After Cash Flow Stabilises, Where Does the Money Go?

A stable cash flow is something many companies spend years working towards. But stability often brings an awkward companion: the money on the books keeps growing while the interest it earns has not moved at all.

The reason is usually not that the company neglects treasury management. It is that nobody has layered the capital. All the money sits in one account and is treated with one set of instruments. The result: the part that could genuinely be used for yield management is held down by margin and working capital needs, while the yield opportunities that would accept a little term risk go unused because nobody wants to disturb overall liquidity.

This article is about exactly that: how to dig out one more layer of yield while the cash flow stays stable.

The first step is not to buy a product — it is to split the money into three layers

Most treasury failures we see begin with a single action: someone sees an attractive yield on a product and puts a sum of money into it.

The problem is that before the purchase nobody answered a basic question: when might this money be needed?

The correct order runs the other way — layer first, then allocate.

Layer 1: money that cannot be moved

Futures margin, letter-of-credit margin, bill margin, prepaid purchase payments. The only property of this money is that it can be called or swept at any time.

There is only one correct action for yield management on this money: keep it in a separate account from operating funds and earn slightly more than the demand-deposit rate without disturbing transfers. Putting it into any product with a term trades operating safety for yield — that trade never makes sense.

One mistake recurs. To lift the yield on the margin account, a company moves the money into a wealth product; then a day arrives when the market turns against it, a margin call comes, the money cannot be withdrawn, and the position has to be closed out under duress. Margin held for hedging was turned into capital for speculation: defensive money became offensive money.

Layer 2: money for short use

Daily working capital, payables falling due within 30 days, and the funding gap behind receivables that have not been collected. This layer has a single requirement: available on demand.

So only the most liquid instruments belong here, even if the yield is poor. Trading yield for liquidity is worth it here, because liquidity has a price: one forced borrowing at a low rate, or one rushed factoring, often costs more than a full year of rate spread.

Layer 3: money that can be allocated

Everything beyond the first two layers with no definite use in the foreseeable period. Only this layer is an object of yield management.

Typical sources: net cash generated by operations, over-raised shareholder funds, dividends held pending payment, and earmarked project money not yet deployed.

This layer is usually the largest and the most easily ignored — on the financial statements it is one line inside cash, and no mechanism exists to track how much of it there is or how long it can stay put.

A worked example you can compute directly

Say a company holds RMB 200 million. Layer it once:

• RMB 30 million is futures and letter-of-credit margin → Layer 1, cannot be moved;
• RMB 70 million is working capital needed within 30 days → Layer 2, available on demand;
• the remaining RMB 100 million is investable cash → Layer 3.

If all RMB 200 million sits in a demand-deposit account at 0.2%, annual interest is about RMB 400,000. Match the RMB 100 million in Layer 3 to a three-month tenor at 1.5%, and place the other two layers in a negotiated deposit and a money market fund, and the blended annualised yield reaches about 0.9% — roughly RMB 1.8 million of interest.

The difference is RMB 1.4 million. For a trading company with RMB 1 billion of revenue and a 2% net margin, that is 70% of a full year's net profit.

That is why treasury management in a mature company is not a finance trick. It is a profit lever of the same order as hedging.

The tool ladder: trading liquidity for yield

How to read the table: match it layer by layer, from the top down. Layer 1 uses the top row; Layer 2 uses the top two rows; only Layer 3 reaches the lower ones. Do not skip layers. Short money does not go long and long money is not used short — better to earn less than to lock money up.

A specific warning on FX products and rate swaps

Companies holding USD balances will come across FX wealth management products and interest rate swaps. They offer a higher nominal yield than RMB products — and FX risk sits inside them.

Specifically: the yield is FX volatility repackaged. If the company is already net long USD (the usual position in transshipment trade), adding an FX product stacks a second directional exposure on top of an existing one.

The test is simple: where does this yield come from? If the part above the equivalent RMB product is still positive after deducting the historical range of FX movement, it is real yield. Otherwise you have not removed a risk — you have swapped a known FX exposure for an unfamiliar one.

Two hard boundaries that must hold

1. Never move margin. That includes moving futures margin or letter-of-credit margin into wealth products. There are no exceptions, and "just a temporary bridge" is not accepted. The worst case is simple: the money is not available on the one day you need it most.
2. Never exceed the authorised limit. Every investment must sit inside the scope and the limits approved by the board or the shareholders' meeting. The most common cause of a treasury incident is not a loss — it is an amount or a tenor outside the mandate, which turns a treasury problem into a compliance problem.

Work out how long you can go without it before you talk about what you can earn

This is the discipline we give every company, and it is the step most often skipped.

The first question in yield management is not "how much can this product earn". It is: if this money cannot be withdrawn for three months, does the business get into trouble?

If the answer is yes, it does not belong in this layer. The essence of yield management is finding yield inside a defined safety boundary, not gambling liquidity in pursuit of the highest yield.

One more thing has to be computed at the same time: the link between the cost of capital and the cost of hedging. Article 14 covered this — forwards and options both consume credit lines and margin, and that capital carries a cost. So any combined hedging and treasury plan needs a net contribution calculation:

Net contribution = locked-in yield − premium − funding cost − management cost

If the numbers do not work, however professional the plan looks on paper, it should be redesigned.

Four things to put in place

1. Build a capital layering table (1 week). By account, by purpose and by available term, put every fund into one of the three layers. This is the basis for everything that follows.
2. Physically separate margin accounts from operating accounts (2 weeks). This step earns nothing on its own, but it removes the most dangerous category of risk.
3. Term-match Layer 3 (1–3 months). Match the term ladder to the foreseeable timing of withdrawals, instead of buying one product with a single uniform tenor.
4. Establish a monthly review (ongoing). Look at three things: the gap between actual and expected yield, whether any liquidity event has occurred, and the impact of funding cost on the net benefit of hedging.

A self-check you can run today

1. Can you say, in ten minutes, how much money is on the books, how much is margin, how much is needed within 30 days, and how much can be placed for three months?
2. Are the margin account and the operating account separate? Has there ever been a case of "borrowing from the margin account for a temporary working-capital gap"?
3. If you term-matched the investable portion once, how much more do you estimate you would get in a year? As a share of your full-year net profit, how much is that?

If you cannot answer the first question, the other two cannot even be asked. Without knowing how the money is layered, there is no managing it — and certainly no earning another layer from it.

Sources and notes

The three-layer capital framework and the tool ladder used in this article are RMC's standard method in corporate capital structure diagnostics and advisory work. The interest rate ranges and example figures cited are market environment references at the time of writing; they do not represent the actual financial position of any specific company and are not specific investment advice. Each company should determine its own position from its own capital characteristics, authorisation scope and liquidity constraints. Nothing in this article constitutes a recommendation or evaluation of any specific software or investment product.