The calculator is a what-if tool: you put BTC up as collateral, borrow against it, and see what reaches you each month over 10 to 50 years, and what can go wrong on the way. Every result comes from a month-by-month simulation that runs in your browser. This page describes that simulation as it is in v13.15.
The model
- Two strategies: Loans to other IPA borrows once at the start and puts the cash into income assets; you live on the dividends left after interest and tax. Loans only starts with no loan and borrows your chosen income every month; nothing is bought.
- Loan: the starting loan = BTC × price × LTV (or this year's cycle-year target, if set). The origination fee comes off the top. Interest is APR ÷ 12 on the balance each month. Picking a lender loads its published terms; for lenders with size tiers, the APR and fee come from the tier of the starting loan (Loans only: the first year of draws) and stay fixed for the whole run.
- Interest options: pay monthly covers it from the dividends (Loans only: from the draw); withhold up front keeps N months of interest as a cash reserve (earning nothing) and invests the rest; let it accrue adds interest to the loan, compounding once a year (and at each maturity) in Loans to other IPA and every month in Loans only. In every mode, interest that isn't covered is added to the loan.
- Loan term, renewal and refinancing: the loan matures every N months (the loan term) from the loan start. In Loans to other IPA it is then either renewed at the same balance, or refinanced: the new loan = the larger of the old loan and target LTV × BTC value. The new loan pays off the old one; the extra, minus the fee, buys income assets. If BTC hasn't risen enough, the loan just rolls over. The renewal/refinance fee is charged on only the new money, the whole new loan, or nothing, per your setting; a fee the cash-out can't cover is added to the loan. In Loans only the loan renews every term, with the fee on increases in your yearly draw, on the whole loan plus that year's draws, or nothing. Renewals are always granted on the same terms.
- Loans only draws: each month you borrow your chosen income (optionally raised each year with inflation). If that would push LTV, measured after this month's fee and interest, above your pause-draws limit, the draw shrinks or pauses until BTC recovers. The loan (and the draws) can start after today; the projection always starts today. "Safe monthly income" is the highest steady draw that never pauses and adds no margin call or liquidation over the run.
- Income target (Loans to other IPA): optional. After any build phase you take up to the target from the dividends left after interest and tax, and anything above it is reinvested. If the dividends fall short you can borrow the gap against your BTC each month, under the same pause limit.
- Income assets: pick one or more and weight them; every purchase is split by the weights. Par preferreds pay rate × $100 par ÷ 12 per share each month (STRC actually pays in two halves each month). The others pay their starting yield on today's share price, growing at the asset's dividend-growth rate, and their price drifts at its price trend (model assumptions, shown in data.json). STRC's price is your purchase price, held flat except for the price-drop setting below. STRC is unsecured and ranks behind Strategy's debt, and Strategy can change its rate every month without notice.
- Income-asset risk: variable-rate preferreds (STRC, SATA) glide in a straight line from today's rate to your "rate in 20 years" over a fixed 20 years, then stay there (SATA shifts by the same amount as STRC); fixed-coupon STRF, STRK and STRD keep their coupons. Each asset's price falls in step with BTC's fall from its running high: a full bear (your cycle drawdown in projections, a fixed 75% in backtests) takes it down by the price-drop setting × the asset's BTC sensitivity, at most 95%. An optional suspension stops Strategy's preferred dividends (STRC, STRF, STRK, STRD) for N months; interest not covered is added to the loan.
- BTC price (model): a trend whose annual growth slides from the yr-1 rate to the yr-20+ rate over a fixed 20 years (then stays there), times a 4-year cycle wave anchored to the halvings (≈Apr 2024, 2028, 2032…). Peaks land about 18 months after a halving and bottoms about 30 months after, like the last three cycles. The wave is sized so each 12-month fall from peak to bottom equals your drawdown setting, and, if you tick the box, each crash after the first is 15% milder. The starting price is placed where today is in the cycle, or, if you choose, at the cycle top (worst case), the cycle low (best case) or a chosen cycle year.
- BTC price (backtest): real month-end closes (blockchain.info) from the start month you pick; the run ends when the data does. Cycle years follow the real halving calendar. Income assets are assumed to exist throughout, though STRC only launched in 2025.
- Margin calls and liquidation: each month the model first checks an intraday dip. In a month that closes lower, BTC is assumed to trade the dip setting below the close at some point, too fast to answer a margin call, so liquidation is checked at that price on the uncured loan, and BTC is sold at the dip price. Then, at the month-end close, a margin call (LTV at or above the margin-call level, or a dip into it) sells income assets, highest basis-to-value lots first, if the defence box is ticked, bringing LTV back to the lower of your target and 5 points under the margin-call level. If LTV is still at or above liquidation, the lender sells BTC. Its terms decide how much: everything owed sells BTC worth the loan plus the penalty and closes the loan (you keep your income assets; any shortfall stays owed, with interest); part sells just enough BTC, plus the penalty, to bring LTV back to the lender's reset level, and the loan carries on. A lender that doesn't publish its terms is modelled as a full liquidation. After a full liquidation there are no more draws, top-ups, contributions or refinancing.
- Cycle-year LTV: optional. Years follow the halving calendar (Y1 = halving year − 2, bear bottom; Y2 bull begins; Y3 halving; Y4 market top). The starting loan, each refinance and margin-call cures use that year's target instead of the main LTV. Income assets are never sold to pay the loan down to a lower target; only the margin-call defence sells them. Without cycle-year targets you can instead re-borrow back to the main LTV once a year whenever LTV has drifted below a set level.
- Build, then pay out: for the first N years, the set share of the dividends left after interest and tax is reinvested into BTC and/or the income-asset mix; after that, the payout-phase share applies. Bought BTC joins the collateral, which lowers LTV. The table shows BTC and income assets added each year: reinvestment, refinance and re-borrow buys minus sales in margin calls or liquidation.
- Add cash every month: your own savings go in each month (optionally rising with inflation) for the years you choose, buying BTC and/or the income-asset mix (Loans only: BTC). Bought BTC joins the collateral at that month's price. Contributions stop after a full liquidation. "Just holding" gets the same cash put into BTC, so the comparison stays fair.
- Taxes (simplified US): every income-asset purchase is its own tax lot. Preferreds (STRC, STRF, STRK, STRD, SATA) are treated as return of capital (ROC); SCHD as qualified dividends; O, JEPI and JEPQ as ordinary income. Each ROC payout lowers that lot's own basis tax-free; once a lot's basis reaches $0 its payouts are taxed at your capital-gains/dividend rate, so each year's batch turns taxable on its own clock (about 8 years at a 12% rate). An opt-in stress test ends ROC treatment for Strategy's preferreds after N years: from then on all their payouts are taxed at the capital-gains/dividend rate, whatever the lot's age (SATA keeps ROC). Unticking ROC taxes every payout. Margin-call sales tax the gain over basis, at the ordinary rate for lots held under a year; a liquidation taxes the BTC gain over your cost basis. Losses are ignored, and so is any deduction for loan interest. Tax comes out of that month's income, so income can go negative. Loan draws aren't income and aren't taxed. Unrealised gains are left untaxed: heirs get a step-up in basis. Treatment can change, so check with a CPA.
- Today's dollars: divides every dollar figure by (1 + inflation)^years (backtests: start-month dollars). LTV and BTC amounts don't change.
- Stress-test scorecard: on request, each stress is its own run of your current settings with one change, scored against your plan as set (outcome, total and first-3-years income, worst LTV, plan total). 2022 bear market, STRC at $71: for the months from the Nov 2021 close until BTC first closed back above it (Feb 2024, 27 months), the BTC price is your starting price × the real month-end close ÷ the Nov 2021 close (down 71% at the Dec 2022 low); after that it follows your own path's month-to-month moves from where the replay left off. Over those months the income assets fall in step with BTC so that STRC (BTC sensitivity 1) is at its worst recent close ($71, data.json) at the BTC low: the price-drop setting becomes 1 − $71 ÷ your STRC price (or your own setting, if larger), with a full bear = the replay's own 71%. The intraday dip applies as usual. Dividend suspension: the suspension above for 12 months from month 1, replacing your own. Loss of ROC status: the opt-in stress above after year 1 (taxed from month 13); not tested with taxes or ROC off. Lender failure: at the lowest BTC close in the loan's first 48 months (among months that start with a loan and come before any full liquidation), the loan, less any unused interest reserve, is set off against your BTC at that close, taxed like a liquidation sale (a debt bigger than the BTC stays owed); the BTC left over is frozen and either returned in full 24 months later (frozen) or never (lost). You keep the income assets and their dividends. While the BTC is frozen (and for good, if it is lost) there is no interest and no new loan: no draws, top-ups, refinancing or re-borrowing. When frozen BTC comes back, the plan borrows again from a new lender on the same terms, from that month: Loans only resumes its draws; Loans to IPA takes a new loan at that year's target LTV (fee off the top, any withheld interest kept back) and buys income assets with the rest; renewals keep the original term schedule. The stresses never change the main results.
- Monte Carlo mode: on request (“Run Monte Carlo”; model price path only, not in backtests), your plan as set is run over 1,000 random BTC price paths instead of the one model path. Each path is its own run, and everything in it works exactly as above, including the intraday dip, margin calls and the lender's liquidation terms. A path starts at your starting price and each month moves by the model path's own move for that month (the trend glide, the cycle wave and its return to the trend, as in the BTC price chart) times one real month: ln P(m) = ln P(m−1) + ln(M(m) ÷ M(m−1)) + e. The e are BTC's real monthly log returns, ln(month-end close ÷ the month before's), from Jan 2014 to the latest close in data.json, minus their average, so the middle path follows the model path while keeping history's fat tails and runs. Because the model's 4-year cycle is added on top of history's moves (which already contain cycles), year-to-year swings run about 6–13% wider than history's (monthly, 19.8% against 19.4%), so the liquidation odds lean high. They are drawn in 12-month blocks of consecutive real months: each block starts at a random month (any of them, with replacement; a block that runs past the latest month carries on from Jan 2014). The random numbers come from a seeded generator (mulberry32; each path has its own stream from the seed and its number), so a seed always gives the same 1,000 paths: seed 1 by default, and “New seed” draws another. The panel shows the share of runs fully liquidated, partly liquidated and with a margin call (in all, and by years 1, 3, 5, 10 and the end); the 10th, 50th and 90th percentiles (linear between neighbouring runs), each taken on its own, of year-1 monthly income (Loans only: draws), total income, the plan total, just holding, and plan total minus holding; the share of runs where the plan total beats holding; and a fan chart of the plan total or BTC price (the p10–p90 band and the p50 each month), all in the page's dollar setting. These are model paths built from history, not forecasts. Monte Carlo never changes the main results.
- Not modelled: lender default, custody or rehypothecation risk (except in the stress-test scorecard's lender-failure rows); a lender refusing to renew, or changing its APR, fees or LTV levels over time; answering a margin call by adding BTC or cash (the model only sells income assets, or does nothing if the defence box is unticked); ordinary-income tax on Strategy payouts after losing ROC (the stress test uses your capital-gains/dividend rate), capital losses, state-specific rules; income-asset price moves not tied to BTC (rate resets, credit events), trading costs and spreads.
Data and sources
- Lender terms (APR, size tiers, fees, margin-call and liquidation levels, liquidation terms, minimums, custody, US availability) come from each lender's published pages, last checked on 2026-10-01. Where a figure isn't published the table shows a dash and the model keeps your own setting. Rates move often, so confirm with the lender before you borrow.
- Income assets (price, dividend rate or yield) are refreshed monthly from the issuers and market data. Dividend-growth, price-trend and BTC-sensitivity figures are model assumptions, not forecasts.
- BTC history for backtests: month-end closes from blockchain.info's market-price chart. The live price comes from CoinGecko when you press "Live price" (and once on a first visit).
- Referral links: a lender link marked * is a referral link; if you sign up through it we may earn a reward. Referrals never change a lender's terms in the model or its place in the list.
- Privacy: on bitcoinloanincome.com the calculator counts visits and feature use with Google Analytics, never your inputs or results. Your scenario stays in your browser (and in the link, if you share it). This page loads no scripts.
This is a what-if tool, not financial advice. The 4-year cycle is a historical pattern, not a law. A deep drawdown right after you borrow is the main way this strategy fails. Tax treatment is simplified and can change; check with a CPA. See the changelog for what changed in each version.