Your idle cash is probably earning nothing. Here's why nobody does anything about it.

Most commodity traders leave hundreds of thousands sitting in current accounts earning little or no interest. At recent base-rate levels, the gap between that and the overnight deposit rate is money quietly being left on the table.

The problem everyone knows and nobody fixes

Ask any CFO at a commodity trading company whether they're optimising their cash balances, and they'll say something like: "We know we should be doing more."

They're right. At recent base-rate levels of around 5%, a company holding an average surplus of £500K across their accounts could be forgoing on the order of £25K per year in low-risk interest income. That's not an investment return, that's broadly the overnight deposit rate. Money market funds and treasury bills may do better, depending on conditions.

So why doesn't it happen?

Because managing cash sweeps manually is operationally painful. Someone has to:

1. Check balances across every bank account, every morning
2. Look at what payments are going out over the next few days
3. Calculate how much can safely be moved without risking a shortfall
4. Place the deposit or money market instruction
5. Diary the maturity date
6. Recall the funds before the next outgoing payment
7. Repeat across multiple currencies and accounts

For a mid-size trader with 3-4 bank accounts across 2-3 currencies, this is a 30-minute daily exercise that nobody has time for. So the cash sits in the current account, and the CFO tells themselves they'll set up a proper sweep arrangement next quarter.

The maths

Let's run the numbers for a typical agricultural commodity trader. The figures below are illustrative, at a base-rate level of around 5%.

That's one currency account. Now add the EUR operating account with a similar surplus, and the USD account that builds up between commodity settlements. Depending on rates, you could be looking at tens of thousands per year in interest income that's currently earning little or nothing.

For context, that's roughly the cost of a junior analyst. Or the annual subscription to a treasury management system. It can pay for itself before it generates a return.

Why cashflow projections change everything

The reason manual cash sweeps are unreliable is that they require a human to know what's coming. If you don't have accurate cashflow projections, you can't confidently sweep, because you might need that cash tomorrow for a payment you forgot about.

This is where treasury management built on top of trade operations makes a fundamental difference. If your system already knows every open trade, every facility drawdown, every repayment schedule, and every receivable, then it knows what's going out and when. Not as a forecast, but as a structured projection derived from actual committed transactions.

When your cashflow projections are system-generated rather than spreadsheet-estimated, you can sweep with confidence. The system knows:

• What's committed: payments against confirmed purchase orders, facility repayments on schedule, salary runs on fixed dates
• What's expected: receivables due within terms, anticipated settlement dates on open trades
• What's projected: pipeline trades not yet confirmed, seasonal patterns from historical data

The sweep decision uses the "committed" layer: the payments that are definitely happening. Expected and projected flows provide additional context, but the sweep algorithm only acts on certainties. This means you avoid sweeping cash that's needed for a confirmed payment.

Multi-currency sweep

Commodity traders operate across multiple currencies, and each one has a different interest rate environment. Rates vary by currency and over time, and so does the yield opportunity.

An intelligent sweep system analyses each currency account independently. A surplus in one currency might earn the prevailing overnight rate, while the same amount in a higher-rate currency could earn more in a comparable instrument. The system prioritises sweep opportunities by yield, adjusted for the liquidity characteristics and risk of each instrument.

It also considers FX settlement dates. If you have a forward maturing in 3 days that will deliver EUR, and your EUR account is currently surplus, the sweep algorithm accounts for the incoming flow, avoiding a situation where you've swept EUR into a 7-day deposit just before you receive another EUR delivery.

What the board sees

CFOs like this feature because it generates a number they can put on a slide: "Treasury yield on idle cash this year."

It's a clean, defensible metric. The cash was sitting there earning little. Now it earns close to the prevailing low-risk rate, with limited credit risk (overnight deposits at your existing banks), limited duration risk (same-day liquidity), and little operational overhead (automated). It is about as low-effort as yield gets in corporate treasury, though, as with any cash instrument, the return and the residual risks should be reviewed against your own policy.

It also demonstrates to the board that the finance function is actively managed, not just reactive. The cost of the treasury management system can be offset by the yield it generates, helping make the business case for finPhlo partially self-funding.

The catch nobody talks about

Automated sweep arrangements have existed at banks for decades. Most corporate banks offer them. So why aren't more commodity traders using them?

Two reasons:

First, bank sweeps are blind. They sweep based on a fixed threshold: "keep £200K in the current account, sweep everything above." They don't know that you have a £500K payment going out tomorrow. They don't adjust for trade settlement patterns. They don't understand that your outflows spike every second Tuesday when your cocoa supplier invoices are due. The threshold is static, and the cash requirements are dynamic.

Second, bank sweeps don't cross banks. If you have surplus GBP at one bank and a deficit at another, the sweep at the first will happily deposit your surplus into an overnight instrument while you're drawing down a facility elsewhere to cover the shortfall. You can end up earning a low rate on one side and paying a higher one on the other. A system that sees all your accounts, across banks, can identify and eliminate this cross-subsidy before it happens.

Getting started

Intelligent cash sweep in finPhlo builds directly on the cashflow projection module. If your trades, facilities, and bank accounts are connected, the system already has the data to calculate sweep opportunities. Configuration involves:

1. Setting sweep thresholds per account: minimum operating balance, safety buffer percentage
2. Selecting target instruments: which deposit products or money market funds to use, per currency
3. Defining the projection window: how many days of committed outflows to protect (typically 3-7 days)
4. Enabling automatic execution: or manual approval per sweep, for CFOs who prefer to review before executing

The system generates daily sweep recommendations. In automatic mode, it executes them. In manual mode, it surfaces them as a morning notification: "£485K available to sweep from GBP current account. Estimated yield shown. Approve?"