Mar

1

The Morningstar stats are:

years: 1945-2010 (count: 66)
mean S&P return: +9.08%
mean S&P return, 3rd year of prez cycle: +17.15%

Setting up a simulation that, for each run, randomly resorts the set of actual annual S&P returns among the actual years and then recalculates the mean return for the set of "3rd year of prez cycle" years:

simulation resorts: 1000
mean return of all simulated "3rd year of prez cycle" sets: +9.07%
SD of the set of means from 1000 resorts: 3.78%
z of actual "3rd year of prez cycle" mean (+17.15%) versus simulated set: +2.14

Victor Niederhoffer writes: 

Wouldn't that have to be adjusted by by a factor of slightly less than 3 or 6 to take account of the fact that each year might be the best or worst, and another factor of 5 or so to take account of different markets like fixed income or grains, and another factor taking into account when they started the years from, thus reducing the 1 in 100 to say 8 in 9 in favor of finding a difference as big as this. 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search