| StockFetcher Forums · Filter Exchange · gap up size relative to the average candle size | << >>Post Follow-up |
| ImperialWhazoo 26 posts msg #34414 - Ignore ImperialWhazoo |
12/17/2004 11:44:23 PM I'm looking for a simple way to decide the following: 1.) Does anybody know how to figure out the average candlebody size over a range of days? 2.) Does anybody know how to compare the size of the white candle on a "gap up on open day" to the average size of the candle bodees over the preceeding days? I would test for 6 preeceeding days up to 22 preeceding days. I've been looking aat this for a while and I want to test my non-scientific observation that there seems to be a relationship between the average size of a set of candles on the days preeceeding the "gap up on open" day. The only way I can see to approach this is to find sideways trading day patterns that last a week or more, measure their average candle body size and then generate a Stockfetcher results list with an added column that compares the gap size to the average candle body size. There may be a relationship or there may not. If there is, then all I need to do is run a Stockfetcher filter and then use that list of candidate stocks to scan on open on my trading platform. If a stock gaps up a certain amount and it was on my list, I could buy with confidence that I was getting in near the open of a big white candle day. The point is to determine if there is a pattern to the preeceeding days candle body size that informs my buying decision when a gap up on open of a certain size occurs the next day. Any ides on this would be welcome. Thanks ahead of time. Imperial Whazoo |
| cegis 235 posts msg #34444 - Ignore cegis |
12/20/2004 3:18:43 PM ImperialWhazoo, I usually like to see some effort on the part of the poster to figure out what they're looking for, but since you described your strategy pretty completely, I'm inclined to help... 1) Average body size: set{bodysize, abs(open - close)} /* This will always be positive. */ set{avgbodysize, CMA(bodysize,5)} You need to use abs(), or else the CMA won't give you what (I think) you're looking for... (I assume you want red body sizeto average the same way as a white body size; i.e., a body of -0.10 is the same "size" as one of +0.10...) 2) White candle on Gap Up on Open: set{gapup, count(open - close 1 day ago > 0,1)} set{white, count(open < close,1)} set{whitecandle, white * abs(close - open)} set{WCgap, whitecandle * gapup} WCgap > 0 /* only select white candle gap up stocks */ add column WCgap / CMA(bodysize,5) add column WCgap / CMA(bodysize,6) add column WCgap / CMA(bodysize,7) add column WCgap / CMA(bodysize,8) /* etc... */ HTH, C |
| bka58 49 posts msg #34452 - Ignore bka58 |
12/21/2004 2:58:44 PM I tried the following variation to what you mentioned. The problem I am seeing is that the values returned for gapup1 and gapup2 are not actually the differences. Look at LU as an example. Please let me know if I have coded something incorrectly. set{gapup1, abs(low - high 1 day ago)} set{gapup2, abs(high 1 day ago - low)} low is above high 1 day ago and gapup1 is above .10 add column gapup1 add column gapup2 add column low add column high 1 day ago |
| ImperialWhazoo 26 posts msg #34461 - Ignore ImperialWhazoo |
12/22/2004 12:02:55 PM Thanx cegis As you seem to have surmised, it would be inaccurate to characterize what I'm asking about as a strategy as of yet. At the moment, I'm looking at a backtesting idea that I can fine tune (or reject, whatever the results prove to be... so be it). I need to construct a Stockfetcher filter (or filter set) to get this accomplished. So, what I did in my original post was describe the general thing I'm looking for and to ask for ideas from the community that may be different from some I've thought of myself in hopes that I'd find somebody out there who was already thinking as I had in mind or in hopes of finding somebody who has already implemented a set of similar backtest ideas. That way, if they had failed, I might avoid wasting a bunch of time and if they had been even partially successful, I might end up pointed in a slightly different direction... maybe, combining my own ideas with somebody else's might generate something better than either had generated alone. If and when there is a real strategy springing forth here, then I can give the community the benefit of it. At present, its just a general idea of a type of experimental backtest I'd like to try to figure out how to do. That having been said, here where the idea came from. Over the last month or so, I've begun listening over the internet every afternoon to a guy named Marc Mandel at www.winningonwallstreet.com. He has a semi-regular guest by the name of David Elliott. Now this guy seems to make a pretty solid case for identifying a chart pattern he calls the "Bull Kicker". A fingerprint of this pattern is the thing I'm looking for. Finding a "Bull Kicker" after the fact is easy. I've already created a whole bunch of Stockfetcher filters that identify them. I could shoot them out to the community for comment and/or use by the community but I think they are essentially useless as a uniform predictive tool for buying the BK move because all the do at present is find the BK after it has occurred (EOD), so I'm not bothering to use them myself or publish them for community comment or use. The reason is that the bulk of a BK move occurs the first day. What I want to do is create a Stockfetcher filter (or set of filters) that finds them in a predictive way. Hence, my query regarding measuring the average size of the candles leading up to a BK event. I've been backtesting thru the filters I've created to find BK events historically and making generalized observation about them. My first idea on approaching the creation of a BK filter set is that the most fruitful results will flow from cases where there is a small variance in average candlestick size in the days leading up to the BK event. In essence, I'm starting by throwing out the BK events where the charts involve stocks with a wide variance in average cndlestick size. This was a logical though admittedly inobjective (non-scientific) starting place, but it creates a subset of BK event stocks that have uniformity as a dominant charcteristic. So, that is why I'm looking for a way to test the average candlestick size over the days leading up to the BK event. I how this helps you, and anyone else bothering with delving into my idea here, get a better toe-hold on the steep rockface I'm going to be attempting to scale. Thanx again & Merry Christmas Imperial Whazoo |
| StockFetcher Forums · Filter Exchange · gap up size relative to the average candle size | << >>Post Follow-up |
Copyright 2022 - Vestyl Software L.L.C.•Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data
Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus