It's all in the shadows

[dpaterso]

Here's another one of those 'hypothetical subject to interpretation and I have a theory about this' questions:

In a short sentence what would you say the shadow of a candlestick or bar is representative of (if anything) and is the size or length of the shadow indicative of yet something more?

[dpaterso]

C'mon John, you KNOW when I post a question here I'm trying to 'hook' an answer from you!!! !!!

[daydreamer65]

Originally Posted by dpaterso

Here's another one of those 'hypothetical subject to interpretation and I have a theory about this' questions:

In a short sentence what would you say the shadow of a candlestick or bar is representative of (if anything) and is the size or length of the shadow indicative
of yet something more?

I still find the webpage below very useful.

Introduction to Candlesticks - Bear v Bull

[dpaterso]

Thanks daydreamer65, I appreciate the answer.

Let me put it this way:

I've noticed that there are certain instruments (pairs) that ALWAYS have HUGE shadows (above and below) (USD/ZAR for example) and there are other instruments (pairs) that, for the most part, have very small shadows (above and below) (EUR/RON for example) (the pairs used in this example of course depict 'extremes').

What I'm 'mulling over' is if this means anything.

Also: if a pair that normally has very small shadows (like EUR/RON in the above example) suddenly starts 'showing' longer and longer shadows then is this indicative of 'something' and would a MA or an EMA of the lengths of the upper and lower shadows over a certain period give some useful information (like a 'dynamic' way of calculating where to place stop orders instead of the old 'a couple of ticks above or below the high or the low'??? I mean 'a couple of ticks above the high' on Gold, for example, would be far 'different' from 'a couple of ticks above the high' on EUR/USD, for example, even if you take into account the different tick or pip values). What you're in effect trying to accomplish is the placing of orders that when they are 'hit' they are INDEED 'HIT with a vengeance' as opposed to just being 'caught' for no real or apparent reason and then the price retraces resulting in a loss until you stop and reverse. This is what I'm 'toying with' at the moment. You cannot use something like the ATR because this is an average of the TR and if you used the ATR as a factor in your calculations as to where to place these orders they would never get 'hit' at least not in my lifetime i.e. the ATR value is always far too big to used for this purpose.

(And I suppose the rest of you thought that I was asking this question because I was too 'bone idle' to look through candlestick pattern documentation)!!!

[akeakamai]

This means volatility. You can get an idea of how the volatility played out during the session by seeing where the shadow is but I don't think it's particularly useful for setting stops. I think finding areas of consolidation, or stop and reverse areas work much better for stop setting.

[dpaterso]

Thanks for the input.

You are right but, following on from my 'Wilder Systems', I'd like to try and find some sort of 'mathematical basis' for these values based on, exactly as you say, the volatility of the instrument concerned. I'm 'playing' with this idea at the moment, let me see what happens. So far it's kept me out of some very bad trades (trades that I definitely would have been in at this point had I been following the 'couple of ticks above or below the high or the low' concept) and has let me down only twice (well, so far, as I type, it appears to have let me down, but who knows what the rest of the day will bring).

[jomni]

longer shadows = more tendency to revert to the mean.
Harder to use trend trading techniques and it may spoil indicators based on the closing price.

[dpaterso]

Thanks for the message.

Are you sure though that you don't mean it will affect indicators that are based on the extreme high or extreme low? The only reason I ask is that I know with Wilder's systems and indicators, because they are mostly based on the close, a quick run up or run down during the day does not affect the results.

I must tell you that 'version 2' of my experiment seems to have even more merit i.e. it's a 14 day MA of the sum of the length of the upper shadow and lower shadow for each candlestick or bar. While it may APPEAR that it causes you to place orders far away from the price I'm surprised at how it keeps you out of potentially bad trades and gets you into the good ones. I've also found that by reducing the the period i.e. from 14 days to 7 or even 5 it reduces the distance. Another way, maybe, is to set the period 'dynamically' i.e. if the shadows over a period are relatively consistent in length but sometime during the period you get a 'spike' that's 'out of the ordinary' then by making the period 'spike day less 1' you're back to what would be considered 'normal' for that instrument.

I'm going to see what the effect is of using an EMA instead of an MA (SMA) today (might solve for the 'spike day').

Just experimenting BUT I do believe that somewhere here there is 'an edge' just waiting to be 'uncovered'.

Put it this way: had I been using the 'good ol' few ticks above or below the high or the low' 'rule' I'd be sitting in a 'world of sh*t right now'!!!

[jomni]

True. But what I meant was if you are using indicators based on close as entry and exit signals, you'll get more false singals as the prices diverge from the indicators (unless you only act upon opening and closing of the bar).

I must commend your research though. Keep it up.

[dpaterso]

Hello, and thanks for the compliment.

You're correct in your observation: in my case the system(s) I trade ONLY give signals based on the 'final' daily 'situation' i.e. 'on the close'.