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

Re: MENSA EasyLanguage Test Question



PureBytes Links

Trading Reference Links

> > Anyway, here is the question:
> > On a tick chart, how can I tell that the current tick bar is the
> > last bar of the day without looking ahead at the next bar?


here's paintbar code for PST SPoos, etc:

IF  time>= 1314 and time <=1315 Then Begin
  Plot1(High,"PBHigh");
  Plot2(Low,"PBLow");
  IF CheckAlert Then Alert = TRUE;
End;

Of course it doesn't work on early closes,
markets with no last minute ticks, etc.

I'll let Gary collect the meal since
he's within driving distance of Tom.

BW


>From: "Gary Fritz" <fritz@xxxxxxxx>
>Reply-To: fritz@xxxxxxxx
>To: Alexander <alexander@xxxxxxxxxxxxxxxxx>
>CC: omega list <omega-list@xxxxxxxxxx>
>Subject: Re: MENSA EasyLanguage Test Question
>Date: Thu, 10 Jan 2002 09:34:37 -0700
>
> > Anyway, here is the question:
> > On a tick chart, how can I tell that the current tick bar is the
> > last bar of the day without looking ahead at the next bar?
>
>You can't.  Not possible.  There's no way of knowing if another tick
>is going to arrive.
>
>I think the best you could do would be to do a statistical
>approximation of the likelihood of this being the last tick.  Look at
>the time, look at the tick rate for the last few minutes, look at the
>normal last-minute behavior in the last several days (since there is
>typically a flurry of ticks just before the close) and you could make
>a WAG* at the probability of the current tick being the last.  But
>you can't know for certain.
>
>(* WAG = Wild-Assed Guess)
>
>Do I still get the free drink?  :-)
>
> > This is a real trick on a tick chart. On an hourly it is easy, you
> > can just look at the session time.
>
>Even that's not 100% reliable.  Some days don't run until the normal
>closing time.  But presumably you're more interested in the realtime
>behavior than in backtesting, so you'd know in realtime if the market
>is supposed to close early today.
>
>Why do you want to know?  Maybe there's a better way to do what you
>want.
>
>Gary
>