Thank you, Mrtools!
John L Person presented a pivot based system in his book Forex Conquered-High Probability Systems and Strategies for Active Traders. I attached below the Easy Language codes for this system from the book. I'm not familiar with Easy Language. The backtest result is quite impressive. However, it seems to me he used data from "current" day in the codes. Hence I don't think the EA is robust. Could you please take a look at the codes and confirm if data from current day is used ? Thanks.
Inputs: NumPIPs( 10 ), StpPIPs( 100 ), PrftPIPs( 40 ), StochBuyPercent( 22 ),
StochSellPercent( 72 ), StochLength( 14 ), SmoothingLength1( 3 ),
SmoothingLength2( 3 );
Variables: S1( 0 ), S2( 0 ), S3( 0 ), R1( 0 ), R2( 0 ), R3( 0 ), PP( 0 ), TodaysHigh(
0 ), YestHigh( 0 ), TodaysLow( 0 ), YestLow( 0 ), TodaysClose( 0 ), YestClose(
0 ),oFastK( 0 ), oFastD( 0 ), oSlowK( 0 ), oSlowD( 0 ) ;
If Time = SessionFirstBarTime(1, 1) then begin
YestHigh = TodaysHigh ;
YestLow = TodaysLow ;
YestClose = Close ;
TodaysHigh = High ;
TodaysLow = Low ;
PP = ( YestHigh + YestLow + YestClose ) / 3 ;
R1 = PP * 2 – YestLow ;
R2 = PP + YestHigh – YestLow ;
R3 = R2 + YestHigh – YestLow ;
S1 = PP * 2 – YestHigh ;
S2 = PP – YestHigh + YestLow ;
S3 = S2 – YestHigh + YestLow ;
if High > TodaysHigh then
if Low < TodaysLow then
Value1 = Stochastic( High, Low, Close, StochLength, SmoothingLength1,
SmoothingLength2, 1, oFastK, oFastD, oSlowK, oSlowD ) ;
If MRO(oFastK crosses over StochBuyPercent, 5, 1) > –1 and MRO(oFastD
crosses over StochBuyPercent, 5, 1) > –1 and
oFastK > StochBuyPercent and oFastD > StochBuyPercent and
(Low > S1 – NumPIPs points and Low < S1 + NumPIPs points) or
(Low > S2 – NumPIPs points and Low < S2 + NumPIPs points) or
(Low > S3 – NumPIPs points and Low < S3 + NumPIPs points) then
Buy next bar at market;
If MRO(oFastK crosses under StochSellPercent, 5, 1) > –1 and MRO(oFastD
crosses under StochSellPercent, 5, 1) > –1 and
oFastK < StochSellPercent and oFastD < StochSellPercent and
(High > R1 – NumPIPs points and High < R1 + NumPIPs points) or
(High > R2 – NumPIPs points and High < R2 + NumPIPs points) or
Trading Systems: Combining Pivots with Indicators 207
(High > R3 – NumPIPs points and High < R3 + NumPIPs points) then
Sellshort next bar at market;
SetStopLoss(StpPIPs points * BigPointValue);
SetProfitTarget(PrftPIPs points * BigPointValue);
Default Values for Rule Inputs
numPIPs = 10
StochBuyPercent = 22
That is the way how it "remembers" current days high and low to be used for next day pivots calculations, so it works OK as far as that is concerned (no flaw in that part)
What I don't see from the equity curve graph is what instrument are those tests results for. In my experience results of using pivot points on forex symbols are not as good as when they are applied to stocks
From context in the book, the instrument tested on is forex. However, he didn't specify which pair it is.
I attached his test report from the book and the EA for MT4 that I coded based on his description of the system. The report shows a winning % of 67.1%. From his description, I found the system only has two exit modes: take profit and stop loss. However, given a take profit vs stop loss of 40 pips vs 100 pips, and a win vs loss % of 67.1% vs 32.%, how is it possible for the system to be profitable in the first place, let alone to be so amazingly profitable as the report shows ? The attached EA I coded for the same system has a winning % of 69% during a 10 year backtest, but my EA only generates net loss during the backtest. However, I know almost nothing about Easy Language codes. Maybe I missed something ? Does his EA really only have the two exit modes like I said ? Do you think this test report can be telling the truth ?
Spread Indication by Bar
Has anybody seen anything like this? If not, anybody up to taking a crack at it?
Without depth of orders or meaningful volume, it is impossible to really characterize accumulation and distribution in forex instrument moves on our MT4 platform.
After watching price action, the Spread'O'Meter and thinking about it a bit, I wonder...
Craft an indicator that shows the spread for the current bar.
Perhaps a stacked histogram with levels of the histo assigned to meaningful divisions (e.g., thirds of a bar where 1st third is 0-2 pips, 2nd third is 2-4 and last third is 4+) with color intensity (blue to red?) to reflect the distribution. (Mladen's creativity and use of buffers would not doubt be in play here).
Or, put the distributions in buffers similarly assigned (e.g., one buffer with running percentage of ticks wherein the spread is 0-2 pips, another buffer is for 2-4 pips and so forth).
Perhaps the sophistication of the indicator is such that it can scale assignments/distributions based on the 'normal' or 'nominal' spread?
The idea, of course, is to be able to query the buffer(s) in order to learn when the move is meaningful because it has broad market participation.
Of course, the other side could be exploited as well; that is, on a low 'volume' move, one could fade the move as it did not had broad market participation.
I need this indi to be a MTF version, and let me have multi indicators(2 more same indis on one chart).
I think using the "UniqueID" on parameter will be useful. of course, if you have alternative of the way, it is better.
thanks in advance.
Here you go No need for unqueID in this one since it does not us objects. Decided not to use objects for the sake of simplicity and since in this one multi time framing the future (extrapolated values) is not as critical as it was in Goertzel browser for example (you will notice the connection of the first future bars with a current bar, but that is the only deviation from the "perfect" connection and is a concequence of interpolating values when there is no mor values to interpolate)
On Chart Rsi
Hmm I wonder if something more can be done for/with this,perhaps a dynamic version with the usual options.Was 4(cutler,rsx,etc...) now 5 (T3 included,please!) If this is possible this indy will become awesome but if I missed something and there is already one floating around,direct me,thanks Max
Goor morning Mladen,
This is the first trendline indicator I like after extensive research (from one of the free sections here actually).
However i would appreciate if some changes can be done to it:
- Make color, width, style of lines fully editable according to their priority (3 kind of priorities I believe)
- Make alarms for trendlines/according to priority
- Option to hide visualization of older/no-more relevant trendlines
- Get rid of orange vertical line
- Get rid of strange letters on top right
Hope this is to the liking of several here.
Thank you and have a nice weekend!
OnChart rsi 2 ...
Here is this version with the 5 additional rsi types added.
0 - rsi
1 - Wilder's rsi
2 - rsx
3 - Cuttler's rsi
4 - rsi T3
1 - Wilder's rsi
2 - rsx
3 - Cuttler's rsi
4 - rsi T3