WHAT MAKES THE PRICE OF OATS 21 
supply of oats apparent in September has become diminished some- 
what, for it is measured by a coefficient of —0.87. 
The correlation between visible supply plus stocks on farms March 
1 and the May cash price is —0.85, whereas consumption to March 1 
seems to maintain a practically constant relationship to production — 
the correlation between final estimate of production plus carry over 
and the stocks on farms plus visible supply March 1 is +0.97. Thus 
the September price of May futures, although closely in line with 
production and carry over at the end of the harvest, may be some- 
what out of line with the cash price which has come to prevail in 
May. 
As between May and the following September a somewhat similar 
but more complex condition exists. For a period of 18 typical years, 
1896 to 1913, inclusive, the May prices of September futures were in 
line neither with the quantity of new oats yet to be matured and 
harvested nor with the September cash prices. The correlation 
between May prices of September futures and final estimates of the 
new crop is measured by a coefficient of + 0.31. 21 Instead of showing 
the negative relationship usually assumed between production and 
price, this coefficient indicates a slight tendency for May prices of 
September futures to be high when the new crop is large, and vice 
versa. 22 It suggests, therefore, that the size of the new crop has no 
significant effect as early as May upon prices of contracts for future 
delivery. May prices of September futures appear in fact to be 
influenced more by prevailing cash prices than by any other factor. 
As the season progresses, however, and as forecasts of the new 
crop become more and more accurate, prices of September futures 
are gradually adjusted toward the average September cash price, 
until in August the average price of September futures corresponds 
very closely with the September cash price. The closeness of agree- 
ment at this time is measured by a coefficient of +0.95. The relation 
between the price of September futures and the August cash price is 
measured by a smaller coefficient, +0.86, showing that the September 
cash price can be predicted more accurately from the August price of 
September futures than from the August cash price. 
Table 11 shows the gradually increasing closeness of the relation- 
ship between cash and future prices as the harvest season approaches. 
The variables in these correlations are expressed in terms of changes 
in cents per bushel over the corresponding months of preceding years. 
These coefficients indicate the increasing accuracy with which Sep- 
tember cash prices can be estimated from the prices of September 
futures. 
RELATION BETWEEN CROP CONDITION ESTIMATES AND PRODUCTION 
In recent years the United States Department of Agriculture has 
published monthly forecasts of production for the principal grains, 
including oats, during the latter part of the growing season. These 
are available for only a part of the period covered in this study, but 
21 The correlation between corresponding changes in the May price of September futures and the Sep- 
tember cash price is measured by a coefficient of —0.03. 
22 This positive correlation is probably due to some tendency for a large crop to be followed by a small 
one, and vice versa, and not to any real relation between the May price of September futures and the size 
of the new crop. Since the May price of September futures is influenced more by the size of the preceding 
crop than by any other factor, tne actual relation measured by the correlation coefficient is that between 
the crop of one year and a price based on the crop of the preceding year. 
