Is a system with 80% accuracy and 1:1 r/r ratio good?

Originally Posted by aroymkp

I have recently backtested a system with about 75-80% successful trades and a 1:1 risk reward ratio? I have tested it on an hourly timeframe for a period of 2 years.

Should I consider trading it? Is the 2 year backtesting sufficient? Theoritically all seems good - I will be grateful for comments.

Originally Posted by ratan jain

Hey Aroy......the % winning seems good, but will u be able to REALLY extract even 1:1?

Actual trading may take it down to 1:0.7
That may be a problem.......

Live trade it in cash mkt with small qty and see, does your ratio of 1:1 hold? I doubt it may not....

Also, I would consider a r/r ratio of 1:2 a minimum to be able to trade any system

First check if 1:1 holds up, if it does, I will guide you with a complex martingale......

Dont worry, its not a double up after every loss type

Originally Posted by rajak1981

80% accuracy with 1:1 r/r is fine.
It 'should' be a profitable system.
http://www.geocities.com/rajak1981/f...fit multiplier
Check out this ppt on r/r and accuracy charts.
P.S: I don't work for MTPRedictor.

This ppt also has other stuff on the abc setup if you wish to look into it.
http://www.************/file/19489416...v11-1ppt1.html

Note: First link is geocities easily overloaded.
Second link is to 4shared. Just browse thru the doc to find the appropriate graphs.

This is all theoretical. You need to be sure that there are no catastrophic losses in real life.
Note: As CV says post an equity curve. I would extend that by saying: post equity curve with drawdowns+intraday drawdowns. I am not sure all software out there will show both types of drawdowns on the same chart.

For further practical caveats read on:
Every system can start to go broke i.e stop working on the exact day you start using it. (Refer to my other posts on backtesting/induction/efficient markets)
Just as investors may diversify across both growth and value picks.
Technicians are also suggested to diversify across systems.

You need a full probability distribution of various outcomes to get a feel for the tradeability of the system. There is a certain piece of software by Van Tharp which will let you run a simulation by entering probability distributions of profits. The gr8 thing about this software is that it will teach you; To not stare at charts but to follow trading systems. Very often when people write a good system they wreck it by trying to make it a gr8 system. OR by trying to trade using other filters in their own head. If you really want to use a filter why not test it out on old data first?? On the fly improvization is something that is best done done only if
a) you were born lucky/gifted.
b) you have sufficient experience i.e you posess intuition and you have the ability to distinguish between intuition/imagination/wishful thinking/hope/greed/fear.
Anyway basic point? "Plan your trades, (test your plan), trade your plan".

http://www.iitm.com/pos-siz-game.htm
P.S : I dont work for van tharp.

Anytime you hold overnight. The risk is a lot higher than you can possibly imagine. Whenever I hold overnight I assume 10% gap down on stocks and 5% gap down nifty. Even this might not be true worst case scenario. The worst case is always yet to come.

Basic ideas for converting theoretical returns into practical returns.
a) be aware of how long lean periods can last in your trading system. days? weeks? months? years??? Backtesting can reveal this if you test properly (read b)
b) torture test your system. if its a trend follower test it in a sideways market. If your system needs volatility test it in a flat market. If its a counter trend system test it in a one sided market. If its a long only system , test it in a bear market, short only , test it in a bull market. etc etc. Work out all combos to figure out worst case scenarios!
(good systems will survive adverse conditions by breaking even. gr8 systems will walk out with tiny profits, weak systems will lose while poor systems will go broke.)
c) Double your max drawdown is the basic wisdom out there in books. In addition I like to halve my theoretical returns. Reason: In real life entries/exits will not be perfect and trading slippage/comissions will always be a bit higher than we anticipate.

Theory is theory, practice is practice. Look at any gr8 athlete or trader. They'll all tell you the same thing. Good technique+practice are the keys to success. Get a coach/mentor/trading partner who will save you from yourself.

P.S: For every piece of wisdom out there in the markets. There is a contrarian piece of wisdom , quite a bit like life.

Confused yet??
Summary:
a) Refer to links to the ppt on risk management, r/r, accuracy and Van Tharp on trading systems.
b) Be aware of practical caveats to theoretical results.
c) Start trading. Even after you have the theoretical stuff down perfect. It will take you a while to perfect the practical aspect. (You can read a million books on golf, do a million wind tunnel tests. You'll only learn to drive 300+ yards by practicing for many many years)
Today I took a trade and exited preemptively for no really good reason on a retracement instead of waiting for a mechanical exit signal.

Very very few traders are perfect, but as the saying goes 'practice makes perfect'.
Other option if you seek excellent execution: automated trading.
I'm still hoping we'll get that in India within a few years.

Note: Leave a comment if you find this info helpful.