The First Day of My New Trading Business

[Xiao si 10]

Equity Curve of the final version ready for testing.

 

Anyone wishing to test this system on the Hang Seng, please contact me.

 

Cheers,

 

 

XS

[Xiao si 10]

I must add this is a Ninjatrader script.

 

Cheers,

 

 

XS

[Xiao si 10]

A brief update on my progress:

 

We have focused on systematic trading, suits my personality better and i enjoy the creative side to it.

 

We now have one system we are testing on AUTO in the RT prices but still on the simulator.

 

I have leased two systems to provide some income to cover expenses during the time my systems are in development.

 

Using Apadtrade's Builder, i have several systems that i am testing on historical data.

 

Still focusing on two markets for my systems, we should have another two systems ready for Live testing by Monday.

 

Hopefully if they test well over the coming months we should be able to go live on a real account in September.

 

I have a long list of lesson's learned that i will present once i deem myself successful, in hopes that other new traders won't make the same mistakes.

 

Cheers,

 

 

XS

[Xiao si 10]

You know, the other thread about 'see my beautiful new trading system' or whatever, got me thinking...and after the last 6- 8 months or whatever its been now, i know why so many 'traders' sell their systems....because its way easier to sell or lease systems, courses, trade rooms etc., than it is to make money on the markets consistently. Its EASY money because there are fools like me who want that holy grail and thought that it was possible.

 

Even with systems that you can test on OOS data and deem trade-able your self, you still need to stick with them through draw-downs. Can you do that? I bet many people that lease systems eventually stop leasing once the draw-downs kick in and reach their pain threshold. I would imagine the vendors count on this, and i would love to see the stats on how many customers stay for the long haul.

 

I could register a domain name,spend 5-10 grand on an eCommerce website and do the same thing quite easily. Maybe someday i will. To do that before i become a successful trader would only be kidding myself. It would be an admission that i wasn't good enough to make my money out of the market and i had to make it on the fringe, from 'wanna be's' just like me. Nope...Not yet...

 

Back to work...perseverance, determination, good research and a little luck...you never know maybe i will be trading one of my system's live on a real account and making money from it one day.

 

XS

[Tams 20]

if you have developed a profitable system,

would you sell it?

would you lease it?

 

why would you?

why would you not?

 

 

if your system is profitable,

with the high leverage of futures,

you can make money yourself.

Why go through all the trouble of marketing, customer service, support, etc.?

 

even if your system can only make US$100 per day,

if the draw down is manageable,

if the profit is consistent,

you can easily scale it up and make more money.

 

If your system cannot make consistent US$100 per day,

there is no point in continuing.

so people put the system up for lease. ;-)

[Xiao si 10]

if you have developed a profitable system,

would you sell it?

would you lease it?

 

why would you?

why would you not?

 

 

if your system is profitable,

with the high leverage of futures,

you can make money yourself.

Why go through all the trouble of marketing, customer service, support, etc.?

 

even if your system can only make US$100 per day,

if the draw down is manageable,

if the profit is consistent,

you can easily scale it up and make more money.

 

If your system cannot make consistent US$100 per day,

there is no point in continuing.

so people put the system up for lease. ;-)

 

If my system makes what the tests say it should, and it trades as it tests and i trade the max contracts that i am comfortable with, why would i lease it for 50 bucks month? Even if i limited the leases to say...20. That's another 1000 bucks a month, hardly worth the hassles of dealing with customers in pain, so yeah i agree.

 

The other thing that strikes me when browsing systems for sale or lease is the lack of commissions and slippage on the BT reports...why would you exclude this when you wouldn't do it yourself...there can be only one reason...to make the results look better than they actually are.

 

Its a mean old world out there!

 

XS

 

Some 'free' code for the SPI....use on a SPI 15 min chart. Enjoy!

 

{Easy Language__

 

Max bars back: 97

 

Created by: Xiao Si

Created: March/2012

 

Compatible with TradeStation 6 or newer

 

Data File: F:\Data - Historical ASCII\SPI\SPI 15 min.txt

Build Dates: 1/07/2010 to 18/04/2011

 

Project File: SPI001_15 min

}

 

{ Strategy inputs }

Inputs: NL1 (53),

NL2 (35),

NL3 (3),

NL4 (46),

NL5 (15),

NL6 (42),

NS1 (22),

NS2 (82),

NS3 (3),

NATREnL (33),

EntFrL (2.2916),

MMStopSz (450.00),

NBarExL (55),

NATREnS (57),

EntFrS (0.6837),

TimeExS (1405),

NTargS1 (3),

NTargS2 (98),

TargFrS (1.2183),

NBarExS (12);

 

{ Variables for average true range for entry and exit orders }

Var: ATREnL (0),

ATREnS (0);

 

{ Variables for entry and exit prices }

Var: EntPrL (0),

EntPrS (0),

TargPrS (0);

 

{ Variables for entry and exit conditions }

Var: EntCondL (false),

EntCondS (false),

EndofSess (false);

 

{ Average true range }

ATREnL = AvgTrueRange(NATREnL);

ATREnS = AvgTrueRange(NATREnS);

 

{ Entry prices }

EntPrL = CloseD(1) - EntFrL * ATREnL;

EntPrS = C + EntFrS * ATREnS;

 

{ Entry and exit conditions }

EntCondL = TriAverage(StandardDev(C, NL1, 1), NL2) >= XAverage(WAverage(Highest(Momentum(H, NL3), NL4), NL5), NL6);

EntCondS = L[NS1] >= WAverage(Average(L, NS2), NS3);

EndofSess = false;

If DataCompression >= 1 and DataCompression <= 4 then

EndofSess = time = SessionEndTime(0, 1);

 

{ Entry orders }

If (MarketPosition = 0 or EndofSess) and (EntriesToday(Date) < 4 or EndofSess) and EntCondL then begin

Buy next bar at EntPrL limit;

end;

 

If (MarketPosition = 0 or EndofSess) and (EntriesToday(Date) < 4 or EndofSess) and EntCondS then begin

Sell short next bar at EntPrS limit;

end;

 

{ Exit orders, long trades }

If MarketPosition > 0 and EndofSess = false then begin

 

If BarsSinceEntry >= NBarExL then

Sell next bar at market;

end;

 

{ Exit orders, short trades }

If MarketPosition < 0 and EndofSess = false then begin

 

If time >= TimeExS then

Buy to cover next bar at market;

 

If BarsSinceEntry >= NBarExS then

Buy to cover next bar at market;

 

TargPrS = EntryPrice - TargFrS * AbsValue(XAverage(O, NTargS1) - Highest(L, NTargS2));

Buy to cover next bar at TargPrS limit;

end;

 

SetStopLoss(MMStopSz);

 

SetExitOnClose;

[Tams 20]

If my system makes what the tests say it should, and it trades as it tests and i trade the max contracts that i am comfortable with, why would i lease it for 50 bucks month? Even if i limited the leases to say...20. That's another 1000 bucks a month, hardly worth the hassles of dealing with customers in pain, so yeah i agree.

 

The other thing that strikes me when browsing systems for sale or lease is the lack of commissions and slippage on the BT reports...why would you exclude this when you wouldn't do it yourself...there can be only one reason...to make the results look better than they actually are.

 

Its a mean old world out there!

 

XS

 

Some 'free' code for the SPI....use on a SPI 15 min chart. Enjoy!

...

 

 

 

 

It is a naive world out there

 

 

 

 

 

 

ps. please use code tag when posting codes. That's the icon at the top of the message window frame.

[Xiao si 10]

It is a naive world out there

 

 

 

 

 

 

ps. please use code tag when posting codes. That's the icon at the top of the message window frame.

 

ok cool

 

{ Strategy inputs }
Inputs: NL1 (2),
       NS1 (1110),
       FirstTradeTm (950),
       LastTradeTm (1528),
       NBarExL (62),
       MMStopSzL (675.00),
       MMStopSzS (350.00),
       NBarExS (41);

{ Variables for entry and exit prices }
Var:    EntPrL   (0),
       LStop    (0),
       SStop    (0);

{ Variables for entry and exit conditions }
Var:    EntCondL  (false),
       EntCondS  (false),
       EndofSess (false),
       TimeOK    (false);

{ Entry prices }
EntPrL = HighD(0);

{ Entry and exit conditions }
EntCondL = XAverage(O, NL1) > LowD(0);
EntCondS = Time < NS1;
TimeOK = time >= FirstTradeTm and time < LastTradeTm;
EndofSess = false;
If DataCompression >= 1 and DataCompression <= 4 then
  EndofSess = time = SessionEndTime(0, 1);

{ Entry orders }
If (MarketPosition = 0 or EndofSess) and (EntriesToday(Date) < 3 or EndofSess) and EntCondL and TimeOK then begin
  Buy next bar at EntPrL stop;
end;

If (MarketPosition = 0 or EndofSess) and (EntriesToday(Date) < 3 or EndofSess) and EntCondS and TimeOK then begin
  Sell short next bar at market;
end;

{ Exit orders, long trades }
If MarketPosition > 0 and EndofSess = false then begin

  If BarsSinceEntry = 0 then begin
     LStop = EntryPrice - MMStopSzL/(BigPointValue * CurrentShares);
  end;

  If BarsSinceEntry >= NBarExL then
     Sell next bar at market;

  Sell next bar at LStop stop;

  If TimeOK = false then
     Sell next bar at market;
end;

{ Exit orders, short trades }
If MarketPosition < 0 and EndofSess = false then begin

  If BarsSinceEntry = 0 then begin
     SStop = EntryPrice + MMStopSzS/(BigPointValue * CurrentShares);
  end;

  If BarsSinceEntry >= NBarExS then
     Buy to cover next bar at market;

  Buy to cover next bar at SStop stop;

  If TimeOK = false then
     Buy to cover next bar at market;
end;

SetExitOnClose;

Try this one...

Forgot commissions and slippage...hehe:roll eyes:

[joshdance 10]

Xiao Si, would you call trying to buy other people's trading systems running a trading business?

[Xiao si 10]

Xiao Si, would you call trying to buy other people's trading systems running a trading business?

 

 

Well its not my perfect idea of a trading business, but it wouldn't be much different than using your own systems in the tax mans view IMO. I am leasing two algo's now to try and earn some extra income and gain some experience while i develop my own, so they're part of my business now, yeah.

 

Ideally i would develop my own strats and only use them myself.

 

XS

[Xiao si 10]

Well holy cow, check out my system! No, its not for SALE!

 

This is what hard work, perseverance, and determination can yield!

 

XS

[Elitny 10]

Well holy cow, check out my system! No, its not for SALE!

 

This is what hard work, perseverance, and determination can yield!

 

XS

 

 

 

I guess with those kinds of returns you should be able to lease a bloomberg terminal

[Xiao si 10]

I guess with those kinds of returns you should be able to lease a bloomberg terminal

 

????

 

Here are some results on OOS data.

[onesmith 10]

The scatter graph shows a 1k stop loss. I am not aware of the precise details of the problems with the built-in stop loss and target keywords because I stopped using them when I found out they had problems. There are people using those things correctly, but the risk of error is not worth the time saved verses the time it takes to write something specific that implements their logic without their inherent defects.

 

Examining these stats further ... shows you've optimized the stop loss separately for longs at 980 and 755 for shorts. If you're doing this using the built-in keywords .. good luck transfering their logic into real life trading.

[Tams 20]

The scatter graph shows a 1k stop loss. I am not aware of the precise details of the problems with the built-in stop loss and target keywords because I stopped using them when I found out they had problems. There are people using those things correctly, but the risk of error is not worth the time saved verses the time it takes to write something specific that implements their logic without their inherent defects.

 

Examining these stats further ... shows you've optimized the stop loss separately for longs at 980 and 755 for shorts. If you're doing this using the built-in keywords .. good luck transfering their logic into real life trading.

 

there is no problem with the built-in stop loss and target keywords,

most people make the mistake of putting them inside conditional loops.

[onesmith 10]

Most people are not aware of the need to enable look inside bar backtesting. That reduces but does not eliminate the problems. It is still possible to write strategies that take advantage of the defects inherent within the stop and target logic.

[Xiao si 10]

Holy smokes can you guys bring the conversation down to my level so i know whats going on?

 

Cheers,

 

 

XS

[Xiao si 10]

Most people are not aware of the need to enable look inside bar backtesting. That reduces but does not eliminate the problems. It is still possible to write strategies that take advantage of the defects inherent within the stop and target logic.

 

I do understand this...but I've not yet backtested any strategy using look inside bar.

 

Should i do this? How critical is this on lower timeframes?

 

XS

[Tams 20]

I do understand this...but I've not yet backtested any strategy using look inside bar.

 

Should i do this? How critical is this on lower timeframes?

 

XS

 

if he takes the trouble to bring it to your attention,

do you think it can be important?

[Xiao si 10]

if he takes the trouble to bring it to your attention,

do you think it can be important?

 

Of course it must be important Tams:roll eyes:, but in reference to the lower time frames how important is it? It seems to me that the higher the time frame the more important it would be, but if you are trading 1 minute bars then maybe not as much of an issue, depending on volitility etc...

 

XS

 

Using the bar magnifier settings, set to 1 minute, there is no difference in the test results....have i missed something?

 

XS

Edited April 5, 2012 by Xiao si

[Xiao si 10]

Here is my latest problem....my system places an order but doesnt get a fill on the IB sim.....

 

Err, sorry..creates an order but does not send it to the sim.

Edited April 5, 2012 by Xiao si

[BlueHorseshoe 10]

in reference to the lower time frames how important is it? It seems to me that the higher the time frame the more important it would be, but if you are trading 1 minute bars then maybe not as much of an issue, depending on volitility etc...

 

It's probably more important in lower timeframes. Say that your strategy has a profit of factor of 1.8 in both daily and 1 minute timeframes. What does that mean in real dollar terms? In the daily timeframe it means that you can afford a point or two (or a few dollars) of error in your calculations (although I would still recommend that you try and avoid this). In the 1 minute timeframe the profit factor of 1.8 might translate to an average trade profit of just a few dollars - if you have miscalculated in this scenario by not using look-inside-bar resolutions (or any of a host of other potential pitfalls) then that few dollars lost will equal your entire profit margin and then some.

 

If you're unclear on what I mean by this then let me know and I will provide code for an EL strategy that works great without look-inside-bar . . . And then falls apart as soon as you test at a higher resolution.

 

Finally, if you are looking at one minute bars then your resolution definitely needs to be single-tick, and you'd better make damned sure that there are no bad ticks in that data - a fund trading intraday will employ staff whose sole responsibility is the development and implementation of algorithms to control the quality of the tick data that their systems are exposed to.

 

Hope that helps

 

Bluehorseshoe

Edited April 5, 2012 by BlueHorseshoe

[Xiao si 10]

It's probably more important in lower timeframes. Say that your strategy has a profit of factor of 1.8 in both daily and 1 minute timeframes. What does that mean in real dollar terms? In the daily timeframe it means that you can afford a point or two (or a few dollars) of error in your calculations (although I would still recommend that you try and avoid this). In the 1 minute timeframe the profit factor of 1.8 might translate to an average trade profit of just a few dollars - if you have miscalculated in this scenario by not using look-inside-bar resolutions (or any of a host of other potential pitfalls) then that few dollars lost will equal your entire profit margin and then some.

 

If you're unclear on what I mean by this then let me know and I will provide code for an EL strategy that works great without look-inside-bar . . . And then falls apart as soon as you test at a higher resolution.

 

Finally, if you are looking at one minute bars then your resolution definitely needs to be single-tick, and you'd better make damned sure that there are no bad ticks in that data - a fund trading intraday will employ staff whose sole responsibility is the development and implementation of algorithms to control the quality of the tick data that their systems are exposed to.

 

Hope that helps

 

Bluehorseshoe

 

 

Thanks for that BH, very insightful.

 

So far my testing and development has shown me that system develop on anything lessthan 15 minute are less robust and hold up poorly over more oos testing and less so in WF.

 

When i have started looking inside bar from 15 minutes i have no difference at all in results...???

 

XS

[Xiao si 10]

Here is my latest problem....my system places an order but doesnt get a fill on the IB sim.....

 

Err, sorry..creates an order but does not send it to the sim.

 

I haven't tested yet because even the Hang Seng is closed for Easter, i think it was my mistake...i simply didn't have the IB chart loaded, only eSignal....I'm still thinking in NT (which i use as well) where you have the primary data source then the broker for orders...

 

Pretty sure this will be the issue.

 

Doh!

 

XS