A Moving Average Trading System

haha nice suggestion indeed!!! :-)

But then there are times when the risk reward ratio in nifty is not so good so have to scan others..............

Alok,

thanks for the suggestion but anways any other comment on the issue and help for me???

This is the continuation of the my earlier post regarding moving average trading strategy (Message No. 109 in this thread). In that earlier post I had shown the failure of 3-13-39 method and explained briefly the reason for the failure. After experimenting with the same strategy with different periods of moving averages I found the method I am describing below to be quite effective and with a lower failure count compared to the 3-13-39 method.

As mentioned in my earlier post, the 3-13-39 method sometimes fails to indicate the entry points correctly. The reason for this failure is the short term SMA period is 3 which is too small and the long term SMA period 39 is a bit too long resulting in less number of cross overs. Therefore, I increased the short term SMA period a little and considerably shortened the long term SMA period. I also adjusted the medium term SMA period accordingly. I found a combination of 5, 10 and 20 periods is quite effective. The reason for selecting these numbers is not based on any theory. I just selected with the reasoning that 5 days data is equal to one trading week, 10 days is two weeks and 20 days is four weeks (approximately one month). The effect of this change can be seen in the attached chart. It is the same chart for TATA STEEL covering the same period as the earlier chart used in my earlier post. We can clearly see that on 28-1-05 the short term (5-day) SMA crossing cleanly above the long term (20-day) SMA giving a buy signal which was missed in the 3-13-39 method. Similarly the reverse cross-over (5-day going below 20-day SMA) occurs on 23-3-05 whereas in 3-13-39 chart it occurs on 24-3-05. Thus we get the signal one day earlier if we use 5-10-20 SMA cross over method.

Therefore, a crossing of 5-day SMA and 20-day SMA such that 5-day SMA goes above 20-day SMA gives a buy signal. Similarly, crossing of 5-day SMA and 20-day SMA resulting in the 5-day SMA going below the 20-day SMA gives a sell signal.

We can also see a couple of cross-overs of 5-10-20 SMA lines in either direction between 12-5-05 and 3-6-05, but they do not give any significant gain. So, one may ask are they not false signals? My answer is they are not. The reason is the cross-overs are not according to the rule. Now I am explaining the rule which validates the correct cross-overs:

A cross-over is valid only if it takes place in the following order: 5-day SMA should cross-over the 10-day SMA first and then cross-over the 20-day SMA. This rule is for cross over in either direction. From the enclosed chart we can see that between 12-5-05 and 3-6-05 the cross overs were not in the order mentioned above. So, these cross overs are not valid.

A second point to remember is that the 5-day SMA crossing over 10-day SMA is an alert but not a buy/sell signal. Because even after such a cross-over the 5-day SMA may not cross over the 20-day SMA and may revert to its original position above 10-day SMA. So, it is only an alert and when it happens we should get ready to buy/sell and watch for the next valid cross-over.

I have tested this 5-10-20 SMA cross over for several stocks with EOD data for the period Jan 2000 till now. It works quite well and I have observed very few failures. The system works for all kinds of stocks.

You may test this method with your own data and see its usefulness.

Happy trading.

-Anant

Attached Files

TATASTEEL-SMA5-10-20.pdf (15.4 KB, 130 views)

Hello TradersEdge,
1.
Please explain the logic of selecting the periods 3-13-39. why not 5-10-20 for example or any other combo?

2.
For intraday trading using 10-min bars, the sma39 would not start forming until ending of trading (total 335 mins). The sma39 plotting wud start only if market hours were longer than 390 mins (39*10) ! How does one get past this?

The logic of selecting 3, 13 and 39 was based on using this system on end of day charts.

3 Day MA - a proxy for price
13 Day MA - a proxy for the short term trend (a moving trend line)
39 Day MA - a proxy for the intermediate trend (a moving trend line).

Since you plan to use this on intra-day charts you will have to modify it to suit your own trading requirements.

__________________
Best Wishes!

Traderji

www.Traderji.com

Traderji,

Thanks for the reply, but my question still remains.

Let's stick to EOD charts, then. I understand the proxy bit. it's the unusual figures 3, 13, 39 that intrigue me. what I am asking is:

why 3 and not 5 (for example)
why 13 and not 15 (for example)
why 39 and not 40 (for example)

Hope I am clearer now.

Hi pgd,

These moving averages have been tried , tested , and have given reasonably good results. However if you have got better results from a different set of averages please let us know.

Regds

Sure, you can use 5,15 & 40 if it works for you. I had done a system test using 3,13, & 39 and 5,15 & 40 and the results were more of less similiar. The objective of this thread was to provide a platform to introduce novice traders to a robust and simple trading system.

You are always welcome to come up with your own trading system and share it with the members here for further discussion.

__________________
Best Wishes!

Traderji

www.Traderji.com

Do you have a success/loses ratio ?

thanks, I also use EOD datas.

don't know about copyrights etc. but nicely lifted from a known source. Let's put something original or give the owner his acknowledgements.

Hi,

Im new here and i find this system very interesting. Could someone please post a screen dump of this system. Im quite new at trading and i plotted this into my computer program. When i select Time period: 3 months. This is quite visible. But when i plott it for one day, freq: 5 min...it comes out almost horizontal lines...hmm...im quite confused. Sorry for my ignorance.

If someone could help.....tanneetan@yahoo.com.au

Most appreciated.