Why stablecoin float is not a reserve floor

Walk into enough treasury conversations and you will hear the same shorthand: “We keep reserves in stablecoins.” Sometimes that sentence is accurate. Often it describes money that moves every week to settle invoices, bridge FX, or wait for a spot purchase that has not yet been sized.

On our desk we draw two boxes before any percentages appear. One is labelled operational float. The other is labelled reserve floor. Stablecoin balances can sit in either box — but not in both at once on the same map.

What float looks like in practice

Float tends to have a rhythm. It rises before a known payment window and falls after. It may live across two or three venues. Someone on the operations team can name the next three uses without opening a policy document.

Reserve, by contrast, is boring on purpose. It has a stated floor. Touching it requires a named decision, not a routine settlement.

How the map records the split

During desktop allocation planning, we ask for the last six weeks of stablecoin movements if they are available. Patterns matter more than precise ticks. If half the sleeve behaves like float, that half leaves the reserve line — even if the wallet nickname still says “treasury.”

Clients sometimes dislike the resulting smaller floor. That discomfort is useful. It is better felt in a quiet sitting than during a week when spot sleeves are already noisy.

A modest rule we keep

We do not argue which stablecoin is “safer” in these sessions. Issuer preference is a different conversation. Our job on the map is narrower: name the sleeve purpose so Monday reviews stop mixing buffer with float.