STRC ("Stretch") is a perpetual preferred stock issued by Strategy Inc. (the company formerly known as MicroStrategy, ticker MSTR). It IPO'd on 29 July 2025 at $90 per share with a $100 par value, and trades on Nasdaq under the ticker STRC.

Unlike Strategy's other preferreds — STRF (10% fixed), STRK (8% fixed), STRD (10% fixed) — STRC is a variable-rate instrument. The company resets the dividend rate each month (Strategy has signalled a switch to fortnightly resets in the near future). Holders receive a cash dividend on each ex-dividend date; the cash amount divided by par × 12 is the declared annualized yield (currently ≈ 11.5%).

STRC was created as a capital-raising vehicle for Strategy's BTC accumulation strategy. The proceeds from each share sale (and the IPO itself) are used to buy Bitcoin on the open market. Through the IPO and weekly ATM offerings since November 2025, Strategy has raised $8.25B and acquired ~98,843 BTC funded by STRC alone.

The instrument is designed to trade close to par because the variable rate is reset to keep yield at the prevailing market clearing level. When STRC trades above par, Strategy can issue new shares at par-or-above and capture the spread as cheap capital. When it trades below par, the company pauses issuance — issuing below par would dilute existing holders.

Threshold mechanic

The model is a deterministic threshold + capture formula. For each day's prediction:

daily_proceeds = capture × volume × close × (1 − haircut) × 𝟙(close ≥ $100)

The indicator function 𝟙(close ≥ $100) is 0 when STRC closes below par (algorithm PAUSED), 1 when it closes at or above par (algorithm ACTIVE). When the close is within ~5¢ of par, the dashboard shows ARMED as a watch state.

Capture rate

On eligible days, Strategy doesn't sell every share that trades — it captures roughly half of the day's volume × close as proceeds. The exact ratio (the "capture rate") is the model's most-uncertain parameter. It is refit every Monday from the previous week's 8-K filing. The current value is shown in the capture rate chart.

The haircut (≈ 0.13%) accounts for fees paid to ATM sales agents and is held constant since the public 8-Ks don't break out fees in a way that lets us identify it cleanly week-by-week.

Forecast: observed + projected (week and cycle)

The "this week" forecast covers Mon-Fri of the current calendar week. The "this cycle" forecast covers all business days from the last ex-div through the next ex-div. In both cases, observed days use actual close × volume; unobserved days are filled with projected close (linear-trend OLS through current-cycle daily closes) and projected volume (10-trading-day median). The threshold gate still applies — if the projected close lands below par, that day's projected proceeds are 0. As the period rolls forward, fewer days need projection and the trend has more anchor points, so the forecast tightens.

Weekly recalibration

Every Monday at 11:30am ET, Strategy files a new 8-K disclosing the prior week's STRC ATM activity (shares sold, net proceeds, BTC bought). The dashboard ingests that 8-K, refits the capture rate, and overwrites the live coefficients. The forecast horizon then projects forward from that point using the new capture rate against the live price + volume stream.

Where STRC fits in the BTC flywheel

STRC's stable price (anchored at par) and premium yield (currently 11.5%) make it a high-quality collateral asset. As volatility falls and Sharpe rises, holders can borrow against STRC at favourable terms to buy more STRC — which Strategy then issues at par to fund more BTC purchases. The cycle compounds: more issuance → more BTC → higher BTC NAV per MSTR → higher MSTR premium → cheaper capital.

Limits and caveats

The model assumes the threshold gate is a hard binary (it is, mechanically) and that capture is roughly stationary week-to-week (it isn't — it varies between 0.27 and 0.79 in the realised history). The capture-by-cycle-phase analysis is currently underpowered (n=10 weeks). Forecasts are point estimates with no uncertainty bands; treat them as midline expectations, not confidence intervals — particularly the projected-day component, which compounds three uncertain projections (close, volume, capture). The cycle forecast carries more projection uncertainty than the week forecast because more days are projected.