292 THE STATISTICAL VERIFICATION OF 
Long Term Investments—what he describes as the record 
of the failure of facts to support a theory. The theory in 
question is that, during the period when the prices of goods 
and services are falling, bonds are betterinvestments than 
ordinary shares. He traces the effect of investments and 
holdings over the period of seventeen to twenty-two years 
and finds that out of twelve tests the investor would have a 
much higher investment return from ordinary shares than 
from bonds. In every case except one, the advantage is 
on the side of ordinary shares, when income and increase 
in capital value are added together. His investigation 
is well worth following. The broad conclusion that he 
arrives at is that, cumulatively, the results tend to show 
that well diversified lists of common stocks, selected on 
simple and broad principles, respond to some underlying 
factor which gives them a margin of advantage over high- 
grade bonds for long-term investment. There are many 
reasons why this particular investigation has no universal 
validity, but it is an interesting example of the verifica- 
tion of theory. A similar test quoted by Mr. Hartley 
Withers for England, though much more limited, tends 
in the direction of confirming it, and Mr. Hartley Withers 
concludes that it has been shown that ordinary shares 
have advantages which make it impossible to regard 
them as necessarily so speculative that we ought to feel 
¢ rather ashamed of possessing them ’.1 
Contributing subscribers to the National Bureau of 
Economic Research, New York, approve the develop- 
ment of ¢ an organization devoted to exact and impartial 
investigations in the field of economic, social, and indus- 
trial science >. The most recent inquiry is entitled Busi- 
ness Annals and gives an analysis of business records of 
seventeen countries for periods ranging in some instances 
as far back as 1790. In many cases these are based on 
description, in others on statistics. A conspectus of busi- 
ness cycles is possible, giving five conventional values 
1 Hints about Investments, Ch. XII.