Cash does three jobs
How Much Bitcoin? is a related exploration, examining the fraction of your portfolio that bitcoin might justify occupying. This page examines the bitcoin allocation strategy from a different angle: how much fiat belongs beside your stack, what that fiat is actually for, and what holding it costs you — the opportunity cost of not holding that fraction in bitcoin instead.
You hold bitcoin, but sometimes cash is needed. Rent, tuition, a new roof, a hospital bill: all of them require cash, and many come as a surprise. That mismatch is why this page exists. A stack that is 100% bitcoin can be a useful strategy, but it is a portfolio that has pre-committed to selling bitcoin at whatever price the market happens to be showing on the day life sends a bill.
A buffer of fiat beside the stack does three jobs, in descending order of weight. The first is that it is a shock absorber: money you can spend without touching your stack, so a surprise expense never becomes a forced sale of bitcoin at the lows. The second is that whatever the shock absorber does not consume is dry powder, and if price happens to visit the floor of the channel while the buffer is still intact, that becomes a contingent bonus. The third is sleep: the buffer you hold so that you can hold the stack.
That preparedness comes with a cost, and this page examines that cost. Every dollar in the buffer is bitcoin that could have been held, but having dry powder might also mean ending up with more bitcoin.
Where you are decides
Raising cash is selling — converting part of one stack into its other form, not setting money aside from it. The full cycle — sell, wait, rebuy — is the round trip, and where in the Power Law channel you run it decides whether it ends in more bitcoin or less. Set the split and a tax rate, then work the two sliders below — Step 1 is where you sell (the ▲ marks today), Step 2 is where you rebuy.
The rebuy side is yours to set. Leave it at your sale for Wait-or-Deploy’s rule — the first lower entry within two years, no timing, no discretion. Slide to a deeper target and the stats change honestly: bigger multiples when it arrives, arriving less often, after longer waits — while the trend compounds under the cash. If the target never comes, the cash rebuys at the two-year price, sometimes higher than you sold.
Deciding the redeploy on its own — cash in hand, no round trip — is Wait, or Deploy Now?’s question; this page couples the two ends of the same round trip.
Exploring your first split, or tracking one you already made? Set the sell slider to where you actually sold, and the rebuy target to what you’re waiting for. Sold, and the price ran past you? You are living the no-arrival branch — the target never came. This page’s honest answer is the two-year rule: at some point the strategy ends and the cash goes back, sometimes above your sale. Deciding when, from here, with cash in hand, is Wait, or Deploy Now?’s question →
Your strategy
One stack, two forms: the share you hold as cash, ready — the rest stays bitcoin. Same monetary energy, split for the round trip.
Assumes a fully appreciated stack, long held. A higher cost basis softens the haircut. Buying back is not taxable — one haircut per round trip.
Optional. Entering a number of bitcoin may help orient you to the potential outcomes; the verdict does not depend on it.
—
In a tax-advantaged account this haircut disappears — set the rate to 0% to model it; Disciplined Rebalancing makes the same recommendation.
—
The Power Law channel across history. The blue dashed line is your sell position — with its neighbourhood dots, the historical entries the answer is built from; the green dashed line is your rebuy target. The gap between them is the strategy’s span. When the target sits at or above your sale, the rule is simply any lower entry — the green line rides your sale. The record before 2014 is dimmed: the model draws its entries only from 2014 on.
—
of sells here saw the target within two years.
to the target, when it arrived.
—
historically, selling at your Step 1 position and rebuying by your Step 2 rule left you with more bitcoin this often — counting both branches.
—
—
Show the work — the record behind these numbers →
—
—
—
Past 80% this stops being a buffer and starts being an exit. Worth remembering before you go further: a stack this hedged has already given up most of its drawdown risk — and most of its upside. Selling this much realizes the tax bill in one season. And the Power Law is a floor observed, not a ceiling promised: if bitcoin ever breaks it upward into an S-curve, the bitcoin you sold is what misses it. The default posture of this site is HODL; the opportunity here is a conservative trim, not a departure.
Where the strategy has paid
This chart is the summary, not a control — it moves with the tax rate AND the rebuy target you set above. For every sell position in the channel: the median bitcoin that came back per bitcoin sold, after tax, following your rebuy rule. Above the line the round trip earned bitcoin; below it, it cost it. Change the settings and watch the crossing move.
The blue point is your sell position — the same blue as the dashed sell line on the channel chart above; the glowing point is today. The shaded band is the stretch your strategy spans, from rebuy target to sale.
—
Dry powder, and its mirror
There is a version of this page that sells cash as a war chest for buying the bottom, and it would contradict a page we have already published. Wait, or Deploy Now? asks a cash holder whether to deploy now or wait for a better entry, and finds that from most positions in the channel, waiting was the wrong call. This page asks the opposite question of the opposite reader: an all-in holder, deciding whether to hold or raise.
They are the same question through opposite doors, and they read the same record with the same machinery, so they cannot disagree. Where Wait-or-Deploy says waiting was foolish, this page says selling to rebuy lower was foolish, and for the identical reason: no lower entry arrived. Where it says waiting earned its keep, this page says a trim earned its keep. That is why the floor branch here is a warning rather than an opportunity, and why the two pages cite each other as one finding read from two sides.
Sleep, and the other way to never sell
The third job does not compute. A buffer is also the thing that lets you sit through a drawdown without doing something you will regret, because the bill that would have forced your hand is already covered. The allocation page puts it as a question worth asking before any of the math: can you hold it? A position you cannot hold through a bad year is not really your position. Some of what a buffer buys is simply the ability to keep the rest.
There is another answer to never selling, and it deserves its paragraph. Borrowing against your stack gets you dollars without a sale, and on paper it dominates a cash buffer: no bitcoin foregone, no tax, no drag. Its failure mode is the whole subject of this page. A credit line against bitcoin is collateralized by bitcoin, so it tightens exactly when bitcoin falls, and the margin call arrives at the same trough where the forced sale would have hurt most. The two options fail at the same moment, and only one of them is still there when it does.
A loan against your stack is a buffer that can be taken away at the exact moment you need it.
That is not an argument against borrowing, which has its own page and its own honest math. It is the reason the two cannot be treated as substitutes for the same job.
This is Disciplined Rebalancing, priced
Disciplined Rebalancing states the protocol — trim strength, rebuy weakness, by zone. This page prices it: what the trim costs after tax, where in the channel it has historically paid, and how long the round trip took. Same discipline, with the work shown.
One identity drives the page. Sell a fraction x of the stack at price P with tax t, and rebuy at P_rebuy: you end with (1−x) + x·(1−t)·(P / P_rebuy) of the stack you started with. The round-trip multiple on the sold slice is (1−t)·(P / P_rebuy) — the number plotted above. Breakeven needs the rebuy more than t below the sale, which is why the tax rate moves the crossing. Your stack size cancels out of all of it, which is why the BTC input changes nothing but the dollar labels. Tax assumes a fully appreciated stack (zero cost basis): a real basis softens the haircut.
Two limits worth naming. When no dip arrives inside two years the rebuy happens anyway, at the two-year price — often higher than the sale, which is how a round trip loses bitcoin rather than merely failing to gain it; those rows are in the table marked "no dip". And at the extremes of the channel the record thins out: the neighborhood widens automatically to keep a workable sample, so a figure at the far upper band rests on fewer entries than one near trend. The entry count and band width are stated with every verdict.