Thursday, March 29, 2012

Moving average based Trading System performance NIFTY

EMA / DMA Moving average based Trading System performance based on India NIFTY Index.
 Period tested - ~ 5 years ,from Jan 2006 to 5 Mar 2012 i.e 1533 trading days
  • NIFTY range in test period - low of 2836 to high of 6318.
  • Bull  / Bear Markets covered in test period - 2 bull market of ( 2006 to 2007 and 2009 to 2011) and 2 bear markets - 2008-2009 and late 2010 - 2011
  • Signal  / Indicator used  - 10 Day exponential moving average EMA
  • Rule - BUY if at EOD (End of Day close) if NIFTY closes above 10 day EMA. Keep position open till NIFTY goes below 10 day EMA (at close) and square off / exit at EOD close.
  • No. of trades generated  - 121 i.e average of 1 trade in 12 days. See below for details
Below is the summary performance along with the scatter plot of the 121 trade performance along with the timeline. ( click to zoom )
Important takeaways
  1. One can see that using a stop loss of 1% of trade significantly improves the avg. return per trade from +1.0% to + 2.2%  and median return goes from (-0.3%) to +1.0%.I have read many traders use 1% as stop loss /max risk per trade and suggest new trades should even use small / stricter stop losses until one is confident and mature enough to play with 1% stop losses.
  2. Just like letting a losing trade lose / averaging a losing trade is a sin even taking profits quickly without riding the big trend is also a sin. To recover 1% stop losses, if one exits at say +1% profit, then obviously trading capital will be almost at same or at a below level than started.For trading capital to grow,one needs to ride the big waves, i.e home runs.We can see the outliers i.e since Jan `2006 at least 16 trades have given returns of +5% or more. And six of them gave above +10% returns ( from 10% to 24%). As they say you all need to be patient with winners and impatient with losers.( and these are all returns based on underlying i.e with 0% leverage. Using F&O Futures or options can magnify these home runs i.e can be career defining trades).
  3. Winning percentage using this simple trading rule is 42%. ( drop from 70% in 2007). Which i think is pretty fine. What matters is maintaining the discipline of using the system. And as with any system , it should be normal to expect losing streak of trades. E.g Longest straight / consecutive losses were 7 in a row in 2 periods - Jan 2008 to Apr 2008. ( during the big reversal) and Sep 2008 to Dec 2008 bear market with added volatility due to Lehman bust). Winning percentage less than 50% seems like even a coin toss is better than even top successful traders such as Peter Brant have winning percentage of ~30%. Good post on this by Mercenary Trader.
  4. Every trade strategy needs an ideal environment or market phase and any trading using moving averages will perform best in trending markets and will not perform that well in sideways/ rangbound markets.
And here is interesting part. I have taking 20 past rolling trades rolling to get a winning percentage. It can be seen that winning percentage has gone down from 70% to 42%. One reason of this drop could be the volatile nature of markets.Moving averages best work in trending markets. Other reason could be market becoming for efficient ?
And lastly, here is the details of all 121 trades. First table, sorted by negative and positive returns.Second table by timeline.

1 comment:

seema said...

Good post. I learn something totally new and challenging on blogs I stumble upon on a daily basis. It will always be interesting to read articles from other authors
and practice something from their websites...