Moving Average

[Nero64]

I have been looking at heaps of charts lately and finding the 200 day MA to have some merit. A lot of stocks have broken the 200 MA and continued upwards. Some bounce off it and some well haven't reached it and others see it as resistance

What got me into MA as an indicator was the reading of Sam Weinstein's book. He drills it home on nearly every page.

He recommends 30 Wk MA for investors and 10 Wk MA for traders.

Just recently the XAO pulled above the 200 MA which some say is a bullish sign.

Anybody prefer one MA tmeframe over another?

 

[CanOz]

At the end of the day the MA or EMA is only an indication of the trend right? Its a reference point, so to me, with all reference points it good to use the same one all the time.

I like the 200 MA, but also use a 50 MA and a 13EMA. Either way though its only an indicator of the trend for me.

Cheers,



CanOz

 

[beamstas]

I don't give moving averages much merit. Instead of giving a picture of what the price is doing now, it just lags behind and shows what the price has done in the past.

Alot of funds use the 200day moving average as a filter.

As a stand alone thing, i'll use a simple system to demonstrate to you what the 200day moving average is like as a buying signal

1992 - 2009 on the ASX 300, here is the code i have used

//Buy when crossing 200day ma
//Sell when going back under the 200day ma

Buy = Cross(H,MA(C,200));
Sell = Cross(MA(C,200),L);
PositionSize = -5;

And here is the result

I hope that from this you can draw your own conclusion on the 200day moving average.

Personally i use a 40 day moving average filter on the XJO for systems design, it is robust and seems to work in most situations.

Cheers
Bradley

 

[moXJO]

I don't mind GMMA as a trend rider.

 

[CanOz]

I don't mind GMMA as a trend rider.

GMMA isn't bad for a nightly scan actually, i occasional used it this past year when looking for new setups. I would run the scan then eye ball the tickers it picked up.

Cheers,


CanOz

 

[Tradesurfer]

The 200 day can be a little lagging. I prefer the 50 day. The challenge with the 200 day is if you enter a trade on a cross above it (price bar closes today above 200 ma and yesterday it was below it), once you get a very strong trend price tends to create more and more distance above the 200 MA line. What winds up happening is that if you exit strategy is to sell when price crosses back below it you might give up quite a bit of profit.

Don't get me wrong, a long holder of a stock would want to see price above the 200 MA certainly.

One thing to pay attention to is the slope of any moving average line. If price is in a downtrend and a candle or price bar moves above the average, if the line is still slanted or sloping down it may not be as high quality of an entry. In general, better breakouts happen when you've had a basing or flatening out period of a stock and it starts to run.

I do you the 50 day moving average but tend to optimize entries with other indicators as well.

One interesting thing I was looking at was the 20 month moving average on the S&P 500 Index. Since 1994 price crossing above or below the 20 month moving average happened 4 times. 2 long signals and 2 sell(or short) signals.

Still think technical analysis is inferior to fundemental analysis?

 

[beamstas]

One interesting thing I was looking at was the 20 month moving average on the S&P 500 Index. Since 1994 price crossing above or below the 20 month moving average happened 4 times. 2 long signals and 2 sell(or short) signals.

Still think technical analysis is inferior to fundemental analysis?

That's just curve fitting - the 20 month moving average has no predictive power at all.

 

[tech/a]

I have been looking at heaps of charts lately and finding the 200 day MA to have some merit. A lot of stocks have broken the 200 MA and continued upwards. Some bounce off it and some well haven't reached it and others see it as resistance

This paragraph has covered every possibility

What got me into MA as an indicator was the reading of Sam Weinstein book. He drills it home on nearly every page.

STAN Weinstein presented the M/A observation way back in 1988.When computers took up buildings and software was limited to Pak Man.

Technical analysis was limited to pad and pencil and hand drawn charts (In the most part).

He recommends 30 Wk MA for investors and 10 Wk MA for traders.

Just recently the XAO pulled above the 200 MA which some say is a bullish sign.

Anybody prefer one MA timeframe over another?

Now 20 yrs later with technical analysis far in advance of the 80s and software which can actually test these Theories and Hypothesis coupled with interest in Tech analysis from a few Doctors of Mathematics like Howard Bandy much which has led the way in technical thinking is now finding its way into the technical scrap yard.

But hey many will still wear a "Mullet" and be happy with their life.
Even though 20 yrs of study has rendered the Mullet Obsolete.

 

[Tradesurfer]

With regard to curve fitting. I am in favor of selective curve fitting actually.

Since trends happen again and again, using indicators or oscillators that get the large piece of trends, curve fitting makes sense since you are setting up the opportunity to take advantage of trends when they happen.

 

[beamstas]

With regard to curve fitting. I am in favor of selective curve fitting actually.

Since trends happen again and again, using indicators or oscillators that get the large piece of trends, curve fitting makes sense since you are setting up the opportunity to take advantage of trends when they happen.

I disagree TradeSurfer.
Patterns do occur, never exactly the same shape, never exactly the same length of time, never the same price action. All curve fitting achieves is showing you what you could have got during the past, but this is not indicative of future performance.

When i design a system, i don't choose the best paramaters, i choose the most robust. If a 20month moving average works, then a 19month and a 21month moving average should give similar restuls, then a 18month moving average should give similar to the 19 months, and a 21month should give similar results to the 20 month. If 20month is the only good timeframe, then when the market changes (as it always does), even a tiny bit, you are going to be left with a broken system.

Regards
Brad

 

[tech/a]

Brad.

A trend is a trend.
Without a trend in the direction of your trading you wont profit.
I agree with Derek in as much as placing yourself in front of any indication of an emerging trend --regardless of timeframe--makes perfect sence.

 

[beamstas]

Brad.

A trend is a trend.
Without a trend in the direction of your trading you wont profit.
I agree with Derek in as much as placing yourself in front of any indication of an emerging trend --regardless of timeframe--makes perfect sence.

I disagree Tech,
If you had to be trading with a trend swing traders would never profit, there are 1000's of ways to profit from trading and trend following is not the only way.. Anyway that is another pandoras box of discussion not for today.

I am not questioning that using a moving average as a trend filter is a bad idea, i use one myself.

A 20month moving average - fine, if that's what you like use it, i have no problem with that, But i wouldn't go as far as to say that it is any better than a 15month or a 10month moving average, just because it has worked the last 4 times.

Brad.

 

[moXJO]

That’s why you need the awesomeness of multiple MMA, everyone knows more is better

On a serious note, I don’t mind using GMMA as a trend confirmation or possible turning points. Or as a guide for just sitting in long term trades. It has the added benefit of band width and turn over points. Needs to be used with the usual stops and trend line break rules if the bands are narrow. Imo they are not to shabby. You can tighten up losing to much, or leaving money on the table.

 

[beamstas]

Here is a simple test, using amibroker, on the XJO.

//Buy On a Cross of the 20month MA
//Short On a Cross under the 20month MA

Period = Optimize("MaLength",20,1,100,1);
MMA = MA(C,Period);
Plot(MMA,"20MonthMA",colorRed,styleLine);

Buy = Cross(C,MMA);
Sell = Cross(MMA,C);
Short = Sell;
Cover = Buy;

BuyPrice = MMA;
SellPrice = MMA;
ShortPrice = MMA;
CoverPrice = MMA;

shape = Buy * shapeUpArrow + Short * shapeDownArrow;
PlotShapes( shape, IIf( Buy , colorGreen, colorIndigo ), 0, IIf( Buy , Low, High));

Regards
Brad

 

[Timmy]

Without a trend in the direction of your trading you wont profit.
I agree with Derek in as much as placing yourself in front of any indication of an emerging trend --regardless of timeframe--makes perfect sence.

I disagree Tech,

What's there to disagree with?
Whether you are trading for a tick profit or you are Mr. Buffet, or any frame in between, the price has to trend in your direction to enable you to take a profit.

 

[beamstas]

What's there to disagree with?
Whether you are trading for a tick profit or you are Mr. Buffet, or any frame in between, the price has to trend in your direction to enable you to take a profit.

I guess then that depends on your definition of a trend. I always percieved the meaning of trend as the general direction in which the market is moving.
I always thought the market could move "against the trend" , and the market can chop around "not trending at all".

I stand corrected.

 

[Tradesurfer]

Brad and others

with regard to my 20 month MA example.

I was doing an interview with a reporter who had the belief that technical analysis would not have helped traders

a) prevent losses with the recent downturn

b) profit from the move (they didn't understand short selling)

Their other assumption was that technical analysis was only for short term daytraders.

I quickly put together a long term use for technicals with a simple 20 month MA which produced those 4 trades to help them see that wasn't the case and thought it was an interesting chart.



I agree that there isn't anything significant with the #20. The 50 day MA is a part of my proprietary indicator but only 1 component. And off course moving averages in itself are lagging. But as I said, my prop indicators only use it as a piece.

By selective curve fitting- I do favor it in that I have my own indicators that WILL capture the meat of large trends. I don't know which will offcourse wind up being large.

As an educator, trader, and technician I also hold the belief that I don't make predictions I simply follow price and I never know when an entry will yield a small loss or become a monster trending position. For me its all about playing the percentages and letting my winners run and putting in place systems to do just that while cutting losses small. A trend following system normally has many small losses and a winning percentage somewhere between 35-45%.

 

[beamstas]

Brad and others

with regard to my 20 month MA example.

I was doing an interview with a reporter who had the belief that technical analysis would not have helped traders

a) prevent losses with the recent downturn

b) profit from the move (they didn't understand short selling)

Their other assumption was that technical analysis was only for short term daytraders.

I quickly put together a long term use for technicals with a simple 20 month MA which produced those 4 trades to help them see that wasn't the case and thought it was an interesting chart.



I agree that there isn't anything significant with the #20. The 50 day MA is a part of my proprietary indicator but only 1 component. And off course moving averages in itself are lagging. But as I said, my prop indicators only use it as a piece.

By selective curve fitting- I do favor it in that I have my own indicators that WILL capture the meat of large trends. I don't know which will offcourse wind up being large.

As an educator, trader, and technician I also hold the belief that I don't make predictions I simply follow price and I never know when an entry will yield a small loss or become a monster trending position. For me its all about playing the percentages and letting my winners run and putting in place systems to do just that while cutting losses small. A trend following system normally has many small losses and a winning percentage somewhere between 35-45%.

Thanks for the response -- and well said.

 

[tech/a]

I disagree Tech,
If you had to be trading with a trend swing traders would never profit, there are 1000's of ways to profit from trading and trend following is not the only way

Other than Arbitrage,Id be really interested in even one other way to profit from the markets which doesnt involve a trend.
A move in price in either direction at any point of time is a trend in that direction.

We analyse and take trades IN ANTICIPATION of a move in our direction.
Not prediction---as those who have little knowledge of T/A presume.

 

[fapturbo]

I have been looking at heaps of charts lately and finding the 200 day MA to have some merit. A lot of stocks have broken the 200 MA and continued upwards. Some bounce off it and some well haven't reached it and others see it as resistance

What got me into MA as an indicator was the reading of Sam Weinstein's book. He drills it home on nearly every page.

He recommends 30 Wk MA for investors and 10 Wk MA for traders.

Just recently the XAO pulled above the 200 MA which some say is a bullish sign.

Anybody prefer one MA tmeframe over another?

Using Amibroker and Using the Rate of Change of the 200 Day Moving Average as a Trend Filter you can design a very good Trend Following Mechanical System