3 Day SMA Trading

[Laksh]

Meanwhile, here are the charts of a few more instruments that I am trading with the 3 SMA.

Here is the Gold daily chart with 3 SMA. Whipsawing very badly...

Any ideas on how you would trade this?

In my 5SMA system I used 13SMA cross of 35SMA as a filter to reduce whipsaws. It does quite a good job of reducing whipsaws. Even now I am trading USDINR daily with 5sma with very good result.

Laksh

 

[prst]

Meanwhile, here are the charts of a few more instruments that I am trading with the 3 SMA.

Here is the Gold daily chart with 3 SMA. Whipsawing very badly...

Any ideas on how you would trade this?

you can use an adx filter to enter only trending moves.

 

[NTrader42]

Before I started trading the 3 Day SMA in November, I did some due diligence by analyzing Nifty data over the years. I am not able to find the file now. So let me download the Nifty EOD data from the NSE site, and redo my analysis, and then share it with you - hopefully by tomorrow.

http://www.volumedigger.com/Software/Data_Downloader.aspx

maybe this can be useful

 

[augubhai]

I was able to get the data from NSE and do some analysis. This analysis is just to get a basic idea about the expectancy from the 3 SMA method. Did it in a hurry.

The attached Excel sheet contains the Nifty Index (not Futures) data from 1994 till today. The logic is to go long when the 3 SMA turns upwards, and to go short when the 3 SMA turns down. Simple!! Dumb??

Column J of the Excel sheet is the cumulative profits if we had traded just 1 Nifty Index ((just 1 unit - not 1 lot).
For those of you who are unable to view the Excel sheet - the chart of the NIfty Index price vs. Cumulative Profit is shown below.

As you can see, the method has a positive expectancy over the years, and that is what I am trying to use in my trading.

But, you can see that there is a big drawdown - 2 years long - from Oct 2008 to October 2010. That's huge. So how do we minimize the drawdown during this period?

Per this system, we would be short on 18 May 2009, that special day when the market zoomed up 17% because of the UPA victory. So how would we minimize the loss on a day like that?

Please share your ideas.

 

[augubhai]

Ok... so 100% of us will not be able to open the Excel sheet, because TJ does not allow attaching Excel files. Let me see if I can upload in some file sharing site...

 

[augubhai]

While we are on this subject... I am also attaching what the NIfty Index price vs. Cumulative Profit would look like if we had a simpler rule.

Go long if Close is above previous day's Close, and short when Close is below previous Close.

Results are better than the ones with 3 SMA, and with lesser drawdowns. So should we discard 3 SMA?

 

[timepass]

Ok... so 100% of us will not be able to open the Excel sheet, because TJ does not allow attaching Excel files. Let me see if I can upload in some file sharing site...

If you have a gmail account, you can upload it in google docs and share the link. Else there other file loading sites like 4shared.com, rapidshare.com etc.. where you can upload the files (but they are removed after some time limit - 15 days minimum).

 

[Laksh]

I was able to get the data from NSE and do some analysis. This analysis is just to get a basic idea about the expectancy from the 3 SMA method. Did it in a hurry.

The attached Excel sheet contains the Nifty Index (not Futures) data from 1994 till today. The logic is to go long when the 3 SMA turns upwards, and to go short when the 3 SMA turns down. Simple!! Dumb??

Column J of the Excel sheet is the cumulative profits if we had traded just 1 Nifty Index ((just 1 unit - not 1 lot).
For those of you who are unable to view the Excel sheet - the chart of the NIfty Index price vs. Cumulative Profit is shown below.

As you can see, the method has a positive expectancy over the years, and that is what I am trying to use in my trading.

But, you can see that there is a big drawdown - 2 years long - from Oct 2008 to October 2010. That's huge. So how do we minimize the drawdown during this period?

Per this system, we would be short on 18 May 2009, that special day when the market zoomed up 17% because of the UPA victory. So how would we minimize the loss on a day like that?

Please share your ideas.

I had done similar testing for my 5sma system. The results were similar. As you say there was a prolonged draw down for 2yrs. The actual draw-down will be worse as because you won't be able to buy/sell exactly at the next day open as the computer does. Can an individual tolerate such prolonged draw down? In fact I got discouraged because of this and don't use this system anymore for Nifty. I have been trying to find methods of reducing such draw-downs. I am enumerating some of my findings/opinion below:

1.Trade this system only when all the 3 SMAs (ie 5-13-35) are moving in the same direction serially.
2. Trade the system only when ADX(14) is greater than 25.
3. Draw a MA of your equity, say 20SMA, and stop trading or do nominal trading when your equity goes below the MA. Again resume trading with full lots when the equity curve crosses above the MA.
Of the 3 options No. 3 seems to be the most promising. I have not been able to test these conclusions in the futures market or with nifty future as that would require a large capital and I wont be able to adjust my position size according to my liking. ( I trade with a small capital). But I am trading the system successfully on USDINR and Goldguinea successfully as that allows me to adjust my position size to my liking.
This brings us to the 4th method of reducing draw-down i.e. trade an instrument which is in a strong trend and where you can adjust your position size to your comfort level.
I would like to add that NTrader's advice on adding additional units at every close in your chosen trade direction in combination with the above may reduce draw-down further.

I hope this helps.

Laksh

 

[NTrader42]

Hi Laksh

I like your concept of trading based on our own equity curve, i had worked on similar concepts but gave up due to same reasons of scalable capital requirements.

Cheers

 

[augubhai]

https://docs.google.com/open?id=0B7XbrH1rDEu0a0NhaXZJOWZSeWFyNy1JM05wbXdxZw

Here is the link to the Excel sheet containing the analysis, uploaded in Google docs.

I had earlier missed out on deducting transaction cost and slippage from the profits. In the uploaded Excel sheet, I have deducted a nominal cost of 0.2%. So the chart is slightly different from the one posted above, but not significantly different.

Note that this analysis is just indicative, since we cannot trade the Index directly. It is just that my trades are based on this analysis.