This [unedited] guest post is by a student in my PSTAT262MC class (background post). Please praise/critique/comment on its quality and importance to you.

varvara [thumbnails].jpgVarvara Kulikova says:

High frequency trading data represent intraday trading activity for an equity on the exchange market. The data set for the analysis had been obtained from the TAQ database of the NYSE via Wharton Research Data Services located at http://wrds.wharton.upenn.edu.

The information available in the data is a record of each transaction occurring at NYSE during regular trading hours, from 9:30 a.m. to 4 p.m. EST. This information includes a time stamp at which each trade had occurred (up to a second), a corresponding price (number of dollars per share) and volume (in number of shares). The trading data had been obtained for three selected stocks: DELL, IBM and Microsoft (MSFT) on October 1, 2009.

A figure below illustrates average price at which trades occur at a particular second and total volume of these trades for three stocks on this day.

The day had been split up into one-second intervals, which correspond to 27,401 time points per trading day. Each of those time points was matched with the time points for average trades for all three stocks. Some of these intervals have no trades for any stocks, some have trades for some stocks, and some have trades for all three stocks.

My forecast (68% interval) for the opening average price of stocks (DELL,IBM,MSFT) in the morning (9:30:00) of next day October 2, 2009 would be 15.1(15, 15.2), 118.8(118.6,119) and 24.87 (24.77,24.97) respectively. Also on October 2, 2009 around noon (12:30 pm EST) the forecast for mean prices would be 14.8(14.7, 14.9), 117.2(117,117.4) and 24.42 (24.32,24.52)


Stockprices_volume_regscale.jpg

blog comments powered by Disqus

Published

06 March 2010

Tags