|
PureBytes Links
Trading Reference Links
|
>BUY @ P3 Limit, where P3 = P1 + Delta points, i.e.
>instead - sign, a + sign is used in Buy formula.
>
>This bad order is placed above the market, and I am
>expecting that PATS will snatch it right away and
>execute it within 2 sec.
>
>Other market participants will see a spike in prices
>above current market equal to Delta.
Often there will NOT be a spike: just because your
buy limit is above the market it does not follow that
you'll be filled at your limit. The the limit is the
worst price you'll get.
I use stop limits with wide "collars" all the time (ES, NQ). More
often than not I'm filled at my stop, and I've NEVER (yet) been
filled at my limit using wide limits.
BW
>From: "DC" <dc010225@xxxxxxxxxxxxx>
>Reply-To: "DC" <dc010225@xxxxxxxxxxxxx>
>To: "david b. stanley" <davestan@xxxxxxxxxx>, <omega-list@xxxxxxxxxx>
>CC: "DC" <dc010225@xxxxxxxxxxxxx>
>Subject: Re: Bad ticks
>Date: Sun, 2 Jun 2002 06:49:52 -0500
>
>Re: Bad ticks caused by Incorrect Orders
>
>David,
>
>thanks for your valuable feedback, my system is fully mechanical
>by passing Buy/Sell signals generated by TS5 directly to PATS.
>.It also compares eSignal and PATS prices on tick-by-tick basis.
>
>You have brought another topic to my attention: aberration in prices
>can also be caused by placing "bad" orders. Let's talk about
>"bad" limit orders.
>
>Let's assume that ES is going up and is hitting a point of resistance
>at price P1. The system is predicting with a very high accuracy
>a retracement to a lower level at P2 = P1 - Delta points and then
>advance to levels above P1.
>
>A correct action for the system is to issue order:
>
>BUY @ P2 Limit, order is placed bellow the current market price.
>
>Let's assume a "bug" in EL code causing the following
>"bad" order to be sent to PATS for order execution:
>
>BUY @ P3 Limit, where P3 = P1 + Delta points, i.e.
>instead - sign, a + sign is used in Buy formula.
>
>This bad order is placed above the market, and I am
>expecting that PATS will snatch it right away and
>execute it within 2 sec.
>
>Other market participants will see a spike in prices
>above current market equal to Delta. They will interpret
>it as a "bad" tick. Only I will know it was a bug in my
>EL code.
>
>You have mentioned bad ticks caused by a poorly placed
>stop orders. My guess is that both stop and limit
>orders incorrcetly placed can cause bad ticks.
>
>DC
>
>----- Original Message -----
>From: "david b. stanley" <davestan@xxxxxxxxxx>
>To: <omega-list@xxxxxxxxxx>
>Sent: Friday, May 31, 2002 6:15 PM
>Subject: Re: Bad ticks
>
>
>You didn't mention if your system was mechanical or descetionary.
>Since you are trading index futures, a bad tick is usually
>off by 100.00 or more. Anything less is often an FOMC announcement,
>an opening gap, or a real poorly placed stop on the minis.
>
>If you are charting anything other than 1-tick charts, the H or L of the
>bar
>should be the bad tick.
>Define bad ticks.
>
>vars:BadTick(0);
>
>if (absvalue(H-H[1])>=100
>or absvalue(L-L[1])>=100
>then BadTick=1 else BadTick=0;
>
>Then nest the code that you do not want to execute on a bad tick.....
>
>if BadTick=0 then begin
> ...your code
>end;
>
>In some cases, you may want to create synthetic values for indicator
>varibles
>when bad ticks occur.
>
>
>
>
|