Lindahl patterns
Ian Williams
Andy
Maximizing Reward and Minimizing Risk with the Wellspring System
By: Alexander, Colin
Pages: 205
Publisher: Windsor Books
ISBN: 0930233506
Type: Book - Hard Cover
Publish Date: 3/1/1992
Five Star Futures Trades: The Premier System for Trading the Biggest
Market Moves
By: Alexander, Colin
Pages: 171
Publisher: Windsor Books
ISBN: 0930233581
Type: Book - Hard Cover
Publish Date: 4/1/1998
axlq
Ian Williams <i...@IanWilliams.worldonline.co.uk> wrote:
>Does anybody know what book they are from and if any good?Many thanks
Books by Colin Alexander.
I programmed a Lindahl pattern recognizer into TradeStation some time
ago. These patterns occur rarely, and when they do occur, they often
don't work.
If anybody is interested, I could post the code here (but not until
7 January -- I'm travelling right now).
-A
axlq
Ian Williams <i...@IanWilliams.worldonline.co.uk> wrote:
They're not good for much, but here's the EL signal code for you to
play with:
{------------------------------------------------------
Lindahl pattern by Colin Alexander
Written by Alex Matulich
The Lindahl Rule
Buying: Within nine bars from the low of the formation:
(a) price must exceed the high of the bottom bar;
(b) price must then take out the low of the preceding bar;
(c) on entry day, price must take out the high the preceding bar and
close above the preceding bar's close and the current bar's open
This formation may be completed in as few as 3 bars or as many as 9.}
variables: newlowbar(10), newhibar(10), lowbar(0), hibar(0), b(0);
if low < low[1] then newlowbar = 0 else newlowbar = newlowbar[1] + 1;
if high > high[1] then newhibar = 0 else newhibar = newhibar[1] + 1;
lowbar = lowestbar(low, 9);
if low[lowbar] < low[lowbar+1] and lowbar > 1 then begin {low identified}
b = highestbar(high, lowbar-1);
if high[b] > high[lowbar] {condition (a)}
and newlowbar < b {condition (b)}
and high > high[1] and close > close[1] and close > open {condition (c)}
then buy("L") tomorrow open;
end;
hibar = highestbar(high, 9);
if high[hibar] > high[hibar+1] and hibar > 1 then begin
b = lowestbar(low, hibar-1);
if low[b] < low[hibar]
and newhibar < b
and low < low[1] and close < close[1] and close < open
then sell("S") tomorrow open;
end;
ynotssor
>
>> Does anybody know what book they are from and if any good?Many thanks
>
> They're not good for much, but here's the EL signal code for you to
> play with:
5 min NQH3, no stops, 2.40/side, 5 days data:
Gross/Net = 1251.20/1181.60 (69.60 Gross Loss)
Winners/Losers = 6/2 (75%)
Max. intraday DD = 239.60
Open Pos. = 570.00 (as of 02:15 EST)
tony
--
use hotmail.com for any email replies
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
ynotssor
>
>> They're not good for much, but here's the EL signal code for you to
>> play with:
> [...]
>
>
> 5 min NQH3, no stops, 2.40/side, 5 days data:
>
> Gross/Net = 1251.20/1181.60 (69.60 Gross Loss)
> Winners/Losers = 6/2 (75%)
> Max. intraday DD = 239.60
> Open Pos. = 570.00 (as of 02:15 EST)
Sorry, forgot to mention a small code change I made:
then buy("L") next bar at market;
then sell("S") next bar at market;
axlq
[in response to my Lindahl EL code]
>
>5 min NQH3, no stops, 2.40/side, 5 days data:
>
> Gross/Net = 1251.20/1181.60 (69.60 Gross Loss)
> Winners/Losers = 6/2 (75%)
> Max. intraday DD = 239.60
> Open Pos. = 570.00 (as of 02:15 EST)
Interesting. I don't have intraday data, so I tested it on EOD
data. Using EOD, the pattern occurs infrequently, not often enough
to take advantage of, and the ones that do appear seem to be no
better than flipping a coin.
I'd be interested in knowing how this does for a longer timeframe.
5 days with 8 trades isn't indicative of what one could expect.
I'm honestly surprised the pattern appeared 8 times in 5 days of
5-minute bars!
-A
ynotssor
> [in response to my Lindahl EL code]
[...]
> Interesting. I don't have intraday data, so I tested it on EOD
> data. Using EOD, the pattern occurs infrequently, not often enough
> to take advantage of, and the ones that do appear seem to be no
> better than flipping a coin.
>
> I'd be interested in knowing how this does for a longer timeframe.
> 5 days with 8 trades isn't indicative of what one could expect.
> I'm honestly surprised the pattern appeared 8 times in 5 days of
> 5-minute bars!
Each of the following has an open position of $1210.00 as of close on 01/17/2003:
--------------------------------------------------------
past 5 days data, 5 min NQH3, no stops, 2.40/side:
Gross/Net = 1251.20/1181.60 (69.60 Gross Loss)
Winners/Losers = 6/2 (75%)
Max. DD/trade = 170.00
Max Winner/Loser = 455.20/(34.80)
Open Pos. = 1210.00
past 10 days data, 5 min NQH3, no stops, 2.40/side:
Gross/Net = 2131.60/1652.40 (479.20 Gross Loss)
Winners/Losers = 8/4 (66.67%)
Max. intraday DD = 564.80
Max DD/trade = 430.00
Max Winner/Loser = 615.20/(274.80)
Open Pos. = 1210.00
past 15 days data, 5 min NQH3, no stops, 2.40/side:
Gross/Net = 3607.60/2724.40 (883.20 Gross Loss)
Winners/Losers = 13/9 (59.09%)
Max. intraday DD = 684.80
Max DD/trade = 530.00
Max Winner/Loser = 775.20/(504.80)
Open Pos. = 1210.00
past 24 days data, 5 min NQH3, no stops, 2.40/side (since H3 becomes "front month"):
Gross/Net = 4098.40/1021.60 (3706.80 Gross Loss)
Winners/Losers = 17/16 (51.52%)
Max. intraday DD = 1988.00
Max DD/trade = 900
Max Winner/Loser = 665.20/(834.80)
Open Pos. = 1210.00
--------------------------------------------------------
Another interesting thing is that I have made the signal optimizable:
Inputs: barsBack(9);
...
lowbar = lowestbar(low, barsBack);
...
hibar = highestbar(high, barsBack);
...
Optimizing for "All: Total net profit" 3:100 gives the default "9" as the best value .
Applied to 1,5,10,15,30,60 min. daily, weekly and monthly ( w&m used $SPX) as well
as the same fractals on ESH3 indicates that 5 min NQH3 is a "sweet spot", as all the
others were quite poor performers.
This current performance may be a chance characteristic of the market for this instrument
in the first weeks of the new year.
Or it may be what we've always read about since the time of Lyle Babbage; "Have an infinite
number of monkeys pounding away on an infinite number of typewriters and eventually they
will produce the entire works of Shakespeare."
It 's probably more like "Have an infinite number of rednecks with an infinite number of
shotguns driving around in an infinite number of pickup trucks shooting stop signs, and
eventually they will produce the entire works of Shakespeare ... in Braille."
What we have discovered here so far is the single word "look" ... but in Braille. =B^)