| Welcome Guest, please sign in to participate in a discussion. | Search | Active Topics | |
Registered User
Joined: 5/26/2005
Posts: 11
|
I copied the list of Boolian? TC formulas for several candle patterns and the one for Belthold is not
getting the right results....Can you forward the correct formula? Thanks Wendy
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
I am not a candlestick guy (at all). One thing I have noticed however is that different sources have different definitions of patterns with the same name. I am quite confident that when Craig created the formula, the results matched the definition of the source he was using.
If you have a different definition in mind, you will need to provide an unambiguous objective description of the pattern in order for me to help you create a formula for your particular idea of what a Belthold Candle Pattern should be.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 5/26/2005
Posts: 11
|
A belthold pattern is a specific candle pattern where there has been a downtrend and then there is a gap down, with a bullish candle by end of day. I am not sure of how big the up move has to be and it may not matter. The formula on your posted list is O=L and C>O. It does not sort for the pattern that I can tell. The pattern should also be able to be done in the reverse direction.
Can you refer this question on to one of the Candle pattern people? Thanks, Wendy
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
While your idea of what constitutes a Bullish Belt Hold pattern might be specific, an internet search will reveal that the actual term itself is not. It's quite ambiguous actually. There are almost as many definitions as there are sites attempting to describe the pattern.
Some sites have single candle definitions. Some sites require multiple candles to define the pattern with the addition of a gap down, several down candles before the Belt Hold, a downtrend before the Belt Hold or possibly even follow-up verification after the Belt Hold.
Even when the surrounding candles are ignored, the only thing all of the sites seem to agree on is that the Bullish Belt Hold candle must Close above its Open. Most require the Open to be the Low of the Day, but many only want a short lower wick. Most seem to want the Close to be near the High of the day with a short upper wick, but there are sites with the requirement that the Close actually be the High of the Day.
The definition used by Craig was a Hollow Candle (or White, Green, Up or Bullish depending on what candlestick practitioner is making the description) that Opened at its Low. There are sites where these are the only requirements given and most sites do agree that these are requirements of a Bullish Belt Hold (but many of those also add additional requirements of some sort - even while disagreeing as to what these additional requirements might be):
O = L AND C > O
You do appear to want a Bullish Candle:
C > O
You also want a Gap Down. There are quite a few ways to define a Gap Down, but the examples used in most of the definitions I found online with a Gap Down requirement seemed to show the Open of the Bullish Belt Hold candle being below the Low of the previous Candle:
C > O AND O < L1
You also wanted this to be at the end of a downtrend (but didn't define what you see as a downtrend). Several of the examples requiring a downtrend as part of the definition seem to show a couple of down candles (with both negative Net Change and the Close below the Open) prior to the Bullish Belt Hold candle. Adding that to the above results in:
C > O AND O < L1 AND C1 < C2 AND C1 < O1 AND C2 < C3 AND C2 < O2
Most of the definitions I found online did require the Open of the Bullish Belt Hold candle to be the Low of the day. That said, the presence of this requirement in Craig's version did not seem to be of interest to you and your description did not mention it as a characteristic of a Bullish Belt Hold candle. Your lack of interest in this pattern notwithstanding, I am not finding definitions that do not require the Open to be at least "near" the Low of the Day. We'll only loosely define near as being less than 20% from the bottom (although I'm guessing this should be more strict).
C > O AND AND O - L <= .2 * (H - L) AND O < L1 AND C1 < C2 AND C1 < O1 AND C2 < C3 AND C2 < O2
While you have stated that you are not sure how big the up move has to be, most sites seem to want the Close to be "near" the High. If we add this requirement and use the same relatively loose 20% criteria we come up with:
C > O AND AND O - L <= .2 * (H - L) AND H - C <= .2 * (H - L) AND O < L1 AND C1 < C2 AND C1 < O1 AND C2 < C3 AND C2 < O2
There are all sorts of adjustments that could be made. We could tighten up (or even loosen) the definition of "near" for the top and/or bottom. We could use a different definition for downtrend. We could require that the Bullish Belt Hold candle at least close the Gap Down or maybe even Close above some component of the previous Candle. There are definitions on the web that have such requirements.
Whatever you decide upon as the definition, it is probably possible to create a "correct" formula for that particular definition. The definition just needs to be unambiguously and objectively defined to create a formula.
I have brought this topic to the attention of the rest of the customer training department. I do not know if any of them are "Candle pattern people" however (maybe they are, I just don't know).
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
Registered User
Joined: 5/26/2005
Posts: 11
|
Thank you Bruce for helping with both the formula and the understanding of the steps to create it. Its now getting better results and I think I can tweak it a bit to get more prior downtrend input. You could have eliminated the less than polite comment about my "lack of interest in the Pattern" and kept that to yourself.
I am not good at taking write ups about pattern concepts from books and translating them into TC phrases very well and that's why I asked for help but it had nothing to do with lack of interest. In any event, I appreciate the time you took to help me work this out.....Wendy
|
|
Worden Trainer
Joined: 10/7/2004
Posts: 65,138
|
I think you have significantly misread and misinterpreted my lack of interest comment if you see it as somehow less than polite.
It merely refers to the fact that you are not interested in identifying the particular pattern described by Craig's version. It is the very reason that O=L was not included in the formulas I created.
As the post described the thought process used in creating the formulas in an attempt to make them easier to understand and modify, I would in fact have been remiss to leave the reason for the omission of O=L from the formulas out of the post.
-Bruce
Personal Criteria Formulas
TC2000 Support Articles
|
|
|
Guest-1 |