Accepting crypto is an acceptance problem. Holding it afterward is a treasury problem. Most teams only budget for the first one.
A week of invoices does something nobody plans for: it builds a portfolio you never chose. You wanted to get paid. Instead you're now long BTC, holding USDT on Tron and Ethereum, USDC on Base and Solana, plus whatever SOL and ETH covered the gas-priced invoices. None of that was a trade. It's just sediment, and it carries price risk you didn't sign up for.
This is the part of running a crypto payment stack that the integration guides skip. So here's how we think about it, and what balance conversion is actually for.
First, the thing it is not
Balance conversion is asset-to-asset. You move value between crypto assets you already hold on the platform: BTC into USDC, USDT on one network into USDC on another, native SOL into a stablecoin. What it does not do is touch a bank. There is no wire, no fiat leg, no off-ramp. If your goal is dollars in a checking account, conversion is the wrong tool and you should stop reading here.
We're deliberate about this distinction because the failure mode is predictable: a finance lead converts everything to USDC expecting it to show up at the bank on Friday, and it doesn't, because that was never what happened. Conversion changes which crypto asset you hold. It does not change that you hold crypto.
What it does change is your exposure. And exposure is the whole game.
Why volatile balances are a liability, not an asset
There's a temptation to treat an incidental BTC balance as a free option. You got paid in it, maybe it goes up, why not ride it.
Because you're not a fund. Your treasury exists to pay suppliers, affiliates, and creators and to keep the business solvent, and every one of those obligations is denominated in something stable. A sharp drawdown in an asset you're holding by accident is not a missed upside — it's a real hole you now have to cover from operating cash. The asymmetry is brutal: the upside is someone else's trade, the downside is your payroll.
The conservative default, and the one we recommend to almost every merchant, is sweep to stable on a schedule. Not because stablecoins are magic, but because a treasury you can forecast is a treasury you can run. You know what you can pay out next week because the number doesn't move overnight.
The second axis nobody mentions: networks
Picking a stablecoin is only half the decision. Which network it sits on is the other half, and it's the one that quietly costs you.
Here's the trap. Your invoices credit USDT on Tron because that's what your APAC customers pay with. But your payout run goes to affiliates who want USDC on Base. So every payout cycle, you're holding the wrong asset on the wrong rail, and you bridge — paying fees and eating settlement latency — to move value you already had to where it needed to be.
Treasury rebalancing is the deliberate version of that. You look at where balances accumulate versus where obligations land, and you convert ahead of time, in bulk, when it's cheap, instead of reactively at payout time when it's not. We wrote up the mechanics in what treasury rebalancing actually means, but the operating principle is simple:
- Inbound concentrates by customer geography. Your customers choose the rail. You don't.
- Outbound concentrates by counterparty preference. Your affiliates and suppliers choose the rail. You don't.
- The two rarely match. Rebalancing is how you close that gap on your schedule instead of theirs.
If you've read our take on USDT across TRC-20 and ERC-20, this is the treasury-side consequence of that same rail fragmentation. Accepting on every rail is correct. Holding on every rail is not.
A policy we actually recommend
Vague advice doesn't survive a Monday. So here's a concrete starting policy you can adapt:
- Define your reporting currency. For most merchants this is a USD stablecoin. Pick one — USDC or USDT — and one or two networks where the bulk of your payouts land. That's your home base.
- Sweep volatiles on a cadence, not on a feeling. BTC, ETH, SOL, and any native gas-token dust convert to your home stablecoin on a fixed schedule. Daily if volumes are high, weekly if not. The cadence removes the "should I hold this" conversation entirely.
- Hold a working balance on each payout rail. If you pay affiliates on Base every Friday, keep enough USDC-on-Base to cover a typical run, so you're not converting under time pressure the morning payouts go out.
- Rebalance the rest toward where obligations land. Anything above the working balances drifts back to home base. When a network's balance runs short of its known commitments, top it up deliberately.
- Never convert what you're about to pay out anyway. If a USDT balance is funding a USDT payout next week, leave it. A conversion you immediately reverse is pure cost.
That last rule is the one teams break most often. The instinct to "just make everything USDC" feels tidy, but tidiness isn't the objective — predictable obligations met at minimum cost is.
What conversion does and doesn't protect you from
Worth being honest about the boundaries.
It protects you from price drift on assets you're holding incidentally. It protects you from rail mismatch between what you collect and what you owe. It gives you a treasury position you can forecast and report against a single number.
It does not protect you from holding crypto at all — you're still on-chain, still subject to whatever you believe about stablecoin issuers and network risk. It is not a hedge, and it is not a yield strategy. And it is, one more time, not a path to a bank account. If fiat settlement is a requirement, that's a different conversation and an off-platform one.
The shape of a good treasury habit
The merchants who handle this well aren't the ones with the most sophisticated model. They're the ones who turned it into a boring, scheduled operation: collect on whatever rail the customer wants, sweep the volatiles to a stablecoin home base on a cadence, keep working balances where payouts land, and rebalance the surplus.
Boring is the goal. A treasury that surprises you is a treasury that's about to cost you. The whole point of converting the crypto you didn't ask to hold is so that the next time you check the balance, the number is roughly the one you expected — and you can get back to running the business.
S. Brandt, halfin solutions