Got a question.
Ive been working with stock prices and gathered data every 5 mins on their returns for 20 days which gave me more than 1300 observations. Viewing its histogram it is leptokurtic (35.2)and was negatively skewed (-2.3).
Now, after doing some conversions, that is sum up the 5 min returns which correlated to one day and did it for 20 days. In this instance the distribution is now normal under the jacque bera test but is slightly negatively skewed (-0.3) and slightly leptokurtic (3.18).
Can anyone explain why? Thanks!