INTRODUCING THE MONSTER MODEL π¨ππ
The laboratory-grade corporate treasury and Bitcoin simulator.
The Monster Model is a fully parameterized discrete non-linear dynamical system that models the complex reflexive interactions between
@Strategy,
@Strive, and
#Bitcoin in a unified framework.
Key features:
β 345 input parameters. Everything from assumed weekly
$STRC /
$SATA net proceeds to target amplification range, mNAV variant, and Bitcoin hodler flows is included. Parameters that don't currently apply to your configuration disable or disappear on their own.
β 159 plottable series. This includes the basics like
$MSTR and
$ASST share price, Bitcoin price, shares outstanding, BTC reserve, USD assets, Duration, mNAV, and amplification, as well as metrics that are novel to the Monster Model such as
$STRK effective in-the-money value, effective sats per share, expected break-even time horizon (EBETH), and Bitcoin's expected one-year forward return. Expected-return premiums for MSTR and ASST, based on work by
@stonychambers, are also provided.
β 36 charts. MSTR and BTC price, BTC reserve, BTC flows, shares outstanding, modeled BTC liquid/illiquid supply, net proceeds, market cap, USD and months of coverage, reserve valuations, mNAV, amplification, sats per share, convertible bond tracking, and more.
β Powered by a bespoke supply-and-demand price clearing model of the Bitcoin ecosystem. The featured method for modeling Bitcoin tracks projected mining emissions, lost coins, Strategy and Strive's own purchases, hodler withdrawals from liquid float into illiquid storage, and releases from illiquid storage back into the liquid float β cleared against a demand trajectory, week by week, into a price. Switch it on and the treasury companies' own buying pressure feeds back into the prices they pay.
β 6 Bitcoin price paths, or compose your own. If you prefer a static price path instead of a dynamic one, the Monster Model offers a menu of choices:
@Giovann35084111 Santostasi's classic power law; the Bitcoin. com and BitcoinFairPrice variants;
@Saylor,
@Shirishjajodia and
@CJ_Bitcoin's Bitcoin24;
@TheRealPlanC's Q1 power law floor; Stephen Perrenod's (
@moneyordebt) 8-parameter log-periodic curve. Or build a custom path out of normalized variants of a power law, an S or Gompertz adoption curve, a linear incline, a periodic wave, a decaying exponential, and Perrenod's 8-parameter curve.
β Full convertible bond accounting. Every bond either company has ever issued or held, retired ones included, with Black-Scholes option deltas producing a novel share accounting method alongside BSO, ADSO and FDSO: *effective* shares outstanding. Parameters govern how eagerly holders exercise puts and conversions, and whether Strategy calls as soon as the bonds' 30% premium condition permits.
β 5 mNAV variants. Multiple of enterprise value (mEV), plus three variants of price-to-book (P/B) that use basic, effective, and assumed-diluted treatments, plus the recently introduced fully diluted multiple of Net Reserve (mNR) in which out-of-the-money convertibles count as debt rather than as future shares. Picking a variant automatically selects the matching share count *and* liability treatment, so the numerator and denominator can never quietly disagree.
β 5 ways to drive mNAV forward in time. A constant expansion scalar that gaps and holds; a multiple of break-even mNAV; a parameterized oscillating sinusoid (with or without a decaying amplitude); a linear rise (or fall) towards a terminal mNAV; or an experimental sentiment algorithm built on fast and slow SMAs plus week-over-week Bitcoin moves.
β Weekly time steps, aligned to the filings. Strategy and Strive announce by 8-K, so the model breathes at that cadence. Quarterly bond interest and monthly preferred dividends land where they actually land, and the rhythm between capital raises is visible instead of averaged away.
β Real history back to 2024. Common share issuance, BSO, ADSO and BTC holdings, taken from Strategy's 8-Ks β used exactly from mid-November 2024 forward, when the filings became regular. The earlier filings are sparser, so a few of those weeks are interpolated to smooth the gaps between them. (Strive's history starts in September 2025, with weekly 8-K coverage from June 2026.)
β An experimental credit collateral subsystem. Adoption curves for Bitcoin pledged against credit, with carry spreads, eligible and pledged fractions, and lender LTV β and some of that credit recycling back into Bitcoin demand.
β STRC volume cycle modeling. If you expect STRC to trade at or above par into its record dates on rising volume, this parameterized sub-model reproduces that rhythm in its projected capital raises. Otherwise, index either an initial weekly net proceeds estimate or a fraction of the total notional value to a growth curve of your choice.
β One-off equity events. Drop a hypothetical issuance or buyback β any share count, any price, any future week β and watch the proceeds, or the funding deficit, propagate through the same allocation rules as everything else. Try using it to explore the potential impacts of various levels of exercise on Strive's outstanding warrants.
β 472 modeled series across 8 interconnected systems: convertible debt, preferred equities, common equity, Bitcoin price dynamics, USD assets, the BTC reserve, mNAV, and amplification.
β
And the model is only half of it. The other half is a premium-grade visualization panel built for tweaking and customization.
β Automatic or manual recalculation. Change any assumption and the entire system re-solves in a fraction of a second. Or, switch to Manual mode and queue up several changes to see their combined effect in a single run.
β A chart grid you arrange yourself. Drag to reorder, insert a slot anywhere, delete several at once with undo, star your favorites. Stretch charts to fit, or size them to the pixel. Embiggen one to the full pane and the rest stop recomputing behind it.
β Charts that follow your parameters. Set Bitcoin to a static curve and the supply series leave the reserve charts, because supply plays no part in that model. Switch amplification from a ratio to a multiple and the axis, the tick format, the series and its target band all change together, in one step.
β Pin a reading and watch it move. Hover a chart to read every visible series at that date; press to pin it in place. It holds while you change parameters, with a column reporting how far each series moved since the last run β so you watch a number move instead of remembering it. Synchronize the grid and one press freezes every chart at a single date.
β Six auto-saving workspaces: five yours, plus a sandbox that catches links other people send you. Keep the bull, base, and bear cases side by side. Give them custom names, copy between them, reset a single tab, export them to JSON and carry them to another machine.
β Shareable links. Your entire parameter set, encoded into the URL. A scenario that differs from the defaults in a dozen places travels in about 50 characters β it records only what you changed and packs each value into the fewest bits its own range allows. It arrives in a reserved sandbox workspace, so opening someone else's scenario never disturbs yours.
β Copy any chart as an image. Not a screenshot β composed for the purpose, with its own title, legend and margins, none of the surrounding interface, your hidden series still hidden, your log axes still logarithmic, and your pinned readout included.
β Live prices. History ends on the last completed week, so the week you're actually in is a live row: mNAV, amplification and reserve value reflect where things stand right now rather than at last Monday's close. Switch it off for the neutral "what if nothing else happens" baseline, or when conducting controlled parameter impact analysis.
β Operable entirely from the keyboard. Not most of it β every control, every menu, every modal. Including lifting a chart off the grid and walking it to a new slot with the arrow keys, or even repositioning a readout within a single chart.
β It even works on your phone. The model was built for a big screen and that is still where it shines, but the mobile version keeps the core of it: tap a chart and the marker lands on that date; drag to sweep it with every value updating as it goes; tap again to pin. Tap anywhere else you would normally click on desktop.
β ~42,000 words of documentation, delivered where you need it. Every one of the parameters, every plottable series and every chart carries a written description on hover β not parked on a docs page you have to go and find.
β
A word on how this was built.
The framework began as a Google Sheets workbook in late summer 2025, built by hand. I had two rules: no formula went in unless I understood how it worked and believed it earned its place; and AI was never allowed to operate on the spreadsheet itself.
The site is that workbook translated into TypeScript and given a thoughtful interface. AI was used heavily and deliberately throughout that process.
I spent 18 years in high-performance computing. That work teaches you one thing above all others: numerical integrity comes first β even before performance β and a number nobody checked is not a result. AI is very good at producing numbers nobody checked.
So every step of the way, the TypeScript output was validated against the source workbook row by row, series by series. Not a spot check.
I've invested over 1,600 hours in the project, the vast majority of that in the last five months. I've never worked harder on anything in my life.
β
The Monster Model is live at
monstermodels.live. No account, no signup β and your scenarios never leave your browser unless you want them to.
For the best experience, I recommend using a desktop or laptop browser.
Please give it a try and tell me what you think! And if you like the tool, please consider reposting this announcement.
This has been an incredible journey, but my hope is that this is just the beginning. Thank you all for your support! π§‘