bn unrealized loss and a $20.76bn net loss. Nothing was sold.

A moving average is the average closing price over a stretch of days, recalculated daily — the 50-day summarises two months, the 200-day a year. Support and resistance are prices where a stock has repeatedly stopped falling or rising; they are conventions, not laws. RSI compresses fourteen days of gains and losses into one number from 0 to 100, where below 30 is called oversold and above 70 overbought.

Realized volatility is how much a stock has actually been moving, annualized. MSTR’s is 84.1%. For scale, we measured Meta at 39.6% last week. That number does more work in this post than any level.

The chart that matters: MSTR against bitcoin

Read the gap between the lines rather than either line. Measured from three different starting points, MSTR amplifies bitcoin every time — and the multiple is not stable:

• From MSTR’s Nov 20, 2024 peak: bitcoin −14.1%, MSTR −69.2%. Roughly 4.9 times the fall.
• From the joint Oct 27, 2025 peak: bitcoin −29.0%, MSTR −50.6%. About 1.8 times.
• From the June 26, 2026 low: bitcoin +35.1%, MSTR +77.4%. About 2.2 times.

We cross-checked bitcoin against IBIT, the exchange-traded fund that simply holds it and which our own data feed carries: −29.6% and +35.8% over the same two spans, within half a point of the coin. The amplification is in MSTR, not in the measurement.

A leveraged way to own an asset is not a better way to own it. It is the same bet with the losses enlarged, and over this two-year stretch the enlargement is the entire story.

What was reported, and what arrives on October 29

Strategy reports third-quarter results on October 29, after the close, with consensus earnings of $8.34 a share. That number is mostly a bitcoin mark, and the direction is already knowable: bitcoin ended June at $58,559 and September at $83,554, a rise of 42.7%. After an $8.3bn unrealized loss in the second quarter — the difference between the $14.46bn loss reported through March and the $22.77bn through June — a large accounting gain in the third is close to arithmetic.

Which is why the coverage going into the print is about a possible windfall rather than about the business. Treat that as a reason to be careful: a gain everyone can calculate in advance is not news, and a stock that has already risen 77% off its low has arguably priced it.

Three other things were reported in the days before we pulled this data, and they matter more than the mark.

The company has been buying its own shares instead of bitcoin. Coverage on October 20 reported Strategy spending roughly three times more on stock buybacks than on bitcoin over five weeks, and a preferred-stock buyback raised to $3bn. For a company whose stated purpose is accumulating bitcoin, choosing its own equity instead is a statement about which it considers cheaper.

An index rule change hurt. MSTR was reported to have lost about 10% in a week on an MSCI rule affecting how companies like it are classified — a reminder that a holding company’s price depends partly on who is permitted to own it.

Analysts disagree violently. In the same three days, one house was quoted seeing 53% upside, another calling the stock 13% overvalued, and a prominent bitcoin figure calling it “massively undervalued”. All of these are reported positions, not our view, and we have not verified the underlying work.

What the trend looks like

The trend function calls this sideways, and that is fair for the last two months while being a poor summary of the last six. MSTR fell from $177.17 on May 1 to $96.85 on August 6 — down 45% — then rose 67% to $162.20 by September 23, and has drifted back to $146.01.

The useful observation is the dashed line. The 200-day average has fallen from $209.54 in April to $152.86 now. The ceiling is descending toward the price rather than the price climbing to the ceiling, which is what the aftermath of a large drawdown looks like.

The levels, and why they are nearly useless here

Support is 3.8% below and resistance 4.7% above: a band 8.5% wide. Now put that next to the volatility. MSTR’s average true range is $9.79, or 6.70% of its price — meaning the stock travels the entire support-to-resistance band in about one and a quarter average sessions.

That is the honest verdict on technical levels for this name. They are real, they are correctly calculated, and they are inside the noise. On a stock moving 6.7% a day, “resistance at $152.86” is a statement about Tuesday, not about a trend. Compare Meta, where the equivalent band was wide relative to a 4.06% daily range and the levels carried some information.

What to watch next

We do not forecast prices, and none of the below is a prediction. These are the published numbers that would resolve the questions above.

October 29, after the close. The quarter itself is largely knowable in advance from bitcoin’s move. What is not knowable is what the company says about buybacks versus bitcoin purchases, and about how it intends to fund further accumulation. That is the part worth reading.

$152.86 — the 200-day average. MSTR has been below it since February. A sustained reclaim would be the first in eight months; losing $140.49, the 50-day, would put the stock back in the range it spent the summer in.

The gap to bitcoin. The clearest single measure of whether this stock is working is whether it stops falling faster than the coin. Track MSTR against IBIT over any window you like; no technical level tells you as much.

84.1% volatility. Whatever happens on the 29th, the move will likely be large. A stock with a 6.7% average day does not react to earnings in single digits.

Why the usual valuation ratios do not work

It is worth seeing how thoroughly the standard toolkit fails on this company, because the failure is informative rather than inconvenient.

Every profitability measure here is dominated by an unrealized bitcoin mark, and every revenue measure by a software business that is not the reason anyone owns the stock. The comparison that would matter — market value against the bitcoin on the balance sheet — needs a current holdings figure and a current coin price measured on the same day. Our last reliable balance sheet shows bitcoin at $73.2bn as of September 30, 2025, against a market capitalisation of $61.6bn today, and we are deliberately not turning those two numbers into a ratio: they are a year apart, and the share count has changed in between. Anyone quoting a precise premium or discount to net asset value should be asked which date each half of it comes from.

What the oversold signal has paid on MSTR

MSTR is not oversold today, so this section is context rather than a setup. We ran the same test we ran on Meta: every occasion RSI closed below 30, and what happened next — over five years, and over the two years our own feed covers.

A median three-month gain of 45.8% looks spectacular until you remember the stock’s volatility is 84% a year. A three-month move of that size is roughly what this stock does when it moves at all; it is not evidence that an oscillator found something. And the one-month figure flips from +14.36% to −2.61% purely on the choice of start date, on 13 and 7 observations respectively.

There is a simpler reading. MSTR’s oversold episodes cluster in bitcoin’s drawdowns, and what followed them was bitcoin recovering. The signal is not telling you about MSTR’s chart. It is telling you, late, about the bitcoin cycle. We reached the same conclusion from a far better sample when we tested 637 oversold episodes across 55 large stocks.

Risks to reading it this way

The amplification is not a constant, and we have treated it as one. Measured from three dates it was 4.9×, 1.8× and 2.2×. That spread is the point — but it also means any sentence of the form “MSTR moves N times bitcoin” is false for most N.

Our bitcoin price comes from a different source than the stock. Tradier does not carry crypto, so BTC-USD is from Yahoo Finance. We cross-checked against IBIT on our own feed and the two agree within half a percentage point, which is why we are comfortable — but it is a seam, and bitcoin trades on weekends while the stock does not.

The balance sheet figures are stale. The last bitcoin carrying value we could pull is dated September 30, 2025. Any statement about premium or discount to asset value today would be guesswork, which is why this post does not make one.

Earnings in four days makes every level provisional. A stock with a 6.7% average daily range, reporting a result dominated by an asset that moved 42.7% in the quarter, can invalidate the entire levels table in one session.

Much of the narrative here is reported, not audited. The buyback figures, the MSCI rule change and the analyst positions come from press coverage between October 20 and 25. The bitcoin losses and the carrying value come from the filings.

What this means for your money

If you want bitcoin exposure, this is a decision about leverage, not about bitcoin. The fund that simply holds the coin tracked it within half a point over both of the spans we measured. MSTR gave you roughly double the move in both directions, and from the November 2024 peak, five times the loss.

Do not use the chart levels on this name as if they were Meta’s. An 8.5% band on a stock that moves 6.7% a day is not a trading range, it is a rounding error with labels.

For option sellers, 84% volatility four days before earnings is the highest-risk configuration on this site. The premium is enormous because the risk is enormous, and a strike 10% away is a day and a half of ordinary movement. Our market weather report covers how we think about event risk, and the Radar scores it on live quotes.

Beware the knowable windfall. A large third-quarter bitcoin gain is arithmetic from a coin price everyone can look up. If a headline presents it as a surprise, the headline is behind the market, not ahead of it.

Data appendix

Every figure above with the date it belongs to. Price studies are the ones this site runs on every symbol it scans — Wilder RSI(14), Bollinger Bands(20, 2), Wilder ATR(14), Chaikin Money Flow(20), the 50- and 200-day EMAs and the support/resistance logic in src/signals/technicals.ts.

Method: highs, lows and all percentage moves use closingprices, so they will not match the intraday 52-week range on a broker screen — both are given in the table above where they differ. Quarterly bitcoin losses are derived by differencing year-to-date filings, the same reconstruction we validated to the dollar against published figures on the Meta post; here the $8.3bn second-quarter figure independently matches what was reported at the time. Backtest episodes are the first session a condition becomes true; forward returns are close-to-close over 5, 21, 63 and 126 sessions. Bitcoin prices are Yahoo Finance BTC-USD because our own feed does not carry crypto, cross-checked against IBIT. Pulled October 25, 2026.