Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.


Paul
Posts:
740
Registered:
7/12/10


Discriminating between cyclical tendencies and random walk behaviour
Posted:
Feb 8, 2013 12:04 PM


Two quants look at the performance of a stock over time. A thinks that the stock growth and loss exhibits clear evidence of regular cyclical patterns, saying things like "It looks so much like a sine curve." B claims that, if you take a completely random walk, it can also exhibit such cyclical patterns, and that there is no evidence of statistically significant cyclical behaviour. What are the standard tools for resolving the debate about if, and to what extent, the growth pattern is "more cyclical" than would be expected under random walk hypotheses?
Thank you,
Paul Epstein



