Miller-Orr cash limits: calculate the return point and transfer amounts

A cash control policy can permit the balance to move within a range, then transfer funds when it reaches a boundary. The Miller-Orr model calculates an upper limit and a return point around a lower limit selected by management.

ACCA's specimen solution illustrates the model's control limits and the transfers that restore cash to the return point. The return point is one-third of the spread above the lower limit. It is not halfway between the upper and lower limits.

The model weighs the cost of making transfers against the opportunity cost of holding cash. It is a stylized planning benchmark for variable cash movements. It does not determine which assets a business should hold, create funds that are unavailable, or replace a forecast of known payment obligations.

Assume a hypothetical business sets a minimum available cash balance of $20,000. Each transfer between cash and an already available liquid reserve has a relevant fixed cost of $50. The opportunity cost of holding cash is assumed to be 0.0002 per day, equivalent to 0.02% daily.

Assume the variance of daily net cash flows is 1,000,000 dollars squared. That corresponds to a daily standard deviation of $1,000. The variance is the input required by the formula; entering the $1,000 standard deviation directly would produce the wrong band.

Use a variance and an opportunity-cost rate measured over the same time interval. An annual rate cannot be inserted unchanged alongside a daily cash-flow variance. The daily rate here is supplied directly so the example does not hide an annual-to-daily conversion assumption.

Let z denote the return point's distance above the lower limit. In the basic formula, z is the cube root of 3 × transfer cost × cash-flow variance / (4 × opportunity-cost rate). All inputs must be positive for this calculation.

Substituting the example gives the cube root of 3 × 50 × 1,000,000 / (4 × 0.0002). The quantity inside the root is 187,500,000,000, producing z of approximately $5,723.57.

The return balance is therefore $25,723.57. The full spread is three times z, or approximately $17,170.71, and the upper limit is $37,170.71. Keep full precision internally and round the operational balances according to the policy's practical requirements.

At the lower limit, the modeled replenishment is $5,723.57, taking cash from $20,000 to $25,723.57. That transfer must come from a reserve or another explicitly available source; the arithmetic does not establish that the source can settle immediately.

At the upper limit, the modeled transfer out is $11,447.14, taking cash from $37,170.71 back to $25,723.57. The upper-bound transfer is twice the lower-bound replenishment because the return point lies one-third of the way up the band.

If a discrete receipt moves the observed balance beyond the upper limit to $40,000, returning to the same target would require about $14,276.43. Mechanically transferring the boundary amount of $11,447.14 would leave cash above the target. The action should refer to the observed balance and the policy's actual execution rules.

The input concerns variation in net cash flows over the chosen interval. The variance of cash balance levels is a different statistic because balances accumulate previous flows. Substituting one for the other can substantially change the calculated limits.

Management transfers can also obscure the underlying variation. If every large movement in the historical series includes a deliberate transfer made under the old policy, the resulting variance may describe that policy as much as the operating cash process.

Separate predictable payments and known receipts from the uncertainty the control model is intended to handle. A scheduled annual payment should still appear on its date in the forecast. Treating it solely as another observation in a random-flow estimate can leave the business unprepared for an event it already knows about.

Because z uses a cube root, doubling transfer cost multiplies the distance by the cube root of two, approximately 1.26. Doubling daily standard deviation quadruples variance and multiplies the distance by the cube root of four, approximately 1.59.

A higher opportunity-cost rate narrows the calculated band, other inputs unchanged, because holding cash becomes more expensive within the model. That result does not override a minimum balance needed for actual obligations or a delay in replenishment.

Raising the selected lower limit by $10,000 shifts both the return point and upper limit upward by $10,000 while leaving the calculated spread unchanged. The formula therefore does not validate the floor. Management must establish the floor from its own payment, availability and operating requirements.

A business with sustained net cash consumption will eventually exhaust a finite reserve even if it follows the transfer rules perfectly. Repeated transfers into the operating account do not solve an underlying funding deficit.

Large scheduled flows, persistent trends, correlated movements and delays in transferring funds can make a simple band a poor representation of the actual problem. Model those features directly when material rather than expecting a wider spread to compensate for every departure.

Use available cash in the policy's defined scope. Restricted balances, unsettled receipts and funds held elsewhere may not support a payment from the account being controlled. Combining them into one headline balance can make a mathematically correct trigger operationally ineffective.

Before adopting a band, apply the proposed trigger and settlement rules to a relevant historical or scenario cash sequence. Record transfer frequency, low balances and reserve use. That exercise can expose a practical timing problem that the three calculated balances do not show.

Present the lower-limit rationale, the daily inputs and the resulting actions together. The model is useful when it gives treasury a reproducible way to compare transfer policies while keeping known obligations and real funding availability in view.