Volume in selloff events

Discussion in 'Trading' started by dmacdonal9, Feb 27, 2002.

  1. I am trading a mechanical EOD system that enters on sharp selloffs that close well below the short term lower bollinger band, in hopes of capturing a bounce. I've added a number of other filters and exit strategies that make it profitable, but nothing I currently do takes volume into account, and that nags at me.

    What sort of volume action during these price shock events might indicate the selloff is over, or has gone too far? I have tried looking for volume spikes, volume declines, volume trend increases, OBV trends, etc, but nothing is better than completely ignoring volume.

    My holding timeframe is only two days, so bear in mind that I'm looking for a bounce, not necessarily a trend reversal. In fact, I'm not even looking for a downtrend as the setup, rather a shock event that should cause a counter-reaction.

    What does everyone think about this? Would you rather enter a very sharp decline after weak or strong volume, and why?
  2. stevet


    anything based on end of day is really just getting you in on a 50/50 bet, and the more indicators you use, thje more they cancel each other out.

    you dont mention your stops for the position, and the chance of you not getting taken out by an early morning move is minimal - so even if you were by luck on the right trend - you could still end up losing
  3. Since you have a profitable end-of-day trading system...why mess with it?

    If your profits are too small...try using intraday analysis the following trading day to improve your entry/exit after you've received those end-of-day chart signals.

    This should improve your profit results.

    Nihaba Ashi
  4. I don't believe it's a 50/50 bet. I've been trading this way for 4 months now and have been profitable every month, which is completely in line with a 5 year backtest. My win% is 58%, with an average P/L of 2.3% and a Sharpe ratio of 3.52. My t-test shows a 9% likliehood of curve-fitting.

    My stops are ATR based, currently 1.5 * ATR(21). That is wide enough to sit through the early morning fades when they happen. Harrowing sometimes, but I know the odds are with me.

    I understand about indicator convergence, and I agree too many are bad. Like Schwager said, indicators are lenses through which we view the same underlying data. Some lenses sharpen our vision, and too many can blur it, IF they're based on the same data. I currently do not use anything that incorporates volume though, and I wonder if I am missing data that I can make use of.

    As to why mess with it, it's gotten as good as it is for me because I constantly look for ways to improve it. Introducing change is only bad if you cannot test the effects of those changes without risking $$. If there's data out there that can improve things that I'm ignoring, then I'm leaving money lying on the table. That's as bad as losing it in my opinion.
  5. stevet end of day analysis 50/50 bet my arse.

    For you maybe, End of day analysis is crucial for determining the medium term trend for inta day trading.
  6. Sounds to me like you already know what your doing but my 2 cents worth is that I definitely like a volume spike for entering on a price spike.
  7. have you tried the tick or the trin?
  8. No, I know trin, but what's tick? How would you use either?
  9. stevet


    you can determine medium term trends as much as you want if that helps you with your trading, but what end of day has to do with medium term i guess depends on how long you consider medium term - personally i want to be aware of the overiding trend which would be indicated by months - not days - and i want to be in at the start of an intra - day trend - thats enough for me to trade off and any more information for me would be pointless
  10. Dmac; I have a similar system, but i use opening prices. I also have looked at volume and found that it has no influence on the results. I found that using the next days volume helped the program, but how do you know what that will be? I am currently looking at volume in the first 1/2 hour of trading and taking positions at that time.
    #10     Feb 27, 2002