Fees Are Near the Floor: Do the Boring Custody Work Now

Most fee commentary is written when fees are spiking, because that is when it feels urgent. This is the opposite piece. As I write this, mempool.space shows recommended fees of 4 sat/vB for the fastest confirmation, 3 sat/vB for a half-hour target, 1 sat/vB for a one-hour target, and 1 sat/vB as the minimum relay fee. With bitcoin trading around $81,067 (roughly €72,509) and the chain at around block 972,000, the mempool is quiet and blocks are clearing at the floor. A window like this will not announce itself. It will just close.

So let me be plain about what this article is. It is not trading advice, and it is not a prediction about where the price goes next. It is about custody — the plumbing of holding your own keys — and the simple arithmetic that says certain one-time jobs cost twenty times less when fees are at the floor than when the mempool is congested. If you have been meaning to do any of the three things below, the window is now.

Why consolidation is cheap right now

If you have been stacking for a few years, your wallet probably holds dozens or hundreds of UTXOs: small deposits from DCA buys, remnant change from old spends, dust from rounding. Each one is a line in your wallet, but more importantly, each one is a future input. When you eventually spend, the transaction has to reference every input it consumes, and inputs are by far the heaviest part of a transaction's size.

A realistic example. Say you have twenty UTXOs you want to merge into one clean output. A transaction with twenty inputs and one output weighs roughly 3,000 vBytes once you account for signatures and scripts. At today's 1 sat/vB, that costs about 3,000 sats — around $2.50. In a congested period, when people were paying 20 sat/vB to get confirmed, the identical transaction cost about 60,000 sats, or roughly $49. Same coins, same move, twenty times the price. The only variable is when you did it.

That is the whole argument. The consolidation itself changes nothing about how much bitcoin you own. It is pure overhead, and overhead should be bought at a discount. Buying it at 1 sat/vB instead of 20 is the custody equivalent of paying list price versus waiting for the sale — except the sale timing is not a mystery, because mempool conditions are public, and right now they are about as cheap as bitcoin has ever confirmed transactions.

The privacy dividend nobody prices in

Fee savings are the obvious benefit, but consolidation does something else that matters more over a holding period measured in years: it cleans up your on-chain fingerprint. A wallet with two hundred fragments leaks. Every future spend that pulls from a handful of tiny inputs links those inputs together, and by extension links everything they were ever connected to. Chain analysts build their business on exactly this kind of stitching.

A consolidated wallet does the opposite. One transaction merges everything once, at a moment you choose, and after that your spending transactions look boring: one input, one or two outputs, nothing to correlate. You are not hiding — bitcoin's ledger is public and pretending otherwise is dishonest — but you are making it harder for passive observers to build a profile of your holdings and habits out of years of small transactions. Do the merge in one cheap, deliberate transaction now, rather than leaking linkage a few inputs at a time every time you spend for the next five years.

One caution while you are at it: consolidate within the same wallet, into addresses you control. Never send everything to a fresh address from an exchange withdrawal or any service that links your identity to the receiving address, because that stamps the new, tidy UTXO with the one label you were trying to shed.

Set up multisig — and actually rehearse it

The second job on the list is multisig. If you hold a meaningful amount of bitcoin on a single seed, a 2-of-3 setup is the standard upgrade: two keys to spend, so one lost or stolen device does not mean disaster, and no single point of failure. The setup work is free. The migration of funds onto it is not, and that is where the fee window matters — moving a wallet's worth of coins onto a new multisig quorum is exactly the kind of large, input-heavy transaction you want to pay single-digit sats/vB for, not congestion prices.

But do not stop at setup, because an untested multisig is a rumor you tell yourself. Rehearse the recovery: sign a small test transaction, then practice the full restore path on clean hardware — seed words, configuration files, wallet descriptor — before you trust it with real money. Multisig failures are almost never cryptographic. They are lost coordination files, mismatched wallet software, a key that lives in a drawer nobody can find. A quiet fee period is also a calm one, which is the best possible environment for doing this carefully instead of under stress. When you are done, your backup plan will have actually been executed at least once, by you, rather than existing as a PDF.

Open the Lightning channels you keep postponing

The third job is Lightning. A channel open is an on-chain transaction, and a typical one weighs 150 to 200 vBytes. At 1 sat/vB that is under 200 sats — about $0.15. Open the same channel during a fee spike and you are paying several times that just for the privilege of starting. For anyone who wants to spend bitcoin day to day without touching their cold storage, Lightning is the tool, and the channel open is a one-time cost per channel. Cheap windows are when you build capacity.

The point of the floor

None of this requires a view on the market. The price could run from here or fall from here, and the argument does not change, because the work is not speculative — it is maintenance. Wallets accumulate clutter, keys age, untested backups stay untested. The only thing the fee floor changes is the price tag on fixing all of that in one pass.

Windows like this one show up a few times a year and never send an invitation. The mempool is quiet at 1 sat/vB minimum right now, at block ~972,000, and the three jobs above — consolidate, set up and rehearse multisig, open Lightning channels — are all cheaper today than they will be the next time the queue forms. Do the boring custody work now. It is the least glamorous thing you will do with bitcoin this year, and quite possibly the most valuable.