[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [amibroker] Re: System Performances



PureBytes Links

Trading Reference Links

Hi Bob,

Tuesday, April 18, 2006, 1:04:44 PM, you wrote:

BJ> Stop loss vs. stop limit?

How could it matter?  The former will not guarantee execution at a
specific price, and the latter will not guarantee *any* execution,
ever.

Also, intermilan04 wrote (below) that his system is a daytrading
system so that there are no gaps.  AFAIK, gaps can develop anytime
intraday, although they are somewhat rare.  But here in Tokyo a stock
gaps any time there is an order imbalance of a requisite amount.  And
while somewhat rare, it is far from unknown to have stocks suddenly
become bid-only or ask-only intraday in response to a sudden influx
of orders (or one big one) on one side or the other.

(Often the best response to this is to immediately fade the impending
gap with a market order, as it's either a mistaken entry amount by
someone, or an emotional overreaction, or maybe a single large seller
who has reached a sell target.)

I think the points of most responders to this thread however would
break down as follows:

1) It is a truly rare system that does not suffer performance
degradation from using stops.  Personally, I don't think I have ever
seen one that didn't suffer thusly.

2) Stops that are very tight (I consider 3 percent pretty darn tight)
are going to get banged when it doesn't really mean anything (except
that the trader has now booked a loss), and it's much better to use
some volatility determined stop, based on say the last 50 or so bars,
rather than to put a flat percentage on it.  Many stocks have
different volatility characteristics.  Three percent on one would be
nothing, while three percent on another would set off alarms.

3) One has to be able to take a 3 percent negative move in underlying
price, most of the time, IMHO.  That doesn't mean taking 3 percent
portfolio hits of course.  It means you have to adjust position size
so that 3 percent of price is within risk bounds.

4) Stops are really to prevent one from getting into *serious*
trouble, rather than to cut losses to a "comfortable" amount.  Yes,
they are risk control devices, but they are best used well outside
normal volatility ranges, as protection against very significant
losses, rather than as triggers for comfortable losses.  (If the
amount is *that* comfortable, the stop will get hit, as I mentioned
above.) If a system needs more than the emergency stop, it probably
isn't a system I'd want to trade very enthusiastically.

All of this ultimately applies to backtesting of course -- the
original thrust of this thread.  The problem with stops and
backtesting is obvious: you cannot guarantee anything, so you have to
use wider slippage, which makes your metrics much less meaningful.
The larger the percentage of exits triggered by stops, the less
backtest results are likely to imitate real trading.

It might be a nice idea as a feature to be able to include that
number however, in a back test report.  I think it would be useful
for some to understand what percentage of their exits were triggered
by stops, and what the possible implications of that might be.

My two yen,

Yuki

BJ> Fred,

BJ> Could you explain as to why 3% wouldn't always limit losses to 3%?
BJ> Assuming the stock has some volume (at least 100K), and I set stop
BJ> loss order as soon as I buy stocks...I'm not quite sure of the
BJ> circumstances where 3% stop loss would not work.

BJ> My system is a daytrading system so there is no gap ups and downs.

BJ> Regards,

BJ> intermilan04





------------------------ Yahoo! Groups Sponsor --------------------~--> 
GFT Forex Trading Accounts As low as $250 with up to 400:1 Leverage. Free Demo.
http://us.click.yahoo.com/lpv1TA/jlQNAA/U1CZAA/GHeqlB/TM
--------------------------------------------------------------------~-> 

Please note that this group is for discussion between users only.

To get support from AmiBroker please send an e-mail directly to 
SUPPORT {at} amibroker.com

For other support material please check also:
http://www.amibroker.com/support.html

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/amibroker/

<*> To unsubscribe from this group, send an email to:
    amibroker-unsubscribe@xxxxxxxxxxxxxxx

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/