central parkInteresting weather data can be downloaded at WeatherUnderground.

For a counting exploration of weather effects, take the daily data for the Central Park location for 2004 (recent year without a lot of large volatility events in the S&P). Then match the weather data to the daily data for the S&P (requires some cleanup). For each S&P day, calculate the High-Low gap as a % of the Open for each day. Two data columns in the WU data are "Cloud Cover" (0-8), and "Maximum Temperature". Sorting the days of 2004 so that one pulls out the 20 days with highest [Cloud Cover] and then lowest [Maximum Temp], one gets the following stats:

All S&P days for 2004:
mean H-L gap: 0.984%
sd: 0.400%
count: 252 days

20 days with:
Cloud Cover (highest) then Max Temp (lowest)
mean H-L gap: 1.158%
z versus All S&P days: +1.95

Of course, other factors bear on the analysis, but a fun and interesting start.


WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '4445' AND comment_approved = '1' ORDER BY comment_date




Speak your mind


Resources & Links