Sunday, March 18, 2007

Maybe A Solution

I did some investigation looking for a filter that might give us an up day on the following day if we bought its outputs for an overnight relationship and I found one - maybe.

I say maybe because I don't want folks rushing off trying to get some free money in what I'm considering a down trending market. So when I give this to you - for the sake of your trading account be careful -

It is, of course, a derivation of the BOB set-up filter - but it has a twist - to give it some oomph I've added a modified ATR/RSI filter and one additional requirement - that the second day of the the BOB set-up volume be greater than the average volume. Here is the filter -

/*This filter is to be run late in the day to find potential overnight trades*/
show stocks where close is between 15 and 35
and average volume(90) > 500000
and volume > average volume(90)
and close 1 day ago < ema(8) 1 day ago
and close 1 day ago < open 1 day ago
and close < open
and high < high 1 day ago
and volume is more than 20% > volume 1 day ago
and low < low 1 day ago
and average true range(10) < cma(average true range(10),10)
and rsi(2) < 5


Note the change in the RSI to a requirement to be less than 5 instead of 2.

Here are some results -



Yes those are all of the selections (there were none Thursday) and yes they were all positive the next day. Now if you lose your butt playing overnight don't blame me.

Keep in mind that if the stock gaps down on you it is possible that it will come back but if it keeps going your best approach is to get out. On a hard gap down no stop loss order is going to save your money. It's gone.

This is a very dangerous play - be very careful if you try it.

This filter tests poorly in the current market so I wouldn't hold these stocks very long. It tests just fine in an uptrending market (figure that).

4 comments:

Marlyn Trades said...

In a sell-off volume increases on the second day - when buying is done volume decreases on the second day.

Pull up any chart over a couple of months and look at the volume spike relationship to the next day where you find two days down or two days up.

Is this absolute? No of course not - it would be too easy that way - but most of the time it is that easy.

slotmonkey said...

The backtest I ran for the most recent 2 year period from 3/16/05 to 3/16//07 returned 51% winners. Holding period is for one day with a 5% stop loss, produced annualized ROI of 43%. Unless I'm not backtesting it correctly, it's really no better than flipping a coin. Please enlighten me if I'm wrong.

Keep them coming though.

slotmonkey

Anonymous said...

11 conditions(!) is idiotic. Ever hear of 'degrees of freedom' ? After testing trading systems for 16 years I can only laugh and then shudder when I read something this ridiculous. Dangerous indeed.

QUALITY STOCKS UNDER 5 DOLLARS said...

And maybe not.