Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Simple Reversal Spotting Trades (MA/Elliott Wave Osc) 1 reply

Trend identification and trend change spotting 8 replies

Trend-spotting 151 replies

Spotting patterns with camera 9 replies

Need Help Spotting Coding Error 6 replies

Attachments: Trend spotting
Exit Attachments

Trend spotting

  • Post #1
  • Quote
  • First Post: Sep 7, 2007 6:12am Sep 7, 2007 6:12am
Hello folks,

This thread is for trend-followers and their personal tricks to spot trending scenarios. I would like to discuss our personal methods to filter out trending vs. non-trending markets when it comes to sniff around for a trade.

Alright, I personally use 3 SMAs for this purpose: the 30, 50 and 100 SMA, and I need them to plot in the chart in a particular fashion before considering an entry.

These are my pre-entry conditions moreless:

1) All 3 SMA in order, meaning 30 > 50 > 100 for an uptrend and viceversa.
2) All 3 SMA somewhat parallel to each other (no crossovers)
3) All 3 SMA pointing in the same direction
4) Price remains in the right side of the 50 SMA most of the time.
5) 100 SMA not breached at any time
6) Price swings harmonically off the area between the 30 and 50 SMA. I search for mild trends and this conditions filters out over-extended moves.

When these conditions apply, I seek for entries in the direction of the SMAs. When price breaches the 100 SMA or there's a crossover, I consider the trend in danger and I seek for entries no more. I apply these conditions to all pairs and to all time frames from the 30 mins to the Daily chart.

See screenshot below for a graphical picture.

So, that's my technique for finding trends. What are your tricks folks?
Attached Image
  • Post #2
  • Quote
  • Edited 7:22am Sep 7, 2007 6:42am | Edited 7:22am
I thought about what a trend is for some time. The way I define an up trend is that there are more ticks up then there are ticks down for the given time frame I am looking at. So How do I measure the strength of a trend? Well, if a sideways moving market = no trend for the given time frame, then that must mean that there is an equal number of buying to selling. There for the ratio of buying to selling should be the same , or it in fact should be 1:1.

So any thing that creates a number that is less than or more than 1:1 is in fact a trend. The larger the ratio (positive for up trend, negative for down trend) the stronger the trend is.

This indicator is my definition of a trend. I only did a text version, because I can keep the values in my head and I do not like to cluter up the chart with abunch of lines.

The first part is a comparison of a short term ATR for a given time frame to a longer term for the same time frame. This tells me if volatility is increasing or decreasing.

The second part tells me the trend and the ratio. This tells me the strength of the trend. a 1:1 ratio is No Trend.
Attached Image
Attached File(s)
File Type: mq4 Aggregate Trend v2.mq4   3 KB | 708 downloads
I'm liking trendlines more and more.

I posted this rather ugly EURUSD chart to jacko's thread the other day, let me put it up here as well...might do some more for my journal and videos later, so might add some here as well.

Basic logic is - hop a trendline for good = end of direction, establish a new one with a higher high or lower low to make a new TL until that in turn breaks, then start over.
Attached Image (click to enlarge)
Click to Enlarge Name: eurusd_290807_gianta.gif Size: 22 KB
Trust price. Know yourself.
Quoting smjones
The first part is a comparison of a short term ATR for a given time frame to a logner term for the same time frame. This tells me if volitility is increasing or decreasing.

The second part tells me the trend and the ratio. This tells me the strength of the trend. a 1:1 ratio is No Trend.
Hello smsjones,

What ratios signify a solid trend according to your indicator? for example, is a 0.5 ratio a strong downtrend?
Quoting WTB
Hello smsjones,

What ratios signify a solid trend according to your indicator? for example, is a 0.5 ratio a strong downtrend?
No, I wanted there to be no ambiguity, so I programmed ratios less that 1 to become negative. There is no number between 1 and zero, only 1 and up or -1 and down. To me a ratio of 2:1 or greater is a strong trend. I have seen ratios as high as 6:1 and as low as -7:1. The market lately for EJ anyway has fallen much sharper than in incline.

The factor input in my indicator is the threshold to determine if there is a trend at all. In the example, I have it set to 1.1:1. So, the indicator shows No Trend, because 1.03: 1 is less that the threshold of 1.1:1.

I hope I answered your question.
Another thing one must consider is the timeframe the trend we are talking about is in. Also, we must consider how many bars back we will include to determine a trend. To me the entire concept of trend is relative to these factors and cannot be defined independant of them.
Quoting smjones
No, I wanted there to be no ambiguity, so I programmed ratios less that 1 to become negative. There is no number between 1 and zero, only 1 and up or -1 and down. To me a ratio of 2:1 or greater is a strong trend. I have seen ratios as high as 6:1 and as low as -7:1. The market lately for EJ anyway has fallen much sharper than in incline.

The factor input in my indicator is the threshold to determine if there is a trend at all. In the example, I have it set to 1.1:1. So, the indicator shows No Trend, because 1.03: 1 is less that the threshold of 1.1:1.

I hope I answered your question.
I see, I think I understand. So, anything between 1.1:1 to 2:1 is a mild trend, and anything beyond 2:1 is a strong trend.

One more question: what does "streght" mean or rather how do I interprete that figure?

Thank you very much!
That number is determined by subtracting the number of up pips from the number of down pips. It is not significant in itself, I just like the info.
The way I determine the up pips and down pips is to take a bar and if the close is greater than the open that is defined as an up bar, so I take the range of the close to the open and add that to an up aggragate. Same for the down pips.
Then I subtract the larger from the smaller and that is the strength number.
When I first wrote the indicator I used the open to high or low, but then felt that was not realisitic, because that could easily be a spike. So I settled on close.
If someone wanted to, they could easily put this in a window with a line graph or a histogram. I just never have, because I see it in my head.

I posted my source code, so feel free to add to or change it as one sees fit.
I've always defined a uptrend as higher highs and higher lows and a downtrend as lower highs and lower lows. That simple definition will beat just about any lagging trend indicator known to traders. I am constantly amazed at the lengths someone will go to define one of the most basic concepts of trading.
Attached Images
Quoting philmcgrew
I've always defined a uptrend as higher highs and higher lows and a downtrend as lower highs and lower lows. That simple definition will beat just about any lagging trend indicator known to traders. I am constantly amazed at the lengths someone will go to define one of the most basic concepts of trading.
Relative to what?

Yes what you have shown is a trend, but is it if you include more bars?

One must define the context. how far back are we looking and in what timeframe.

I think subtracting the ups from the downs is a pretty simple definition.
Thanks SMSJones for the detailed explanation
Quoting philmcgrew
I've always defined a uptrend as higher highs and higher lows and a downtrend as lower highs and lower lows. That simple definition will beat just about any lagging trend indicator known to traders. I am constantly amazed at the lengths someone will go to define one of the most basic concepts of trading.
The problem I have with simply defining a trend as "higher highs and highs lows" is that, while ultimately true, it doesnt tell me anything about the trend in itself. Yes, sure, price is trending, but HOW exactly is it trending? what does the trend actually look like? is it a mild trend? is it a strong trend? is it a whisawing trend? is it steady or is it violent?

That's why I like to "entrap" the trend within my SMA conditions in order to make sure Im searching for a "tamed" trend.
Quoting smjones
Relative to what?

Yes what you have shown is a trend, but is it if you include more bars?

One must define the context. how far back are we looking and in what timeframe.

I think subtracting the ups from the downs is a pretty simple definition.
It's trading 101, one of the oldest definitions in the book. You compare highs to highs and lows to lows. You only need to look at the current peak/valley and the immediately preceding one, no more, no less. The addition of more bars is irrelevant. You are either in a trend now or you're not.
Is this an up trend?
How about this?
How about this?

The context must be defined.
Attached Image (click to enlarge)
Click to Enlarge Name: ScreenHunter_025.gif Size: 12 KB
Attached Images
UP
UP
DOWN

Until the definiton of higher highs and higher lows (or lower highs and lower lows) is broken then the prevailing trend is defined by the move that got you to that point. Clearly, if you can't break S/R then you are in no trend.
Quoting philmcgrew
UP
UP
DOWN

Until the definiton of higher highs and higher lows (or lower highs and lower lows) is broken then the prevailing trend is defined by the move that got you to that point. Clearly, if you can't break S/R then you are in no trend.

Well that is my point exactly. the fact that you have answered the same chart three different answers with three different contexts shows that the context is important.
Quoting smjones
Well that is my point exactly. the fact that you have answered the same chart three different answers with three different contexts shows that the context is important.
I could care less if it's the same chart or not. You only compare the current peak/valley to the previous one. No other "context" is necessary. The fact that you expanded and contracted the same chart has no bearing on the outcome of the determination.

My intent was to express one of the oldest definitions in trading that still holds true today. Not to ruffle your feathers and incite you to prove me wrong. If you have to put moving averages on a chart to figure out if it's going up or down, I'm happy for you.
Quoting philmcgrew
I could care less if it's the same chart or not. You only compare the current peak/valley to the previous one. No other "context" is necessary. The fact that you expanded and contracted the same chart has no bearing on the outcome of the determination.

My intent was to express one of the oldest definitions in trading that still holds true today. Not to ruffle your feathers and incite you to prove me wrong. If you have to put moving averages on a chart to figure out if it's going up or down, I'm happy for you.
Phil my friend, no ruffling here at all. I always enjoy and learn from your point of view. I just happen to not agree with you on this.. I do understand your argument, though I simply reject parts of it, as not well enough defined. I do not have any moving averages on my charts.. ???
a long term stochastic for me
it doesnt exactly define an up or downtrend but when the stochs are trending strongly or at extremes, thats where i look for trades because they are high probability trend trades.

hidden divergence usually gets me in sometimes with teh help of a trendline break.

example trade:

short woulda been after the big bar that closed down right under 2.4275. wait for a small retracement then get in.
Attached Image
0 traders viewing now