324

A TREATISE ON MONEY

BK. IV

For, as more and more commodities reach their peak anddecline, the rate of increase of the composite price-levelwill be, in the case of Fig. 1, slower and slower, until apoint comes when the commodities of which the price hasdeclined outweigh those of which the price is still rising.

There is also a further qualification to our previousconclusions, if we relax our assumption as to uniformlength of process but not our assumption as to perfectforecasting. For if the rate of input of a commodityhaving a short production-period is increased in thesame proportion as that of everything else, its pricewill fall below the normal at the end of its production-period, because the purchasing power available for itwill not yet have increased in as large a proportionas its output. It follows that the rate of input ofcommodities having a relatively short production-period should be increased somewhat more slowly thanthat of the others. On the other hand, as a partiallybalancing factor to the above, there will be somediversion of consumption from articles of long pro-duction-period to articles of short production-period,to take advantage of the lower price of the latter.

(6) Assumption Zeta.

Any departure from this assumption has the sameeffect as a variation of the rate of saving already con-sidered with reference to Assumption Alpha.

( 7 ) Assumption Eta.

In practice Credit Cycles have a strong inherenttendency to be “ overdone ” as compared with theStandard Case.

So far we have presumed accurate forecasting andhave only made a few passing allusions to the resultsof imperfect forecasting. But in fact forecasting iscertain to be imperfect, and likely, moreover, in thepresent state of ignorance, to have a bias in onedirection. For, with the existing mentality of entre-