Ask Colin

I read in ATAA Journal that some data are padded but I do not know which is better or does is really matter much?

First, what is that "data padding". I understand that some data vendors insert the last closing price for a day where a security does not trade. The volume will be shown as zero, and Open, High, Low and Close shown as the last closing price. Also, some charting software packages will do this same thing where there is no data for a day.

Second, is it a good idea? Daryl Guppy has written at length about it and tried to justify it. I will have to leave you to make a judgement on his arguments. However, my view is totally opposed. In a nutshell, I trade what happens, not some synthetic data. If a market does not trade, it does not trade, notionally or otherwise. To include some padded data on the basis that the market would have traded there or that it notionally accepted that price is simply unrealistic to me. Just consider the example where a market last traded at 250c. If some disaster struck and the spread was buyer 150c, seller 210c and no trades took place, what is the possible realism in using 250c as the price for that no-trade day when calculating an indicator?

However, this argument is really so much hot air way off the whole point. Indicators are meant to be used on deep liquid markets that trade all the time. They are of doubtful use in markets that are thinly traded and have days when they don't trade. My feeling is that they should not be used in such a situation.