Timeframes & Indicators

[rfk_from_za]

Hi Guys,

I have been dabbling in intraday trading and am wondering what timeframes people are using to determine trends. Currently I look at 1, 5, 15, 30 and 60 minute timeframes.

I use MACD, Momentum and and a 3/10 Oscillator are some of the indicators I use.

My question is, should I be using different periods for each timeframe, or can I keep my period settings the same across all the time frames? If I keep them the same across the time frames I often see the indicators contradict each other at different timeperiods.

For example the 5 min MACD Histogram will be moving up towards the 0 line from below, while the 30 minute MACD histogram will be moving down towards the 0 line from above.

Are these contradictions correct, or should I be adjusting my period settings so that the indicators more or less match up across all time periods.

Thanks,

Rob

 

[options]

Tooooooo many charts, toooooo many indicators.

This will/has led to trigger freeze. You won't know when to pull the trigger to enter or exit a trade.

Work out what indicators are working in line with the price action. That suit you!

Look at all the time frames before you start by all means. In fact it's advisable to. Then settle on a max of 3.

A lot of traders, after deciding which way their trend is heading only work with one chart, and one lot of settings.

Saves confusion and makes life a lot easier.

 

[rossored]

I have suffered from "trigger freeze" until very recently, perhaps for the reasons that Options has said.

I have two screens, one with one instrument on in 1m, 5m and 13m timescales. I also have a daily chart on the same screen, but it is hidden during market hours - I just look at it pre-market. I tend to use the 1min for entries and then move up to 5m or 13m, depending on whether its a quick scalp or I think it might go somewhere.

On the second screen, I have two other instruments on 5m timescales.

No indicators, just price, volume and horizontal S&R (that's not to say you shouldn't use indicators if that's what floats your boat)

 

[Roberto]

My question is, should I be using different periods for each timeframe, or can I keep my period settings the same across all the time frames? If I keep them the same across the time frames I often see the indicators contradict each other at different timeperiods. For example the 5 min MACD Histogram will be moving up towards the 0 line from below, while the 30 minute MACD histogram will be moving down towards the 0 line from above.

Yes; you will get that. The idea of "high probability trades" is that you leave the indicator settings the same at different timeframes, and only take the trade when the indicators agree. So the example you mention is a lower probability trade which you would "filter out" by using charts of different periodicities with unchanged indicator settings.

Are these contradictions correct, or should I be adjusting my period settings so that the indicators more or less match up across all time periods.

They are correct. If you think about it, there'd be no point in adjusting the period settings for the indicators, because if you're going to do that, you are effectively looking at one set of information twice rather than trying to corroborate it "independently" with something else. In other words, you're adding nothing.

(There are a very few opinionated puritans who would contend that even doing it "correctly" is adding nothing. But they're just wrong. Usually. You can often tell from the patronising way they express their forecful opinions. )

At another forum, there was recently a discussion about using a MA-based system on 10-minute charts, and "confirming" the "entry signals" by looking at 5-minute charts ... but doubling the periods used for the MA's. This was all rather illogical ... "it's technical analysis, Jim, but not as we know it!"

At the moment I don't use indicators myself, but I've heard some good things from people I trust about this "3, 10 oscillator", and I know it relates to something used by Chick Goslin, for whom I have great respect. Have you had good experience with it, yourself?

 

[rfk_from_za]

I have had some reasonable successes with 3/10 oscillator. I use it as a confirmation of the current trend. If the oscillator is trend up and so it the price I usually take that as a sign that the price is gonna keep moving in its current direction.

Thank you for your detailed reply with regards to the time frame question, your thoughts pretty much match up with what I felt was the correct way. (since I had placed some trades when the indicators at different time frames all agreed and these have been successful on most occasions)

Since you are in agreement with me would you agree that if I am looking at a 15 minute chart I can safely say that what my indicators are telling me are at least valid for the next 15 minutes. Same for indicators at other time frames. Does that make sense, since this would help me time my trades with a little more accuracy.

Out of curiosity since you don't use indicators how do you decide when to enter or exit a trade, are you using TA patterns to identify reversals etc?

Rob

 

[peterpr]

We've all been there

Indicators are just that - indicators. They are simply mathematical derivatives of historical price/volume/period action. They have predictive power only to the extent that there are genuine cycles or pattern repeats contained in the data series itself and which can therefore be projected into the future. However, there are no short-cuts. You need to understand exactly how the indicator you wish to use is derived and exactly what it is supposed to 'indicate' when it behaves this way or that. You then have some hope of adjusting its parameters to best fit the data series and time frame you choose to provide trading triggers or other useful info.

I have a fond belief that there is propably the odd autistic-type mind out there that can see such cycles clearly without recourse to indicators at all. You kow the sort of thing - like the guy who can solve the 13th root of a random 22 digit number in 10 seconds without recourse to pencil and paper, let alone a calculator. For us lesser mortals, indicators are the way to go, but they're all hard work and no magic - unless you put your faith in someone else's 'system' that is - and you wouldn't want to do that now would you?

 

[Roberto]

would you agree that if I am looking at a 15 minute chart I can safely say that what my indicators are telling me are at least valid for the next 15 minutes. Same for indicators at other time frames.

I don't know about "safely" but that ought to be the theory, IMHO.

Out of curiosity since you don't use indicators how do you decide when to enter or exit a trade, are you using TA patterns to identify reversals etc?

Yes indeed; I'm using two specific TA patterns of price-bars (1-2-3-formations and Ross-hooks) to try to identify reversals and trend-continuation respectively. But I've used indicators before, and next time I fancy some "research" I'll be looking at this 3/10 oscillator of yours, of which I hear such good things.

 

[ZDO]

It's amazing (to the point of being an anomaly to look into ?! ) just how little of the total time shorter intraday time frame 'trend' indicators are concurrent with longer intraday time frame 'trend' indicators. And, when you do find a set of tf's that are actually 'harmonic' that's a sure indication that they are getting ready to go out of harmony... Makes using one to filter above or below timeframes fairly useless unless you are extremely selective (and patient). Using one timeframe trend to filter for another has only worked for me by only using 2 timeframes - 1 intraday and one of a much longer time frame (daily or weekly) and taking more lots when lower timeframe signal is concurrent with a sensitive long time frame trend indicator (and fewer when they are not concurrent)

Also consider looking at Gosling's work. Instead of multiple timeframes he uses multiple indicator 'length' settings to accomplish basically the same 'filtering' objective...

hth

 

[rfk_from_za]

Do you have any links to 'Gosling's work' I would be interested in seeing how he calculates the his increasing periods. Since I find the MACD to be a particularly useful indicator of trend I wouldnt mind seeing 3 (5,15 & 50 minute) MACD's all in the same view.

I seem to remember something about this in one of the articles on this site. Must research it a bit more and I will post the results of what I find and how successful they are.

cheers

Rob

 

[ZDO]

Actually the correct spelling of his name is Goslin and the primary source is his book Trading Day By Day. Up here the main thread for his work is the topic "Trading Day by Day" by Chick Goslin. Just search on Goslin to find it.

The book reveals no secrets - or better said his secret = you must learn to interpret price and the multiple oscillators yourself. For him the best trades are those where all the indicators are concurrent (ie going in the same direction) and he wetware computes what price action it would take to change a trend then carefully takes 'anticipatory' positions ... I brought him up as one alternative to multiple timeframes. His perspective and methods may not resonate at all with yours...

hth