Moving Average

[bunyip]

Hi Bunyip,

As per your diagram, I understand that you would place your initial stop-loss just under bar 3 – is that correct? What would you do if bar 4 turns out to be a bearish candle like bar 2 (ie. moves up triggering your entry, then heads down closing below your entry price)? Do you exit immediately, or wait to see if the initial stop gets hit?

thanks

That's correct - the stop would go under Bar 3.
If my buy order is executed and then the bar closes below my entry but without hitting my stop, I don't exit the trade - I wait for the stop to be hit, which hopefully it won't be.
The stop is to shield me from catastrophic loss if the trade goes against me, but at the same time it should be far enough from entry to allow the trade a bit of wiggle room. If you bail out of the trade because it's closed below your entry, you're allowing it virtually no wiggle room at all.
After closing below your entry, the trade could well head the right way and put you in profit. You have to at least give it a chance to do that, rather than eliminate that chance by immediately closing the trade.
If your stop is hit, well, the loss will be only a very small percentage of your trading account if you're prudent with your money management.

 

[beamstas]

Moving Averages smooth out Price Movements, they give a very clear way of measuring a Trend. If you want to devise a Mechanical System you need a way of being able to code a "Trend"

Can you tell me of a method to do that??

There are plenty of ways!
The relation of various highest high values or lowest low values to each other is one way

 

[fapturbo]

If your stop is hit, well, the loss will be only a very small percentage of your trading account if you're prudent with your money management.

What has the Stop Placement got to do with how much you decide to risk?

 

[fapturbo]

There are plenty of ways!
The relation of various highest high values or lowest low values to each other is one way

Over what period??

A week?? Two weeks??

 

[seasprite]

Back to moving averages, don't use them. Think simple trend structure would tell you a lot more.

What a load of toss. One MA will give nothing , however two or more at a crossover in the sp in an uptrend or downtrend will more or less indicate the sp direction, and it's this trend you play.

 

[fapturbo]

What a load of toss. One MA will give nothing , however two or more at a crossover in the sp in an uptrend or downtrend will more or less indicate the sp direction, and it's this trend you play.

Crossover is not the point.

Identifying a tradeable trend is the point.

What do you consider a tradeable trend??

How do you rank stocks based on trend??

Is stock xyz in a better trend than abc??

 

[MRC & Co]

What a load of toss. One MA will give nothing , however two or more at a crossover in the sp in an uptrend or downtrend will more or less indicate the sp direction, and it's this trend you play.

Sorry, I meant to say I don't use them.

But why don't you try use the MACD while your at it, heard it's the goods too!

 

[beamstas]

What a load of toss. One MA will give nothing , however two or more at a crossover in the sp in an uptrend or downtrend will more or less indicate the sp direction, and it's this trend you play.

Seasprite, have you read this thread?
https://www.aussiestockforums.com/forums/showthread.php?t=15817

Over what period??

A week?? Two weeks??

I guess it depends what timeframe you are trading. If you are trading for 10months the HHV of this week and the one last week won't really help

I just coded this quickly in amibroker.. it does ok. im sure someone with more experience could write something better. this code doesn't use any moving averages.. just HHV values.. im sure it could be done ..

It's the best i can do in 10minutes..
(not perfect and maybe never will be)

 

[fapturbo]

Seasprite, have you read this thread?
https://www.aussiestockforums.com/forums/showthread.php?t=15817



I guess it depends what timeframe you are trading. If you are trading for 10months the HHV of this week and the one last week won't really help

I just coded this quickly in amibroker.. it does ok. im sure someone with more experience could write something better. this code doesn't use any moving averages.. just HHV values.. im sure it could be done ..

It's the best i can do in 10minutes..

How about the rate of change of the moving average over a set period of time?

Do you think this would give a better indication of trend??

The slope of a moving average in other words. The higher the gradiant the better the trend.

Take the value of the moving average today and compare it to the value of the moving average "x" days ago.

 

[seasprite]

How about the rate of change of the moving average over a set period of time?

Do you think this would give a better indication of trend??

The slope of a moving average in other words. The higher the gradiant the better the trend.

Take the value of the moving average today and compare it to the value of the moving average "x" days ago.

no , pick the trend before it begins.

 

[beamstas]

How about the rate of change of the moving average over a set period of time?

Do you think this would give a better indication of trend??

The slope of a moving average in other words. The higher the gradiant the better the trend.

Take the value of the moving average today and compare it to the value of the moving average "x" days ago.

No probs, which one? 200day? or 50?
I'll compare it to what i just posted see if we can't get a better indicator of trend

(see this thread is being constructive after all)

 

[beamstas]

No probs, which one? 200day? or 50?
I'll compare it to what i just posted see if we can't get a better indicator of trend

(see this thread is being constructive after all)

Here you go
Same graph
(left the writing on it)
but using the 50day MA to define trend (ie the 50day MA higher today than it was yesterday?)

Again, very basic and fast coding

 

[Chorlton]

How about the rate of change of the moving average over a set period of time?

Do you think this would give a better indication of trend??

The slope of a moving average in other words. The higher the gradiant the better the trend.

I would suggest (without any testing) that this would depend on the length of time that you would prefer to be in a trend.

Surely those with the steepest gradients have more chance of failing quicker than those which have a more steady gradient.

However, if you intend to trade very short-term then maybe there is some merit in this.

 

[fapturbo]

no , pick the trend before it begins.

Good luck doing that mate... the idea is not to pick 100 stocks and hope you get on one that trends.... nice try...

Identifying one that is trending already is key.

 

[fapturbo]

No probs, which one? 200day? or 50?
I'll compare it to what i just posted see if we can't get a better indicator of trend

(see this thread is being constructive after all)

200 Day moving average today compared to the 200 Day moving average 50 days ago.. for long term trend.

Can't amibroker run optimisation test so that you can optimise the moving average period and the look back peroid??

How about rotate stocks you hold based on the fastest rate of change of the moving average.

So you are always holding the fastest moving stocks based on the rate of change of the moving average.

So take the ASX stocks for example and apply the rate of change of moving average filter to these stocks. Apply some sort of liquidity filter and volatility filter.

Then when one stock moves into the top 5 stocks we buy this one and then sell the one that drops out.

Keep rotating weekly. That is once a week check which ones are the fastest moving stocks and make sure you own them. Might be top 5 or top 10 or top 20 depending on Capital size. Maybe just top 2

 

[>Apocalypto<]

Sorry, I meant to say I don't use them.

But why don't you try use the MACD while your at it, heard it's the goods too!

i use moving averages and a MACD MRC on 15min time frames.

not the tools u use but how u use em.

 

[seasprite]

Sorry, I meant to say I don't use them.

But why don't you try use the MACD while your at it, heard it's the goods too!

my apologies

 

[fapturbo]

Surely those with the steepest gradients have more chance of failing quicker than those which have a more steady gradient.

However, if you intend to trade very short-term then maybe there is some merit in this.

Trend is a trend right...

The stronger the trend the better right??

Why would you be worried about it falling quicker when you are riding the trend for as long as you can??

Having a predefined exit you will not be worried about whether it will fall or not

 

[MRC & Co]

i use moving averages and a MACD MRC and i profit 4.5 nights out 5 on 15min times frame.

not the tools u use but how u use em.

Yep, and how much emphasis you give them.

I bet you use a lot of other reasons for your trades which give more weight.....

 

[bunyip]

Yes i was wondering this also.

Also one other thing. I get the entry using the highs of the canldes with limit buy orders.

Do you filter stock based on "Trend"

If so can you please describe how you determine if the "Trend" is trade worthy?

I use the classic definition of a trend, i.e. higher peaks and higher troughs for an uptrend, lower peaks and lower troughs for a downtrend.
I like moving averages to confirm the trend.
I agree with anyone who says moving averages are unnecessary. But unnecessary doesn't mean they're not useful.
As for which moving averages to use, I've borrowed some ideas from Dave Landry and Alex Elder.

Elder says that whatever timeframe you trade from, you should first identify the trend in a timeframe that's four to five times higher, then you go back to the shorter timeframe and trade in the direction of the trend of the longer timeframe.
A trader from daily charts would first identify the weekly trend, which is five times longer than a day. Then he'd trade the daily chart in the direction of the weekly trend.
A trader working from 15 minute charts would first identify the trend on an hourly chart, which is four times as long as 15 minutes. Then he'd trade the 15 minute chart in the direction of the hourly trend.