What's your exit strategy?

[newguy05]

My 5 year backtesting result with a mixed sample of 500ish nyse+nas stocks between $5-55 range is at 45/55 win/loss and -5%

What i found interesting is the 55 percent losers, more than half has good entry, meaning the stock does go up, just that my system doesnt exit until it becomes a loss. My current exit is a simple ma crossover.

I tried using a trailing stop but that didnt work as it limited my winners more than helping reduce losers.

What exit strategies do you use in your system?

 

[Martin Gale]

100 sleeping pills

 

[newguy05]

Quote from Martin Gale:

100 sleeping pills

Sorry but i think everyone is only interested in good exit strategies. Yours obviously didnt work too well as you are still here alive and posting.

 

[frostengine]

newguy,

What you are describing is something I've seen in all of my automated systems. I have yet to find a truly good answer to it.

 

[btowntrader54]

Couldn't you take tighter profits?

 

[quantparis]

May be you could test by applying two different exit rules;
1 entry,2 exit
entry size 10, exit 1 size 5 and exit 2 size 5;

 

[pneuma]

Quote from newguy05:

My 5 year backtesting result with a mixed sample of 500ish nyse+nas stocks between $5-55 range is at 45/55 win/loss and -5%

What i found interesting is the 55 percent losers, more than half has good entry, meaning the stock does go up, just that my system doesnt exit until it becomes a loss. My current exit is a simple ma crossover.

I tried using a trailing stop but that didnt work as it limited my winners more than helping reduce losers.

What exit strategies do you use in your system?

Can you give us a bit more info about your trading style - trend or swing or what? They all have different approaches.

Obviously you are concerned about win:loss. I personally would be more concerned about expectancy and Sharpe ratios. You can make significant returns with 70% losers if you know what you are doing.

With some more info would be useful.

pneuma

 

[newguy05]

My system only looks at the daily bar not intraday. I am not sure if you call it trend or swing, basically it enters order when ma10 moves above ema20 above ema30 (among many other conditions), and can hold a position open for months.

 

[lescor]

I either let the market tell me i'm wrong and exit, or I use a profit target combined with a time out.

I'm a big fan of time stops actually. If the trade has a positive expectancy, I might as well give it as much time to work as I can, often right to the close. Or I try to identify a time by which if the trade hasn't worked, the odds start to deteriorate too much. The key is to spread myself around in many smaller sized trades and try not to get too lopsided long/short or concentrated in one sector. It's rare that a stock triggers my 'uncle point' where the pain is just too much and I start to puke it out. So the bad trades can be bad, but most of the time there's enough good winners to offset the not so bad losers. And I don't stress over all the whipsaws that are triggering other trader's stops.

 

[pneuma]

Quote from newguy05:

My system only looks at the daily bar not intraday. I am not sure if you call it trend or swing, basically it enters order when ma10 moves above ema20 above ema30 (among many other conditions), and can hold a position open for months.

Since you are a EOD trader and seem to be a med-long trend follower I would do three things:

1. Use a Day Only stop entry say 1xATR(10) above the close.
This will help confirm the trend you are hoping to jump aboard.

2. Place your initial stop at a couple of ATR(10) below your entry price say 3xATR(10)

3. Place you trailing stop at a % below highest price since entry OR at a x day low (say 20 days) OR at a multiple of the ATR below the highest price since entry say 7xATR(10).

You could use a profit target, but i find for this style it is more a hindrance. However, i also agree with time exits say hold for 6 months - 125 trading days. But for these systems you would be wise to rank your stocks by the speed of their up trend, usually through a custom indicator, and open positions in the top 10.

pneuma