The Hidden Mechanism: Why Freed Reserves Can Delay Large Payments

In Colombia’s CUD, the reserve-requirement change is an input, not a timetable. The specific problem is to explain how funds released by a lower requirement can coexist with later settlement of high-value obligations. The mechanism is an allocation choice: a participant can use liquidity to settle now, preserve it for a later need, place it in another short-term use, or wait for an incoming flow. Following those choices shows why greater financial room can produce a later payment profile without implying that institutions suddenly became unable to pay.

The same reserve can serve two uses

An early payment has a clear benefit. It removes an obligation, gives the counterparty final funds, and may allow the recipient to meet its own payment. In a connected RTGS system, that last effect is important: one payment can become the liquidity for the next payment. Early settlement therefore supports coordination and reduces the amount of unfinished business left for the end of the day.

But early settlement also consumes balance. A bank that pays now cannot use the same funds for another obligation that may arrive later. It may need to obtain replacement liquidity, accept a less favorable funding cost, or risk having too little flexibility when an unexpected payment arrives. Waiting has the opposite profile. It preserves optionality and can allow the bank to learn whether an incoming payment or a cheaper funding source will materialize. Its cost is that the obligation remains outstanding, the counterparty may be unable to recycle funds, and the system has less time to resolve an error or shortfall.

A reserve-requirement reduction changes the starting point of that decision. It can release funds previously tied to the regulatory requirement, but it does not determine the use of those funds. If the opportunity cost of leaving liquidity idle at the central bank is high, a participant may prefer to deploy some of the released balance in short-term money-market activity. The institution may still be more liquid in an overall balance-sheet sense, yet choose to submit a particular outgoing payment later because the funds have another valuable use or because the bank expects to receive liquidity from elsewhere.

This is the difference between availability and deployment. A policy can make funds available without making every payment urgent. The timing result depends on the institution’s competing claims on those funds and on the price of waiting. The BIS paper presents redeployment into short-term market operations as one possible explanation for the CUD pattern. The claim is not that every bank followed the same strategy, but that a lower reserve requirement changed the incentives under which payment timing was chosen.

The size of the reserve release and the size of a payment also need not match. A broad reduction in requirements can give many institutions more discretion, while the obligations they face remain lumpy and scheduled around particular trades or settlement events. The extra balance may be enough to clear a set of smaller transfers immediately without being the best source for one large payment that has a different funding date. In aggregate, the system has more room; in the tail of the distribution, the timing of a few large obligations can still dominate the day’s profile.

This is why the mechanism is not a claim about a bank “hoarding” cash in a pejorative sense.