I think I may have misspoken. The modified VTO system I was referring to was the one *I* modified, by setting the upper limit to 60 and also adding the declining volume confirmation. I think it is important not to use volume in an absolute sense, so what I do is compare that day's volume to the average over the previous 5 days, so it is more accurate with respect to the most recent trading volume, and about -15% looks to be about right to signal declining buying interest. It doesn't always catch the top, but it doesn't let it get too far. It correctly signalled the turn on three occasions, 1/13, 1/30, and 2/27, but then also signalled a sell/short on 3/15, 3/20, and 3/31, and nowhere a buy/cover in sight in the interim since the last three, so it missed this latest breakout. I should probably back test it more now that I have the NDX data back to 1985 from TC. The problem with the volume confirmation is that you can only implement the signal on a next day basis. I haven't yet figured out how to trigger an order on the condition that the volume is less than some predetermined amount, since it will be less for the whole day until it passes it. However, as a day trader, you could track it hour by hour, but I think there are more effective strategies than this for day trading, like Nocona's.
Market Data 
Markets 
AI
