Flip a coin for entry

Random entries can work if you have the proper exit strategy - theoretically speaking. you need to keep your losses small and let your winners run long enough to pay for the losers. So exit only at VERY obvious technical levels after a big run.


However, you are better off entering 'smart' too- near a double top or double bottom, this will increase your chances dramatically.

Peter
 
Quote from peter ritter:

Random entries can work if you have the proper exit strategy - theoretically speaking.
More...

Oh, God. Please. No. Not again. Aaaaahhhhhhh!!!!
 
I am updating everyone on how this experiment is working out. So far, so good I have to say.

Samples: 21
Wins: 9
Losses: 12
Win Ratio: 42.3%
Longest consecutive win streak: 2
Longest consecutive losing streak: 4

Heads/Longs: 12
Tails/Shorts: 9
Ratio: 57.2%

Net pips: 100 (net of interest and spread)

Attached is the actual trading chart for anyone who is skeptial of my trades. This is from Oanda's brokerage. Real screen shot that I cropped in Paint Shop Pro. I run 1600x1200 resolution and did not resize it down so it might be a big picture for those running a lower res, however I had to jpg compress it by 70 to get it small enough to attach it to ET.

I will update again at around 40 samples.

Cheers,
DNAJ65000
 

Attachments

Not sure, but I also want to know when you flip it again to either double up or get out?

Quote from balda:

how do you pick when to flip your coin and do you flip it using same hand?
More...
 
I say, go for it, and if you make out well, remember that I encouraged you and that I accept gifts.

Having said that, since random entry doesn't have an edge, you're better off investing in an index fund.
 
I've discovered that if you use the "coin flip" method, you'll get better results by flipping only on even numbered days, when the wind is out of the southeast, after drinking 7 to 8 Coors Lights, with Van Tharp's book open to page 173, and waiting 12 seconds to look at the coin after it hits the floor. Can someone with Tradestation backtest this method for me?:D
 
Relax guys, the biggest reason why this strategy fails on stocks/futures is because you're paying coms. With forex you have a 2 pip spread to pay, he factored this into calculations. I say good luck and keep us posted.
 
P/L is a combination of entry, trade management, size, and exit. When you go to random entry, what you are really doing is testing your skills at the latter three, taking entry out of the equation. If you also do nothing but wait to hit the stop or target, you are also taking trade management out of the equation. Probably you are also trading a fixed size.

So really, what you are doing here is testing this exit strategy (-10 or +25), holding everything else constant.

Once you get enough results you will have a nice "base of operations" to fiddle with. You can then, say, try out a strict entry rule and see how it changes the results. You would soon find if your new entry method beats random or not. Either way you've made progress.

So I say what you are doing now is a very good thing.
 
Quote from peterfigliozzi:

P/L is a combination of entry, trade management, size, and exit. When you go to random entry, what you are really doing is testing your skills at the latter three, taking entry out of the equation. If you also do nothing but wait to hit the stop or target, you are also taking trade management out of the equation. Probably you are also trading a fixed size.

So really, what you are doing here is testing this exit strategy (-10 or +25), holding everything else constant.

Once you get enough results you will have a nice "base of operations" to fiddle with. You can then, say, try out a strict entry rule and see how it changes the results. You would soon find if your new entry method beats random or not. Either way you've made progress.

So I say what you are doing now is a very good thing.
More...

he would also need random entry times not just random directional trades to separate entry rules from exit rules. course then you probably find out it averages out with no edge - commissions.

wouldn't it be better to start with entry rules with profit targets at say +50 and stops at -50. That will show you if you have an edge. then just put a variable in the profit target and stop. try everything, within reason - with equal stop & profit target - to see how far out your edge goes.

what might be interesting is - random entries with scaling in - if the tradeable is trending you should eventually get going in the right direction and add to it. maybe add in some statistical analysis of peak to trow (of when the tradeable is trending) to fine tune the stop loss and profit targets.
 
You must log in or register to reply here.
Back
Top