← All Insights

Uber Shows Why Cash Management Is Moving On-Chain

FintechOffice of the CFO

Uber has demonstrated something more important than a faster payment. It showed how a corporate treasury could keep cash invested until the moment it needs to pay someone, including outside market hours.

In an October 30 pilot, Uber used HSBC tokenized deposits to buy Franklin Templeton tokenized money market fund shares through Tradeweb. The purchase and settlement happened together on Canton, with Fireblocks providing wallet infrastructure. The group also demonstrated liquidating a portion of those fund holdings over a weekend to access cash for payments. This was a pilot, not a live global cash-management product. But the operating model is clear.

I put the step-by-step mechanics of the pilot, who moved what and where the extra yield came from, in a two-page paper: Uber’s on-chain treasury pilot.

The headline is being described as a stablecoin story. Technically, that is wrong. The cash leg was a tokenized bank deposit, and the yield-bearing asset was a tokenized money market fund. A payment stablecoin is a third instrument. Calling all three a stablecoin obscures the part that creates the economic value: cash can move between a bank claim and an investment claim without waiting for separate institutions to open, trade, settle and reconcile.

After raising a Series B, I learned how crude treasury management could be. Our bank could not support Treasury purchases, so I opened an account elsewhere. Building a ladder meant phone calls, email chains and manual trade confirmations several times a month, every month. Settlement took days. The process stopped for the weekend.

That experience is why the Uber pilot matters to me. A fund share is an investment. A tokenized deposit is a bank liability. Neither one, by itself, is a completed payment. The value appears when a treasury team can move between them under its existing permissions without rebuilding that phone-and-email workflow each time cash needs to move.

That is the metric I would put on a CFO dashboard: uninvested dollar-hours caused by the operating workflow, not token volume.

The cost is the cash you do not invest

Corporate treasury has long managed a tradeoff between yield and immediate access. Leave too much cash in an operating account and it sits idle. Invest too much and a payment may require a redemption, a settlement window and a transfer before funds are usable. To manage that uncertainty, a company holds a buffer. The bigger and more international the company, the more places that buffer can accumulate.

The math is not mysterious. At a hypothetical 4% annual yield, one additional day invested on $100 million is about $11,000 before fees and risk. That is an illustration, not a claim about Uber’s balances, returns or pilot results. The more important number for a CFO is the total idle balance that can be released without increasing the chance of a missed payment.

On-chain infrastructure changes the sequence. The treasury system can see an obligation, liquidate only the amount required, exchange the investment claim for bank money in a synchronized transaction and release that money through an approved payment route. The rest stays invested. The participants demonstrated a weekend liquidation of eligible holdings. It does not make every recipient bank or payment rail available on Sunday night. The useful question is how much of the buffer exists because an investment cannot be converted to usable bank money when the payment obligation arises. A treasury team can answer it by matching scheduled obligations to fund liquidity limits, available payment destinations, entity permissions and compliance approvals, then retaining the cash that still cannot be released safely.

That is a much larger proposition than making a payment a few seconds faster.

I saw the other side of this problem in B2B disbursements. Payments moved through checks, cards, ACH and wires, each on a different clock. A mailed check could take roughly 10 days to reach its recipient. I spent time exploring bank arrangements that could put funds to work during the period before a payment was due. That is where yield, payment speed and international reach converge. An on-chain product that only turns a Monday payment into a Sunday payment is incomplete. The opportunity is to keep funds productive until they are needed, then make them available through the right payment route.

The operating system for cash

My Treasury ladder had two costs. One was the time spent coordinating trades by phone and email. The other was the cash left in an operating account because accessing an investment again would take days. Automating a trade is useful. Eliminating the need to choose between earning and being ready to pay is the larger change.

That is why I would measure uninvested dollar-hours before and after deployment. If an on-chain treasury product leaves the cash buffer unchanged, it has not solved the problem, however fast its tokens move. If implementation and control costs exceed the recovered yield and operating savings, the business case fails too.

The architecture has two jobs. A tokenized fund holds the investment and earns the yield. A tokenized deposit, or in another design a regulated payment stablecoin, moves the cash claim. Uber used the deposit, not a stablecoin. Permissions, fund liquidity and the recipient’s payment route have to work around the conversion. A ledger does not make a bank, a fund or a wallet safe by itself; controls over identity, approvals and endpoints do the work.

HSBC’s Uber pilot joined those two layers. BNY has separately demonstrated tokenized deposits and a Treasury collateral workflow. By my measure, those are the important building blocks: regulated money linked to an asset that can be put to work. Neither announcement proves that a company can retire its cash buffer in production.

The closed-loop economics emerging in B2B payments point to the same prize: control the movement between a balance and its use. And as with the financial interface, the treasury dashboard has value only when the money movement and controls beneath it work.

I expect basic cash management to move on-chain because the old alternative is a bank portal, a brokerage account, a spreadsheet and a recurring chain of calls and confirmations. A treasury team should be able to keep funds invested until an obligation is due, convert only what it needs and send through an approved route. The test is not a token balance or a weekend demo. It is fewer idle dollar-hours without a missed payment.

That is the product.

Sources