|
PureBytes Links
Trading Reference Links
|
Don't waste your time: SPH loses after commish and slippage,
ESH is a loser net before....but then I didn't optimise it
for the "most robust" settings..... ;-)....oh, and BTW, it
still trades WAY too much (20+/day).
>If I remember correctly, the system was something
>like: put three RSI's (or other indicator as you are investigating) with
>differing lookback periods on the 1-minute chart. When they are ALL
>pointing in the same direction AND they are pointing with an angle of some
>magnitude(can this be quantified?)<<
Inputs: RL(14), RZ(9), RX(5);
If rsi(Close Data2, RL) > rsi(Close[1] Data2, RL)
and rsi(Close Data2, RX) > rsi(Close[1] Data2, RX)
and rsi(Close Data2, RZ) > rsi(Close[1] Data2, RZ)
and roc(rsi(Close Data2, RL),1) > 13
and roc(rsi(Close Data2, RX),1) > 13
and roc(rsi(Close Data2, RZ),1) > 13
Then Buy;
If rsi(Close Data2, RL) < rsi(Close[1] Data2, RL)
and rsi(Close Data2, RX) < rsi(Close[1] Data2, RX)
and rsi(Close Data2, RZ) < rsi(Close[1] Data2, RZ)
and roc(rsi(Close Data2, RL),1) < -13
and roc(rsi(Close Data2, RX),1) < -13
and roc(rsi(Close Data2, RZ),1) < -13
Then Sell;
>From: "Wes Williams" <softexcl@xxxxxxx>
>Reply-To: <softexcl@xxxxxxx>
>To: "Jim Johnson" <jejohn@xxxxxxxxxxxxxxxx>, "Wes Williams"
><softexcl@xxxxxxx>
>CC: "Bill Wynne" <tradewynne@xxxxxxxxxxx>, <mr_bond@xxxxxxxxx>,
><ebonugli@xxxxxxxx>, <omega-list@xxxxxxxxxx>, <rfurse@xxxxxxxxxxxx>
>Subject: TIKI as a market timing tool
>Date: Thu, 21 Feb 2002 17:26:19 -0700
>
>Hi Jim,
>
>Personally, I've investigated the TIKI in depth but I prefer automated
>systems rather than discretionary ones. Still, I am hopeful that your
>persistence may pay off. If I remember correctly, the system was something
>like: put three RSI's (or other indicator as you are investigating) with
>differing lookback periods on the 1-minute chart. When they are ALL
>pointing
>in the same direction AND they are pointing with an angle of some magnitude
>(can this be quantified?), it is an indication that strong program buying
>just kicked off. We must keep in mind that the TIKI is a reliable indicator
>only at extreme levels. The stronger the program buying, the stronger the
>TIKI moves. My friend who was using this system was profitable but I did
>not
>like it because I thought it was discretionary so I went elsewhere.
>Further,
>the angle bends show up strongly on Ensign software on tick-level data. So,
>what does it look like on TradeStation 1-minute bars? I do not know what
>you
>will find. Well, that's all I know about it.
>
>I think that your Summation idea is an excellent one. Would it not make
>sense to take a weighted MA of the Summation since you may go from low
>periods to high periods and the Summation will hide the change in increased
>activity? With the creativity you've demonstrated I'm sure you will think
>of
>something fine.
>
>Sincerely,
>Wes Williams
>
> > -----Original Message-----
> > From: Jim Johnson [mailto:jejohn@xxxxxxxxxxxxxxxx]
> > Sent: Thursday, February 21, 2002 4:18 PM
> > To: Wes Williams
> > Cc: Bill Wynne; mr_bond@xxxxxxxxx; ebonugli@xxxxxxxx;
> > omega-list@xxxxxxxxxx; rfurse@xxxxxxxxxxxx
> > Subject: Re[2]: OddBall as a market timing tool
> >
> >
> > Hello Wes,
> >
> > One thing I've played with a bit is to create a summation of tiki. No
> > real nice looking curves so far but I do think there is a kind of
> > "balance of power" reading to be had. Not for scalping but to see if
> > stealth accumulation is taking place prior to the rest of the market
> > reacting. something like this:
> >
> > summation(tiki's > +20, 15 minutes) 0 summation(tiki's < -20 , 15
> > summation(tiki's > minutes)
> >
> > a sort of net extreme tiki over the last 15 minutes.
> >
> > Thursday, February 21, 2002, 3:34:06 PM, you wrote:
> >
> > WW> Bill,
> >
> > WW> I am unsure of your intent or the timeframes you used, but
> > were you using
> > WW> 1-minute bars on the TIKI or something else? Program trading
> > only lasts for
> > WW> a short while and when it starts, the TIKI will move but then
> > return to
> > WW> "normal," whatever that may be in a few minutes. So, if you
> > were using a
> > WW> longer timeframe for Data2 perhaps you have a reason why? The
> > TIKI is for
> > WW> day trader timing only, in my opinion. Even then, there are
> > other things
> > WW> that can make the TIKI move other than program trading, which
> > is why it is
> > WW> imperfect. But program trading will move it.
> >
> > WW> Sincerely,
> > WW> Wes
> >
> > >> -----Original Message-----
> > >> From: Bill Wynne [mailto:tradewynne@xxxxxxxxxxx]
> > >> Sent: Thursday, February 21, 2002 1:00 PM
> > >> To: softexcl@xxxxxxx; mr_bond@xxxxxxxxx; ebonugli@xxxxxxxx;
> > >> omega-list@xxxxxxxxxx
> > >> Cc: rfurse@xxxxxxxxxxxx
> > >> Subject: RE: OddBall as a market timing tool
> > >>
> > >>
> > >> {S&P data1, TICKI data2}
> > >>
> > >> Inputs: RL(14);
> > >> If rsi(Close Data2, RL) > rsi(Close[1] Data2, RL) Then Buy;
> > >> If rsi(Close Data2, RL) < rsi(Close[1] Data2, RL) Then Sell;
> > >>
> > >> Here you go, check how "robust" it is for yourself.....sorry,
> > >> I just can't help it some times. :-)
> > >>
> > >> BW
> > >>
> > >> Disclaimer: DON'T USE THIS SYSTEM! As far as I can tell it loses
> > >> money across all time frames and parameter sets.
> > >>
> > >>
> > >> >From: "Wes Williams" <softexcl@xxxxxxx>
> > >> >Reply-To: <softexcl@xxxxxxx>
> > >> >To: "Bill Wynne" <tradewynne@xxxxxxxxxxx>, <mr_bond@xxxxxxxxx>,
> > >> ><softexcl@xxxxxxx>, <ebonugli@xxxxxxxx>, <omega-list@xxxxxxxxxx>
> > >> >CC: <rfurse@xxxxxxxxxxxx>
> > >> >Subject: RE: OddBall as a market timing tool
> > >> >Date: Thu, 21 Feb 2002 12:16:18 -0700
> > >> >
> > >> >I agree with David's assessment on the TIKI as it stands, but do
> > >> this: Run
> > >> >an RSI against it and use that as your indicator rather than the
> > >> TIKI per
> > >> >se
> > >> >(you won't read that in the books). It is still imperfect but I had
>a
> > >> >friend
> > >> >who used it to verify the trade he was about to take. If the
> > RSI on the
> > >> >TIKI
> > >> >was strongly pointing the other direction, he wouldn't take the
> > >> trade until
> > >> >more information came in.
> > >> >
> > >> >Jim Johnson cited the Larry McMillan book I read on the TIKI.
> > I wouldn't
> > >> >recommend the book for just the chapter on the TIKI unless
> > you can get it
> > >> >from the library or borrow it.
> > >> >
> > >> >Sincerely,
> > >> >Wes
> > >> >
> > >> > >
> > >> > >
> > >> > > David:
> > >> > >
> > >> > > My take is the ticki is generally too noisy, while in
> > >> > > strong trends it may not get to some arbitrary point
> > >> > > (say -20) and you miss the move. But hey, if someone
> > >> > > has a ticki system that beats the basic OB, or whatever,
> > >> > > please post the results.
> > >> > >
> > >> > > Perhaps my comment "a complete waste of time" is too
> > >> > > harsh, but I think there are better places to look.
> > >> > >
> > >> > > BW
> > >> > >
> > >> > >
> > >> > > >From: "David Folster" <mr_bond@xxxxxxxxx>
> > >> > > >To: "Bill Wynne" <tradewynne@xxxxxxxxxxx>, <softexcl@xxxxxxx>,
> > >> > > ><ebonugli@xxxxxxxx>, <omega-list@xxxxxxxxxx>
> > >> > > >CC: <rfurse@xxxxxxxxxxxx>
> > >> > > >Subject: Re: OddBall as a market timing tool
> > >> > > >Date: Thu, 21 Feb 2002 10:28:01 -0800
> > >> > > >
> > >> > > >Bill,
> > >> > > >
> > >> > > >I found TIKI had some value as a daytrading technique if you
> > >> > > used it as an
> > >> > > >oversold measure in a rising market and vice versa for a falling
> > >> >market.
> > >> > > >i.e. If 5 minute SP > than 25 period MA and tiki = -20 or
> > less, buy.
> > >> > > >Didn't
> > >> > > >define an exit technique but one possibility would be exit when
> > >> > > you have a
> > >> > > >profit and the majority of a 5 minute tiki bar is < 0. Possibly
> > >> > > adding my
> > >> > > >OB DB as a filter for longs/shorts would work better.
> > >> > > >
> > >> > > >Any comments?
> > >> > > >
> > >> > > >David
> > >> > > >
> > >> > > > > Raschke, Conners, Schwartz all talk about tick and tiki
> > >> >methodologies
> > >> > > > > in recent books..... (IMHO a complete waste of time...the
>tiki
> > >> >systems
> > >> > > >that
> > >> > > > > is). Ned Davis and many others have been doing breadth (i.e.
> > >> > > > > A/D) research for decades.
> >
> >
> >
> > --
> > Best regards,
> > Jim mailto:jejohn@xxxxxxxxxxxxxxxx
> >
> >
>
|