How much is a good Stoploss system method worth?

pstrusi said:
My suggestion: I've seen pretty often the advantage to have a stoploss order triggered by the system logic and not by a an arbitrary amount. Perhaps include some logic in the core of your system to generate these signals might be more efficient. By the way, this what I did. Obviously there's a toll, perhaps several little loser trades but few big winners which compensates them all up
More...
Who said anything about it not having logic?
I agree with you however but not about several losers, too much work and fees.
Thanks for post.
 
Writing/coding a trendline is complex, they also make sense after the fact in hindsight. So you code up a trendline in Amibroker for example, can you transfer that to Supercharts or any other charting program? NO! Can you code it simply into Excel? NO!
Some may wish to code up a % trailing stop and use that as a trendline, well same thing applies, it's complex and for the most part unpractical.

CODING FOR TRENDLINE IN AB
//////////////chart & back ground color////////////////////
SetChartBkGradientFill( colorBlack, colorBlack, colorBlack );
//Plot( C, "Close", ParamColor("Color", colorDefault ), styleNoTitle | ParamStyle("Style") | GetPriceStyle() );
Plot( Close, "C", colorWhite, styleCandle );
separator = Day() != Ref( Day(), -1 );
Plot( separator, "", colorPink, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 5 );
starttime = 091500;
endtime = 151500;
separator = Day() != Ref( Day(), -1 );
insession = timenum() >= starttime && timenum() <= 151000;
endsession = timenum() >= endtime;
endsession = ( endsession - Ref( endsession, -1 ) ) == 1 ;//OR Ref( separator, 1 );
startsession = timenum() >= starttime;
startsession = ( startsession - Ref( startsession, -1 ) ) == 1;
timelmits = insession && !endsession;
Plot( endsession, "", colorWhite, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 1 );
Plot( startsession, "", colorViolet, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 1 );
//GraphXSpace = 85;

_SECTION_BEGIN("Long R/S");
farback = Param("How Far back to go", 100, 50, 5000, 10);
nBars = Param("Number of bars", 12, 5, 40);
aHPivs = H - H;
aLPivs = L - L;
aHPivHighs = H - H;
aLPivLows = L - L;
aHPivIdxs = H - H;
aLPivIdxs = L - L;
nHPivs = 0;
nLPivs = 0;
lastHPIdx = 0;
lastLPIdx = 0;
lastHPH = 0;
lastLPL = 0;
curPivBarIdx = 0;
aHHVBars = HHVBars(H, nBars);
aLLVBars = LLVBars(L, nBars);
aHHV = HHV(H, nBars);
aLLV = LLV(L, nBars);
aVisBars = Status("barvisible");
nLastVisBar = LastValue(Highest(IIf(aVisBars, BarIndex(), 0)));
_TRACE("Last visible bar: " + nLastVisBar);
curBar = (BarCount - 1);
curTrend = "";
if (aLLVBars[curBar] < aHHVBars[curBar])
{
curTrend = "D";
}

else
{
curTrend = "U";
}

for (i = 0; i < farback; i++)
{
curBar = (BarCount - 1) - i;
if (aLLVBars[curBar] < aHHVBars[curBar])
{
if (curTrend == "U")
{
curTrend = "D";
curPivBarIdx = curBar - aLLVBars[curBar];
aLPivs[curPivBarIdx] = 1;
aLPivLows[nLPivs] = L[curPivBarIdx];
aLPivIdxs[nLPivs] = curPivBarIdx;
nLPivs++;
}
}
else
{
if (curTrend == "D")
{
curTrend = "U";
curPivBarIdx = curBar - aHHVBars[curBar];
aHPivs[curPivBarIdx] = 1;
aHPivHighs[nHPivs] = H[curPivBarIdx];
aHPivIdxs[nHPivs] = curPivBarIdx;
nHPivs++;
}
}
}

curBar = (BarCount - 1);
candIdx = 0;
candPrc = 0;
lastLPIdx = aLPivIdxs[0];
lastLPL = aLPivLows[0];
lastHPIdx = aHPivIdxs[0];
lastHPH = aHPivHighs[0];
if (lastLPIdx > lastHPIdx)
{
candIdx = curBar - aHHVBars[curBar];
candPrc = aHHV[curBar];
if (lastHPH < candPrc AND candIdx > lastLPIdx AND candIdx < curBar)
{
aHPivs[candIdx] = 1;
for (j = 0; j < nHPivs; j++)
{
aHPivHighs[nHPivs - j] = aHPivHighs[nHPivs - (j + 1)];
aHPivIdxs[nHPivs - j] = aHPivIdxs[nHPivs - (j + 1)];
}
aHPivHighs[0] = candPrc;
aHPivIdxs[0] = candIdx;
nHPivs++;
}
}

else
{
candIdx = curBar - aLLVBars[curBar];
candPrc = aLLV[curBar];
if (lastLPL > candPrc AND candIdx > lastHPIdx AND candIdx < curBar)
{
aLPivs[candIdx] = 1;
for (j = 0; j < nLPivs; j++)
{
aLPivLows[nLPivs - j] = aLPivLows[nLPivs - (j + 1)];
aLPivIdxs[nLPivs - j] = aLPivIdxs[nLPivs - (j + 1)];
}
aLPivLows[0] = candPrc;
aLPivIdxs[0] = candIdx;
nLPivs++;
}
}

for (k = 0; k < nHPivs; k++)
{
_TRACE("High pivot no. " + k + " at barindex: " + aHPivIdxs[k] + ", " + WriteVal(ValueWhen(BarIndex() == aHPivIdxs[k], DateTime(), 1), formatDateTime) + ", " + aHPivHighs[k]);
}

a1 = ahpivs == 1;
a2 = alpivs == 1;

Para = ParamToggle("Plot Parallel Lines","Off,On");
ColorS= ParamColor("Support",colorLime);
ColorR= ParamColor("Resistance",colorRed);
x = Cum(1);
s1 = L;
s11 = H;
pS = a2 == 1;

endt = LastValue(ValueWhen(ps,x,1));
startt = LastValue(ValueWhen(ps,x,2));
ends = LastValue(ValueWhen(ps,S1,1));
starts = LastValue(ValueWhen(ps,S1,2));
dtS = endt - startt;
aS = (endS - startS) / dtS;
bS = endS;
trendlineS = aS *(x - endt) + bS;
g3 = IIf(x > startt - 10, trendlineS, Null);
Plot(g3, "", colors, styleLine+styleDashed+styleNoRescale+styleNoLabel);
pR = a1 == 1;
endt1 = LastValue(ValueWhen(pr,x,1));
startt1 = LastValue(ValueWhen(pr,x,2));
endr = LastValue(ValueWhen(pr,S11,1));
startr = LastValue(ValueWhen(pr,S11,2));
dtR = endt1 - startt1;
aR = (endR - startR) / dtR;
bR = endR;
trendlineR = aR *(x - endt1) + bR;
g4 = IIf(x > startT1 - 10, trendlineR, Null);
Plot(g4, "", colorr, styleLine+styleDashed+styleNoRescale+styleNoLabel);

_SECTION_END();
https://forum.amibroker.com/t/trendline-code-customization/8949
 
I use a large combination of factors to determine when to exit. On backtests whenever I use a single method (e.g. Chandelier Exit, Parabolic SAR, Donchian Channels) it doesn't give great results. Stop losses at a fixed price/rate/ATR multiplier also seem to yield worse outcomes if it's not conditional to some other state (e.g. lost 1% already, it's not picking up, and didn't get another buy signal from a bunch of indicators in that last x minutes).

A lot depends on whether the trade is a winner or a loser as well, and for how long it's been running.

themickey said:
Writing/coding a trendline is complex, they also make sense after the fact in hindsight
More...

It's simpler to use crossovers to state whether it's an uptrend or downtrend. There is also this "Instantaneous Trendline" indicator which can give you a hint. You can use that in combination with ADX to try to have a notion of where in the trend you are (as in: the start, middle or end of trend)

Hope this helps
 
themickey said:
Writing/coding a trendline is complex, they also make sense after the fact in hindsight. So you code up a trendline in Amibroker for example, can you transfer that to Supercharts or any other charting program? NO! Can you code it simply into Excel? NO!
Some may wish to code up a % trailing stop and use that as a trendline, well same thing applies, it's complex and for the most part unpractical.

CODING FOR TRENDLINE IN AB
//////////////chart & back ground color////////////////////
SetChartBkGradientFill( colorBlack, colorBlack, colorBlack );
//Plot( C, "Close", ParamColor("Color", colorDefault ), styleNoTitle | ParamStyle("Style") | GetPriceStyle() );
Plot( Close, "C", colorWhite, styleCandle );
separator = Day() != Ref( Day(), -1 );
Plot( separator, "", colorPink, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 5 );
starttime = 091500;
endtime = 151500;
separator = Day() != Ref( Day(), -1 );
insession = timenum() >= starttime && timenum() <= 151000;
endsession = timenum() >= endtime;
endsession = ( endsession - Ref( endsession, -1 ) ) == 1 ;//OR Ref( separator, 1 );
startsession = timenum() >= starttime;
startsession = ( startsession - Ref( startsession, -1 ) ) == 1;
timelmits = insession && !endsession;
Plot( endsession, "", colorWhite, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 1 );
Plot( startsession, "", colorViolet, styleHistogram | styleOwnScale | styleNoLabel | styleNoRescale, 0, 1, 0, -2, 1 );
//GraphXSpace = 85;

_SECTION_BEGIN("Long R/S");
farback = Param("How Far back to go", 100, 50, 5000, 10);
nBars = Param("Number of bars", 12, 5, 40);
aHPivs = H - H;
aLPivs = L - L;
aHPivHighs = H - H;
aLPivLows = L - L;
aHPivIdxs = H - H;
aLPivIdxs = L - L;
nHPivs = 0;
nLPivs = 0;
lastHPIdx = 0;
lastLPIdx = 0;
lastHPH = 0;
lastLPL = 0;
curPivBarIdx = 0;
aHHVBars = HHVBars(H, nBars);
aLLVBars = LLVBars(L, nBars);
aHHV = HHV(H, nBars);
aLLV = LLV(L, nBars);
aVisBars = Status("barvisible");
nLastVisBar = LastValue(Highest(IIf(aVisBars, BarIndex(), 0)));
_TRACE("Last visible bar: " + nLastVisBar);
curBar = (BarCount - 1);
curTrend = "";
if (aLLVBars[curBar] < aHHVBars[curBar])
{
curTrend = "D";
}

else
{
curTrend = "U";
}

for (i = 0; i < farback; i++)
{
curBar = (BarCount - 1) - i;
if (aLLVBars[curBar] < aHHVBars[curBar])
{
if (curTrend == "U")
{
curTrend = "D";
curPivBarIdx = curBar - aLLVBars[curBar];
aLPivs[curPivBarIdx] = 1;
aLPivLows[nLPivs] = L[curPivBarIdx];
aLPivIdxs[nLPivs] = curPivBarIdx;
nLPivs++;
}
}
else
{
if (curTrend == "D")
{
curTrend = "U";
curPivBarIdx = curBar - aHHVBars[curBar];
aHPivs[curPivBarIdx] = 1;
aHPivHighs[nHPivs] = H[curPivBarIdx];
aHPivIdxs[nHPivs] = curPivBarIdx;
nHPivs++;
}
}
}

curBar = (BarCount - 1);
candIdx = 0;
candPrc = 0;
lastLPIdx = aLPivIdxs[0];
lastLPL = aLPivLows[0];
lastHPIdx = aHPivIdxs[0];
lastHPH = aHPivHighs[0];
if (lastLPIdx > lastHPIdx)
{
candIdx = curBar - aHHVBars[curBar];
candPrc = aHHV[curBar];
if (lastHPH < candPrc AND candIdx > lastLPIdx AND candIdx < curBar)
{
aHPivs[candIdx] = 1;
for (j = 0; j < nHPivs; j++)
{
aHPivHighs[nHPivs - j] = aHPivHighs[nHPivs - (j + 1)];
aHPivIdxs[nHPivs - j] = aHPivIdxs[nHPivs - (j + 1)];
}
aHPivHighs[0] = candPrc;
aHPivIdxs[0] = candIdx;
nHPivs++;
}
}

else
{
candIdx = curBar - aLLVBars[curBar];
candPrc = aLLV[curBar];
if (lastLPL > candPrc AND candIdx > lastHPIdx AND candIdx < curBar)
{
aLPivs[candIdx] = 1;
for (j = 0; j < nLPivs; j++)
{
aLPivLows[nLPivs - j] = aLPivLows[nLPivs - (j + 1)];
aLPivIdxs[nLPivs - j] = aLPivIdxs[nLPivs - (j + 1)];
}
aLPivLows[0] = candPrc;
aLPivIdxs[0] = candIdx;
nLPivs++;
}
}

for (k = 0; k < nHPivs; k++)
{
_TRACE("High pivot no. " + k + " at barindex: " + aHPivIdxs[k] + ", " + WriteVal(ValueWhen(BarIndex() == aHPivIdxs[k], DateTime(), 1), formatDateTime) + ", " + aHPivHighs[k]);
}

a1 = ahpivs == 1;
a2 = alpivs == 1;

Para = ParamToggle("Plot Parallel Lines","Off,On");
ColorS= ParamColor("Support",colorLime);
ColorR= ParamColor("Resistance",colorRed);
x = Cum(1);
s1 = L;
s11 = H;
pS = a2 == 1;

endt = LastValue(ValueWhen(ps,x,1));
startt = LastValue(ValueWhen(ps,x,2));
ends = LastValue(ValueWhen(ps,S1,1));
starts = LastValue(ValueWhen(ps,S1,2));
dtS = endt - startt;
aS = (endS - startS) / dtS;
bS = endS;
trendlineS = aS *(x - endt) + bS;
g3 = IIf(x > startt - 10, trendlineS, Null);
Plot(g3, "", colors, styleLine+styleDashed+styleNoRescale+styleNoLabel);
pR = a1 == 1;
endt1 = LastValue(ValueWhen(pr,x,1));
startt1 = LastValue(ValueWhen(pr,x,2));
endr = LastValue(ValueWhen(pr,S11,1));
startr = LastValue(ValueWhen(pr,S11,2));
dtR = endt1 - startt1;
aR = (endR - startR) / dtR;
bR = endR;
trendlineR = aR *(x - endt1) + bR;
g4 = IIf(x > startT1 - 10, trendlineR, Null);
Plot(g4, "", colorr, styleLine+styleDashed+styleNoRescale+styleNoLabel);

_SECTION_END();
https://forum.amibroker.com/t/trendline-code-customization/8949
More...
What language is that?

I only know VBA, is there a translator for that?
 
themickey said:
For years I've struggled over knowing what best stoploss method to use.
I have one on Amibroker which I had coded, had it for years, it's a percentage trailing stop system, but it is clunky.
Difficult to code, restricted to Amibroker, a bit complex and time consuming in its application.

I mentor a couple of people and I must say, my entry system I'm very happy with but a good exit system - well, a bit embarrassing really.

So anyhow I created some notes on various styles of stoploss but never answered the question - "What is an ideal system?

So once again was forced to ponder the question and and with a stroke of luck came up with what I think is about the most user friendly practical and workable method.
Can be used on charts or data with minimal coding, it's the old old story, 'If you don't know it is hard, once you know it is easy'.

How many traders are there who struggle with this issue?

Nobody wants to supply their secret sauce entry method, but what about an exit system? Does revealing this create me as a trader issues?
Is there any merit in selling such a method?
More...

Nothing is perfect, so you develop multiple methods with each seeking different outcomes, scalping is 90/10 scalp/runner, hourly 1st target for 3 digit outcomes, daily shoots for 1st targets of 4 digits, long term 1st targets 5 digits. Time allows for bigger profits cause they happen less. I believe the toughest is acceptance of what you end up, it never be the best, someone else smarter with find anything better, but since I concentrate on having very low drawdowns, upside becomes less important. So instead of keeping 50% of account on sidelines to handle losses, I can utilize 85% with 15% in reserve. As you age, you have to come to realization it is easier to accept and in long range, so long as equity curve keeps going up, accept.
 
ironchef said:
What language is that?

I only know VBA, is there a translator for that?
More...
It's called AFL (Amibroker formula language).
I know of no direct translator, but I would imagine many good coders could look at it and suss most of it out.
That particular code I just copied from internet, it's not something I use.
The general opinion seems to be, AFL is not an easy language to understand and that too is my experience, I think the software is great but understanding the coding difficult.
Even if you go online looking for answers, many of the AFL crowd be making comments; "work it out yourself, that's the only way you'll learn".
 
Handle123 said:
Nothing is perfect, so you develop multiple methods with each seeking different outcomes, scalping is 90/10 scalp/runner, hourly 1st target for 3 digit outcomes, daily shoots for 1st targets of 4 digits, long term 1st targets 5 digits. Time allows for bigger profits cause they happen less. I believe the toughest is acceptance of what you end up, it never be the best, someone else smarter with find anything better, but since I concentrate on having very low drawdowns, upside becomes less important. So instead of keeping 50% of account on sidelines to handle losses, I can utilize 85% with 15% in reserve. As you age, you have to come to realization it is easier to accept and in long range, so long as equity curve keeps going up, accept.
More...
And 'equity curve keeps going up' is the result of two predominant things, good entries and good exits.
The former I'm ok on, the latter not so good previously. The biggest risk period seems to be entries, because if you get it wrong there you go into an immediate loss.
Entries is most vulnerable due to mindset you have it right, when in fact you have it wrong.
 
themickey said:
And 'equity curve keeps going up' is the result of two predominant things, good entries and good exits.
The former I'm ok on, the latter not so good previously. The biggest risk period seems to be entries, because if you get it wrong there you go into an immediate loss.
Entries is most vulnerable due to mindset you have it right, when in fact you have it wrong.
More...

I've often seen, backtesting several models, that it's always more profitable a Strategy or Model that always has a position than others that includes Flat positions. Intuitively, it seems a Flat position might be better as a defense mechanism, but the statistic evidence shows otherwise.
 
pstrusi said:
I've often seen, backtesting several models, that it's always more profitable a Strategy or Model that always has a position than others that includes Flat positions. Intuitively, it seems a Flat position might be better as a defense mechanism, but the statistic evidence shows otherwise.
More...
Is this for long & short trading?
 
You must log in or register to reply here.
Back
Top