How companies earn yield on idle treasury cash

Six structures compete for the same corporate balance, and they differ on liquidity, minimum size and who holds the claim rather than on headline rate.

Idle corporate cash earns yield through one of six structures. Those are a bank sweep, a fixed-term fund, a directly held T-bill ladder, an embedded treasury API, a tokenised money market fund, or an onchain vault. They mostly draw on the same underlying short-dated government and bank paper, so the rate converges. What separates them is how fast you can get the money back, how much you need to commit, and who your counterparty is.

That last point is where most treasury decisions actually get made. A finance team choosing between these is rarely optimising for 20 basis points. They're deciding whether the cash has to be available on a Friday afternoon, and whether the board will accept a new counterparty. Then whether the reporting lands in the month-end close without someone rekeying it.

What are the six structures?

Each row below is a different answer to the same question. The yield source column matters most, because several of these deliver near-identical exposure through very different wrappers.

Four of these six are wrappers around short-dated government paper. As of 4 August 2026 the aggregate value of tokenised US Treasury funds sits at $16.17 billion on rwa.xyz, with a 7-day average yield of 3.28% across the category. That number is a useful floor. Any structure charging you for access to the same T-bills has to beat it after fees, or justify itself on liquidity and operations instead.

What problem does each wrapper solve?

The wrapper exists to solve an access problem, and each one solves a different one. Buying T-bills directly needs a brokerage relationship, an operations person to run the ladder and somewhere to book the accounting. Every structure above removes some part of that burden and charges for it.

• Bank sweeps solve deposit insurance and effort. A sweep network spreads the balance across partner banks for expanded coverage, and the cash stays immediately available. You take whatever the bank passes on after its spread, which is usually the lowest rate on this list.
• Fixed-term funds solve counterparty concentration. Instead of leaving several million in unsecured exposure to one bank, you get clean pass-through exposure to a chosen issuer for a chosen term. TreasurySpring has issued fixed-term funds in the tens of billions to a client base including FTSE 100 corporates, charities and family offices.
• Managed ladders solve operations. Meow runs the buying, the auto-roll and the maturity management at 1 basis point per month annualised, and the accounting integrations mean the gains and losses land in your books without manual work.
• Embedded APIs solve build cost for a platform rather than for a treasurer. AtomicVest gives a fintech treasury bills, money market funds and insured sweep behind its own front end, so the platform ships a yield product without standing up the regulatory and compliance apparatus itself.
• Tokenised money market funds solve settlement and composability. The fund share is a token, so it moves on a blockchain rail rather than through a transfer agent's calendar, and it can be posted as collateral elsewhere.
• Onchain vaults solve mandate flexibility. The strategy can span several instruments at once, and the perimeter it reaches is enforced in code rather than described in a document.

Read that list as a menu of problems rather than a ranking. A treasurer with a clean bank relationship and a patient balance has one problem. A fintech that needs a yield toggle live by quarter end has a different one.

Where does an onchain vault sit in this?

An onchain vault is a layer below most of the others rather than a competitor to them. Upshift is non-custodial vault infrastructure for onchain yield. Fintechs, exchanges, wallets, neobanks, chains and asset managers use Upshift to launch custom yield products without building vault contracts, risk controls, NAV accounting, fee logic or withdrawal processing themselves, and depositors remain self-custodied throughout.

The distinction that matters commercially is between a surface and an engine. A surface owns the customer, the onboarding and the brand. An engine holds the balance and produces the return. Some firms in this category run both, and some sell only one.

The last two rows occupy the same slot on different rails. An embedded treasury API is the closest structural analogue to a vault in this list, because both sell the engine rather than the customer relationship. What separates them is the instrument set each rail can reach, and how fast a new one gets added.

Those two rails also compose. An embedded treasury API already carries a menu of instruments, and a vault-backed option can sit on that menu next to the treasury bills and the insured sweep. We build for exactly that: a partner distributes the Upshift vault catalogue through its own API, its downstream clients switch the option on inside their own product, and the partner takes a share of the fee on the balances it brings. A platform like AtomicVest could add onchain treasury management to its instrument set that way, without building vault contracts or withdrawal processing. Vault-as-a-Service covers what the partner side involves.

What does a vault do that a fund wrapper can't?

Three things, and none of them is a higher headline rate. Treat the rate as a function of the strategy inside, which is covered in the strategies onchain vaults run.

The trade against those three is real and worth stating plainly. Contract risk has no analogue in a fund wrapper. Upshift's contracts have been through 10 audits by 6 independent firms as of August 2026, and audits reduce that risk without removing it. These vaults carry no built-in smart contract cover, though third-party cover exists in the market from underwriters like Nexus Mutual and is arranged separately. Yields move with borrowing demand rather than being guaranteed. Always make sure to do your own research and be aware of the above and any other risks before depositing.

How do you choose between them?

Start with the calendar your cash actually runs on, because liquidity eliminates options faster than anything else. Work through these in order.

The custodian row is the one that surprises teams most often. Your custodian holds the receipt token, which makes them a party to the arrangement rather than a bystander. Confirm it early with Anchorage, BitGo or whoever holds your assets.

What if you're the platform rather than the treasurer?

The calculation inverts when you're the one offering the product. Your customers hold balances with you, so the question is which engine sits behind the earn tab you want to ship. An embedded treasury API answers that with treasury bills, money market funds and insured sweep. A vault answers it with a programmable mandate that can reach further. Running both is a normal outcome, with the conservative tranche on one rail and the yield tranche on the other.

Both routes remove the build. What differs is the instrument set you inherit and how fast you can add to it. A vault deploys in days to weeks, runs on more than 30 chains, and can hold a strategy that spans money market funds, onchain lending and private credit inside one wrapper. Using a vault inside a treasury product walks through the four patterns firms actually ship, and launching your own vault covers what it takes to stand one up.

As of 4 August 2026 Upshift holds $461 million across its vaults. Peak deposits have passed $550 million across 50+ vaults for more than 60,000 users, as of July 2026.

Frequently asked questions

Where should a company put idle treasury cash?

Match the structure to how soon you need the money rather than to the headline rate. Cash needed within days belongs in a sweep, a managed T-bill ladder or a vault with an instant redemption buffer. Cash you can commit for a known term earns more in a fixed-term fund. Most treasuries end up splitting the balance across two of these rather than picking one.

Is an onchain vault an alternative to a business banking product?

They occupy different layers. A business banking product owns the account, the customer relationship and the front end. A vault is one of the engines a balance can sit in while it earns, alongside brokerage ladders, funds and insured sweep networks. Plenty of fintechs run a TradFi engine and an onchain one side by side for different customer segments.

Can a vault sit behind a treasury API we already use?

Yes, and that's a common shape. An embedded treasury API carries a menu of instruments, and a vault-backed option can be added to it alongside treasury bills, money market funds and insured sweep. The API partner keeps the client relationship and the front end, and takes a share of the fee on the balances it routes. Nothing about it requires the partner to build vault contracts, NAV accounting or withdrawal processing.

How does a tokenised money market fund differ from an onchain vault?

A tokenised money market fund is one fund, tokenised, and its strategy is fixed by its prospectus. A vault is a wrapper whose mandate you define. It can hold a basket of money market funds, an onchain lending position or private credit, with the perimeter enforced by a policy engine. A vault can hold tokenised money market funds inside it.

What yield should you expect on treasury cash?

Short-dated government paper set the reference. The 7-day average across tokenised US Treasury funds was 3.28% as of 4 August 2026 on rwa.xyz. Anything materially above that is taking credit, duration or strategy risk somewhere, and the useful question is where. Yields are variable and never guaranteed.

Do you have to hold crypto to use an onchain vault?

You hold a stablecoin rather than a volatile asset. Most treasury vaults are denominated in USDC, so your exposure is to the stablecoin issuer and the strategy, without directional crypto price risk. Your custodian holds the receipt token the same way it holds other tokens.

How long does it take to get a treasury yield product live?

Onboarding to a fixed-term fund platform or a managed ladder runs in weeks, and the variable is your own counterparty approval. A vault deploys in days to weeks, where the long pole is again your compliance review rather than the technology. Building the equivalent in-house takes quarters.

Can you use more than one of these at once?

Most treasuries do. A common split leaves operating cash in a sweep for immediate access. A committed tranche goes into fixed-term funds for the term premium, with a smaller allocation somewhere running a higher mandate. Splitting also spreads counterparty exposure, which is often the actual reason.

Keep reading

• What an onchain vault is: the wrapper explained for treasury and fund teams, including who holds the assets.
• Using a vault inside a treasury product: the ETP and tokenised fund comparison, fees, and what to decide first.
• Launching a vault for your own clients: the sequence, who holds which keys, and what's hard to change later.
• The DeFi yield strategies onchain vaults run: the strategy shapes, the risk in each, and measured returns on live examples.
• How neobanks and fintechs offer stablecoin yield: the same decision from the platform side of the table.