CH. 20 PURE THEORY OF THE CREDIT CYCLE 321

pect of a rising price-level will tend to evoke hoarding—unless the rate of cost of hoarding is greater thanthe rate at which the price-level is expected to rise—and the prospect of the restoration of the price-level to its previous lower figure will cause stocksto be cleared off. The removal of the no-hoardingassumption is therefore important.

If we assume that the course of the price-level iscorrectly foreseen, a certain proportion of the newbank-credit will be employed not to replenish work-ing capital but to augment liquid hoards, with theresult that the initial rise in the price-level will begreater than would otherwise be the case, because lessliquid goods will be available for consumption. Ifhoarding cost nothing, the price-level would rise toits maximum in the first time-interval (instead of, asin the Standard Case, in the (2 r - l)th time-interval),and this maximum will be the price which can bemaintained without change throughout the 2r -1time-intervals. This price will be half-way betweenthe initial price-level and the maximum price-levelon our old no-hoarding hypothesis, that is to say

hoarding will continue to increase up to the rth time-interval and will diminish thereafter until the hoardshave been completely absorbed in the (2r - l)thinterval, after which, as before, the price-level fallsback again to p.

It will be seen that the possibility of hoarding (ifdone at no cost) halves the amplitude of the pricefluctuation, and is in the general interest since it bringsabout a better distribution of consumption throughtime. For hoarding to be practicable along with theincrease of employment at the rate we have stipulated,

1 For if there were to be a greater rise than this it would only be possibleto sell at a loss in subsequent time-intervals some of the stocks taken off themarket in the first time-interval.

approximately pi 1 +

In this case the volume of

VOL. I

Y