Hi,
Sorry to be troublesome.
I've been studying data and running tests for a long time because I was never getting an accumulated volume count to come anywhere near the numbers reported by PSS and the exchanges (the volume I'm referring to is the total volume so far during the day). For busier symbols my count might be off by a factor of ten or twenty. The end result is that I suspect PSS ignores all but the last trade that occurs within the same second.
I ran a test Dec. 20 to write out selected data to a file for every stock in the portfolio (at that time, 200 stocks). The data is: Symbol, LastTradeTime (the time reported by the exchange for that trade), CurrDateTime (the time I write this to the file -- can be delayed by some seconds), the Price (just for double checking), PrevVolume (I store the volume reported with each trade in a database - on the next trade I retrieve that volumne and it becomes PrevVolume), LastSize (the number of shares in the current trade), Volume (the accumulated daily volume reported by the exchange). I put this into a spreadsheet and totalize the LastSize column and calculate a column "Missing Trades". PrevVolume plus LastSize should equal Volume, if it doesn't, the difference is MissingTrades.
Of the 200 stocks, I just picked the first two in the portfolio: AA (heavily traded) and AAWW (lightly traded). As reported in the spreadsheet, the majority of the AA trades are missing. AAWW is better, but still much data is missing.
To eliminate the possibility that the system was just too busy, I repeated the test on Dec. 29 (a light day) with only those two stocks in the portfolio. Results were the same. I tried to attach that spreadsheet to this entry - but it didn't work. I can supply it to you if you are interested.
In any case, I've given this a thorough test, and I think it bears your investigation.
thank you
