To: Don Maher who wrote (627 ) 6/23/1998 1:40:00 PM From: Bruce A. Bowman Respond to of 1176
Hi Don- I should have suggested this already, so I need to backtrack and make this recommendation: take a look for charts you wish you'd traded if you'd known about them in time. Back up one or more bars and look at indicators you're comfortable with and see what the preconditions are. What you want is consistency in the preconditions. That's what you want the scan to find for you. You may find that what you really want is 2 or more scans. For the list you have here, I see some redundancy in that looking for 3 ESAs moving up and also looking at an MACD will tell you essentially the same thing since the MACD is constructed from a pair of ESAs. OBV is a good screen because it tells you something about buying if you look at the slope. You can also consider money flow, volume accumulation-%, accum/distrib or any of a few others. Phase is an MACD and is available in TF+ only in that form. You already have MACD listed, so it's covered. (Phase uses specific settings of the underlying ESAs. I think it's in the reference manual). So this could be boiled down to: - MACD positive and breaking upward - OBV trending up Since I trimmed your list a little (you may want to not drop the 3-ESAs... I was only illustrating a point about the content of the info), you now would prefer to see MACD positive as well as breaking upward. You also have to decide what period you want to use to look backward when evaluating the trend of the OBV. AIQ uses a 21 day period. They look for OBV making a new 21-d high. You may find this too restrictive an approach to OBV trend and prefer to use OBV > SMA(OBV,21) instead. A somewhat traditional measure of testing trend is to look at the 50-d SMA. It's important to include because many growth stock investors use it. You don't want to be on the wrong side of a trade if it gets very extended and they all head for the woods. Ian Woodward says that a stock which is greater than 25% extended above the 50-d SMA is in the red zone; 15% - 25% is the cautionary (yellow) zone. So a test to assure that the 50-d SMA is moving up (today is greater than yesterday) and that price is less than 125% of the 50-d SMA would loosely fit the trend criteria. To allow a little room, you could use a slightly smaller value than 125%. This gives you the pulse of Ian's followers. As an alternate, you can use +DMI(14) (positive directional movement indicator) being greater than -DMI(14) and ADX(14) greater than 26. I prefer the later for trend but have used both. The reason I prefer ADX/DMI is that the 50-d SMA scan will give a buy when a stock is trading in a range where the ADX/DMI combo will often screen it out. Are you willing to consider stocks that are in the 100/sh or 200/sh range? When I was trading ST I didn't bother with anything that cost more than 30/sh because I couldn't buy a sufficiently large number of shares. Admittedly a %-gain is a %-gain and it doesn't matter what the base number is, but: - as the number of shares get smaller, the commissions become more significant. - a lot of trades will turn against you and getting out with a 1/4 pt gain can mean a profit if there are enough shares (again, this is related to commission) If price is important to you in light of commission impact, then add it to your list of criteria. You might also want to consider volume. Trading stocks with thin volume can be profitable, but it can also be dangerous: - if SOES traders start hitting them, they can completely destroy a trend in a heartbeat. - if you need to get out of a position quickly, you may not be able to. - worst of all, the spreads will kill you. Here's a cautionary note about scans: don't take the list from a scan as trade recommendations. A scan won't give you trades automatically. You still have to look at each chart and satisfy yourself by whatever means you use right now that you have a stock you want to trade. It sounds like you're comfortable with trading stocks you've looked over carefully, so you'll still be doing the same thing. Bruce