A purchasing cap should be derived from a rolling cash-flow forecast rather than picked by feel: list incoming payment timing, committed fixed costs and in-transit inventory over the next months, then work out how much stock can be tied up. Update the forecast weekly and record reasons for variance instead of smoothing the numbers.
Keep a buffer in the cap: supplier lead-time swings, return reversals and FX timing all push actual occupancy above plan. Write the buffer ratio into the metric definition and adjust it by category or sourcing tier — one flat ratio across the whole store tends to fail on one class first.
This article covers the calculation basis for forecasts and caps only; it makes no investment judgement about any supplier or category. Record cash and stock figures as they are; if numbers have not moved, record that too, without pressure to buy more or change direction.