Price pathThe market every Agreement lives on. Everything downstream — non-performance, settlements, recoveries — is a consequence of this path.
Start price?BTC price at month 0 — the entry environment for the first cohort.
Mode?Bridge: pinned start-to-end, volatility fills the middle. Historical replay: BTC's actual monthly moves from a chosen start month, rescaled to your start price. Real returns, trend removed: real monthly returns reshuffled in blocks with the average trend stripped out — history's volatility, no assumed appreciation.
End price?Where the simulated path is pinned to finish; the bridge fills everything in between with the chosen volatility.
Volatility %/yr?How violently the path swings between its pinned endpoints. 43% matches the recent coin (2023+ realized ≈ 42%); the 2017+ regime ran ≈ 70%.
Shock designer?Overlay a crash on any path: drop Z% starting at month X over W months, then stay down or recover.
LiquidationWhat a forced coin sale realizes when a customer stops paying. The sale executes at the following month's simulated price (the 25-day clock, taken literally).
Haircut % (sale below spot)?Extra discount on forced sales below the reference price. Default 0: a 1-BTC sale has no market impact, and the 25-day delay is already priced by selling at the NEXT month's simulated price.
Market-sale cost (bps)?Exchange execution cost on liquidation sales, in basis points. 25bps = 0.25%.
Paper termsThe contract economics of each Agreement: what the coin costs, what the customer pays, and what BTC Now takes along the way.
Term (months)?Length of every Agreement. Payment size, the servicing fee and the default curve's shape all re-derive from this.
Price multiple ×?The Agreement price as a multiple of the coin's cost: 1.475× means a $60,000 coin is sold for $88,500, paid in equal monthly payments.
First N payments → BTC Now?The first N monthly payments route entirely to BTC Now as origination compensation. The Purchaser receives payments N+1 onward. 0 = off.
Purchase price % of coin cost?What the Purchaser pays per Agreement, as a share of the coin's cost — 100% is par. The negotiable in every forward-flow term sheet.
Servicing fee %/yr?BTC Now's servicing rate, charged as a flat skim on every dollar actually delivered to the Purchaser: rate × term-in-years. Dead paper pays no fee.
fee on delivered $: 5.00%
PacingHow the book is built over time. Paced entry ladders the strikes — the built-in dollar-cost averaging that decides how a crash lands.
Monthly cohorts?How many months in a row new Agreements originate — the length of the forward-flow purchase commitment.
Agreements per cohort?How many Agreements are signed in each originating month.
Intramonth strike dispersion?ON: each Agreement draws its own entry price around its month's level. Changes texture, not economics.
Origination stops at month?Hard stop on new purchases from this month on (0 = never). Existing Agreements run to completion.
Seed?Random-number seed. Same seed + same inputs = the identical run, dot for dot.
Defaults & behaviorWho stops paying, when, and why — the credit and behavior assumptions. The engine draws individual customer outcomes from these.
Lifetime default %?Share of customers who ever stop paying over the full term. With a custom timing curve it is the sum of the per-year bars — editing it here rescales every year proportionally.
timing shapeHow the lifetime number is laid out over the term. Hump: the program's actuarial curve — heaviest in years 1–2. Flat: the same total spread evenly. Late-loaded: mass ramping toward the final years — an openly synthetic stress.
when they stop payingThe lifetime number laid out over the term, from the last engine run. Each bar is a share of the WHOLE book; together they sum to the lifetime %. WHEN a stop lands matters as much as whether — a year-1 stop has collected almost nothing against full exposure, a final-year stop is nearly harmless because the schedule is amortized away.
Rational default?ON: only customers whose coin is worth less than their remaining obligation stop paying — the same default budget, timed as adversarially as possible. OFF: stops land randomly in time, like real-life job losses.
Drawdown-scaled hazard?The graded middle ground (memo §09): each month the stop probability is the baseline hazard × a state multiplier on the coin vs entry — in the money ×0.5, drawdown ≤30% ×1.0, 30–50% ×1.5, 50–70% ×2.0, deeper ×3.0. With it ON, this engine and the independent Python model agree within ±0.3pp on the crash grid.
Lost-conviction rule?Deterministic capitulation: anyone whose coin sits X% below their entry price for Y consecutive payment dates walks away — the behavioral stress on top of the random/rational engine.
Early settlement %/mo?Chance per month that an in-the-money customer pays off the whole remaining schedule early, scaled by how far in the money they are — so even 100 is not instant: it means the fully-rational exit at the moneyness-weighted rate.
Take-profit gate?ON: they only settle once the coin beats their ALL-IN cost — the full Agreement price, sunk payments included — by the % below. OFF: anyone whose coin beats their remaining payments may settle.