Double Top and Double Bottom

[cherryblossoms]

What is the time period considered for identifying a double top (or bottom) formation in a stock?

 

[alexaherself]

It's very hard to answer in abstract, because the predictive value of such chart formations wears off gradually, as a probability-function, rather than having a specifically defined cut-off point.

It varies, according to quite a few parameters, including volatility, fundamentals, the strength of the trend before the double-top/bottom, and so on. But for a well-traded, liquid stock, I'd say up to 3 months or so, from a daily chart, other things being equal. It would still be a double-top/bottom after much longer than that, too, but how "predictive" it would realistically be is another matter.

And it's going to depend on the longer history of the stock (specifically with regard to how predictive its own previous double-tops/bottoms have been), as well.

 

[malaguti]

What is the time period considered for identifying a double top (or bottom) formation in a stock?

a double bottom/top can happen across all timeframes, from 1 minute to monthly so its not really a question that can be answered. Its the characteristics of the pivots that determine not necessarily the time its taken to form

 

[Atilla]

a double bottom/top can happen across all timeframes, from 1 minute to monthly so its not really a question that can be answered. Its the characteristics of the pivots that determine not necessarily the time its taken to form

I see it as trench war-fare between bulls and bears.

Agreed the ground in the middle and time-frame is not necessarily important.


For me it's a case of which side pushes through to a higher high or a lower low once the double levels have been breached.

Onwards and forwards as we say

 

[cherryblossoms]

Thanks and what is the best source to explore on the characteristics of pivots?

 

[TechQuant]

cherryblossoms, you seem keen to keen to explore TA in an empirical manner and I salute that. DBs [or any chart pattern for that matter] are more of an art than a science. Even when DBs have 'shown' some predictive power in the past, it doesn't necessarily hold this time, or the next.

I strongly suggest in support of your endeavours you review all occurrences of DB [or any other TA pattern], as many as you can in fact, and tabulate just what the percentage win rate is for each of them. Before you do that, google Confirmation Bias just to be sure what to look out for.

 

[Solas0077]

For me it is 10s to a 30 bars on any timeframe. I do not like shorter or longer. But it depends and I don't know of any general rules along these lines.

 

[bredin]

Thanks and what is the best source to explore on the characteristics of pivots?

mathematical, or actual?

for actual, take whatever tf you are looking at and zoom out as far as you can in your platform so that thhe candles/bars are as horizintally compact as possible. The actual pivots will be glaringly obvious at this point.

mathematical pivots (those derived mathematically, like in a pivot indicator) are a waste of time.

B.

 

[tomorton]

bredin - Thanks for this, I have always thought mathematical pivots were just smoke and mirrors. But is there a study or source that backs us up on this?

 

[Splitlink]

I take all chart patterns with a big pinch of salt. In the beginning (Genesis?) there was sense in what patterns told us. However, other people study them as well. New employees at big trading firms are taught to watch out for them because that is where the suckers put their stops.

When they have learned that lesson they are then taught how to manipulate prices to create these patterns. They have so much money to do this and, although I have heard indignant cries of "not possible", I believe that to be the case.

You take care with those doubles--stick to Scotch.

 

[wallstreetwarrior87]

When they have learned that lesson they are then taught how to manipulate prices to create these patterns. They have so much money to do this and, although I have heard indignant cries of "not possible", I believe that to be the case.

You take care with those doubles--stick to Scotch.

Very true, so if a pattern can be formed out of manipulation, it suggests that there will have been a likely reason/need for this to occur, which suggests that there is generally a grand plan(possibly known) in advance, which then helps counter any argument that the markets (in general) are not random.

So rather than asking about what a double top or bottom is, it is best to know the cause and effect of one, then you can be more confident of riding a bottom to the top (your target) rather than getting dumped out of your long as soon as you enter.

This has the makings of good thread!

 

[bredin]

bredin - Thanks for this, I have always thought mathematical pivots were just smoke and mirrors. But is there a study or source that backs us up on this?

Apart from my own trading? Im not a great reader of other people trading ideas.

However I do have hundreds (lit) of markups showing real pivots, and they form an important part of my trading method.

They became much more important after investigating point and figure charting, which really highlights pivots by making price movements more granular. Of course the usual way P&F charts are displayed obsfucates this aspect of them, so a more readable P&F method needed to be developed. Luckily MT4 serves this purpose nicely (see below).

Pivots are specific S/R lines that are characterised by the tendency of price to fail to close over that line, or close over and not close back (ie not form a FatCat, at least not immediately), but 'retest'.
They are also characterised by price being unable to travel between them in a small number of candles. If price can travel between pivots in a few candles it is, in fact, a lower timefame pivot. The "power" of the pivot is determined by the largest timeframe it can be seen on.
Pivots tend to occur at higher timeframe extremes. Thus they are good indicators of where the larger orderstacks are sitting in a market, and since those large stacks need to be overcome, once overcome they tend to leave a large orderstack behind; Supply becomes Demand at that point (R->S in the notation i use on my charts)

Simply put, they are lines that price tends to stay on one side or the other of.
The key aspect of this is that one must allow for "false breaks" just like in any trendline/channel analysis.

Here is AN Daily with *some* pivots (im sure you can see more). Please note that the chart is zoomed right out so we cant see any individual candle. Some say that its easier to draw these off line (close) charts, but I like to see the full range of price movement (wicks) whcih can often give a better idea of rejection/retesting.

"But wait, thats just S/R!"

Not quite, a pivot cant exist until after a "retest" of a "breakout", S/R exists as soon as the extreme is formed.

So, whats up with mathematical pivots?
No idea. since they only hit real pivots by coincidence (draw some real pivots then load a pivot indicator to see what i mean), but do so with enough probability that cognitive bias comes into the picture and gives a false impression of how often they "work", like fibonacci levels.

How to you trade that?

Since all trading comes down to "draw a line and trade away from it" the answer should be obvious.

B.

** Point an figure done right, in MT4

and zoomed out to see the pivots

edit: ignore that blue line, i did not realise it was there until after posting.

 

[Splitlink]

I do not think that the money market guesses. I believe that the retail trader does. If a pivot is formed Mr Market figures out, with all the considerable means at his disposal, where the stops are. Those stops will be triggered, for sure.

Once triggered, the trader has only two possibilities. Be right, or be wrong.

I think that it is as simple as that. Day trading, ie. very short term, is random.

I have read the posts written daily by the analysts on the site and, please notice, their predictions are always written with the conditional "could".

It could go up or down, couldn't it?

I've been doing this for a long time. Because of my misgivings, I am a moderate trader. Thinking as I do, I would be a fool to be otherwise. However, the bug has hooked me and I must keep trying for something new. At my age, it's good for the brain box!

 

[bredin]

splitlink,

please answer the following question in 10 words or less:
"what makes a market move?"

when you can answer this, you will also know why pivots (and most other 'chart patterns') form. Even market manipulators must adhere to this principle; even ghosting or other forms of faking S/D needs this to work properly.

Pivots are awesome for one reason only: they provide the places on the chart with the smallest risk and greatest rewards, outside of huge timeframe extremes.

And while we're at it, heres something ive been telling my students for many years:
"Only a fool allows their stop to be hit."

B.

 

[bredin]

I have read the posts written daily by the analysts on the site and, please notice, their predictions are always written with the conditional "could".

It could go up or down, couldn't it?

Its not a matter of where it could go, but of where it can not go.

B.

 

[timsk]

. . . Point an figure done right, in MT4

Hi bredin.
Is this a free EA for MT4 and, if so, could you supply a link for it please. If not, can you tell us where you got it?
Thanks,
Tim.

 

[alexaherself]

Its not a matter of where it could go, but of where it can not go.

Indeed.

I've just been re-reading those books in which Jack Schwager interviews some of the world's most famously successful traders (Market Wizards and The New Market Wizards) and although I hadn't particularly noticed this when I originally read them, this time around I was really struck by this regularly recurring refrain from those interviewed.

 

[Splitlink]

splitlink,

please answer the following question in 10 words or less:
"what makes a market move?"

when you can answer this, you will also know why pivots (and most other 'chart patterns') form. Even market manipulators must adhere to this principle; even ghosting or other forms of faking S/D needs this to work properly.

Pivots are awesome for one reason only: they provide the places on the chart with the smallest risk and greatest rewards, outside of huge timeframe extremes.

And while we're at it, heres something ive been telling my students for many years:
"Only a fool allows their stop to be hit."

B.

Today, the market is rigged.

A lot of those stops are to enter a trade, not exit and there is no reason why a dealer should not enter his own stop or limit orders and remove them as necessary in order to get things moving as he wishes.

I don't pretend to be an expert of the "other side" but Greece was sucked into a big debt problem by Goldman Sachs. This type of firm is what we are trading against. No one should underestimate their power in the money markets.

 

[Fugazsy]

Patterns in larger time frame are useless because conditions change from one day to the other and it becomes more a guess than anything else, in a shorter time frame future prices are more predictable and patterns can be more reliable, I know a trader that trades only H&S's as reversal and also as a continuation pattern on the 5 minute and she is doing pretty well, patterns needs a lot of subjectivity, a break of a neck line does not necessary need to be traded as a break or traded at all....in a strong market are not advisable, in that case a fail break is a better option....because many traders that take not in considerations the market conditions are only going to be trapped....

 

[Splitlink]

Its not a matter of where it could go, but of where it can not go.

B.

It can go anywhere and the position that the retail trader takes has to be one of two---right or wrong. So if he is lucky enough to have entered in the direction that GS & Co have decided to trade. Then he is right.

Pivots are only understood in hindsight.