RealTest - Portfolio Level backtesting software

[houtman]

What data source are you using to populate RealTest - or to put it another way what fuel are you putting into your Lamborghini?

Just Yahoo.

It's free, if I ever get Realtest out of the driveway I'd look at something better.

I'm about 95 %FA, 5% TA atm so Yahoo is fine.

 

[Sid23]

I use Realtest.

It's main strengths appear to be running multiple strategies and simplicity of coding.

The simplicity was why I got it.

I only use it in the most basic way, probably comparable to sitting in a Lamborghini in the drive way while making car noises.

I've got no coding abilities, but after a week of a few hours a night had a decent understanding of how it works.

Those are also the 2 main reasons that sparked my interest. I did code up a simple BBO system 18 months ago (made a few updates since) & bought Nick's WTT turnkey (with a few parameter changes) 12 months ago but find it difficult to grasp coding with things like looping and "If" statements. So ease of coding would be a big plus.

Slowly working towards multiple systems (when/if funds permit, currently 3 systems), having a system that includes multiple strategies that can be backtested would be worth having, such as testing a US & Aus strategy together to balance out the system.

@Warr87 received email newsletter from Chartist today, pretty sure you don't need to be a member to access it.

RealTest Free Trial

RealTest Intuitive multi-strategy backtesting software. Download your free 30 day trial

www.thechartist.com.au

Sid

 

[Bluboy1969]

Would be good to see if they can add Tiingo to there list of data providers for the US data, its cheap and more accurate that Yahoo I am told. I use it at the moment through Amibroker.

Stock Market Tools | Tiingo

A financial research platform dedicated to creating innovative financial tools for all, while adopting the motto, "Actively Do Good."

www.tiingo.com

 

[Nick Radge]

Nick Radge just confirmed (via video on website) he is moving to RT beginning next year.

@Sid23 I've already ported my own systems to RT and moving the company systems as we speak. All turnkey strats have been ported and will be available soon. Hopefully we'll add some more.

The mentor course will be rewritten this year and officially launched with the RT changes in Jan 2024

-Nick

 

[BoNeZ]

After 12 months, are you solely using RT and if so, was it worth the change?

Yes and Yes.

Amibroker is still installed on my PC and I occasionally use it to check a single stock but all my development, testing and trading is done using RealTest.

My RT code is simpler, smaller and runs faster plus support from the RT forum or direct from Marsten is rapid and always helpful.

I don't think I could go back.

 

[BoNeZ]

Would be good to see if they can add Tiingo to there list of data providers for the US data,

It already supports Tiingo and TiingoCrypto. You need to edit the Realtest.ini file to add your personal API key

There is a section in the manual plus a sample script which both have instructions.

 

[Sid23]

@Sid23 I've already ported my own systems to RT and moving the company systems as we speak. All turnkey strats have been ported and will be available soon. Hopefully we'll add some more.

The mentor course will be rewritten this year and officially launched with the RT changes in Jan 2024

-Nick

Nick, will turnkey for Amibroker be available till end of year only? Just incase I stick to AB, though strong case for moving to RT.

Will evaluate trial, may have to wait & see if tax refund this year to buy it, as after last year performance unlikely I'll owe. ?

Sid

 

[Warr87]

@Sid23 I've already ported my own systems to RT and moving the company systems as we speak. All turnkey strats have been ported and will be available soon. Hopefully we'll add some more.

The mentor course will be rewritten this year and officially launched with the RT changes in Jan 2024

-Nick

Are you making the RT versions available for those who previously purchased your turnkey strats?

 

[Nick Radge]

Nick, will turnkey for Amibroker be available till end of year only? Just incase I stick to AB, though strong case for moving to RT.

Will evaluate trial, may have to wait & see if tax refund this year to buy it, as after last year performance unlikely I'll owe. ?

Sid

Amibroker code isn't being removed, we'll continue to support it, so no stress for those who want to continue with AB.

Turnkey code will be available to prior buyers but with a nominal cost, yet to be determined.

The new RT files, where applicable, will enable the user to test either ASX, US, or the combination of both.

 

[MovingAverage]

Anyone here making serious use of RTs monte carlo capability? Would like to hear your thoughts.
AB monte is completely useless at best so I'm currently exporting trade data from AB to TradeSim for MC testing. TradeSim is great for MC testing and does it properly. I've looked through the RT doc's for specifics about how it's doing MC testing but can't find much detail.

 

[Roller_1]

Anyone here making serious use of RTs monte carlo capability? Would like to hear your thoughts.
AB monte is completely useless at best so I'm currently exporting trade data from AB to TradeSim for MC testing. TradeSim is great for MC testing and does it properly. I've looked through the RT doc's for specifics about how it's doing MC testing but can't find much detail.

​

I have used it a fair bit. What are you wanting to test? I have used it for trade skipping, data variance, signal variance things like that. It's so fast that you can run these tests a lot quicker than AB

 

[MovingAverage]

​

I have used it a fair bit. What are you wanting to test? I have used it for trade skipping, data variance, signal variance things like that. It's so fast that you can run these tests a lot quicker than AB

I'm curious about how it does the MC test? In AB all it does is takes a SINGLE trade set and then shuffles those trades around in time--that's ridulous. The way I'm currently doing it is to dump ALL possible trades into TradeSim, which then will randomly generate DIFFERENT trade sets to give a better MC analysis. I'm curious to know whether RT does the same approach as AB or it is more like the TradeSim approach?

 

[Roller_1]

I'm curious about how it does the MC test? In AB all it does is takes a SINGLE trade set and then shuffles those trades around in time--that's ridulous. The way I'm currently doing it is to dump ALL possible trades into TradeSim, which then will randomly generate DIFFERENT trade sets to give a better MC analysis. I'm curious to know whether RT does the same approach as AB or it is more like the TradeSim approach?

Oh sorry, I misunderstood your question. I more use the optimiser to test different scenarios.

Like you said with AB I never like that either. Doesn't really make sense to me.

"in this case, 100 random cumulative profit curves were generated by taking random samples (with replacement) from the trade list"

This was quoted from the guide. Which to me seems like it is doing a similar thing.

 

[MovingAverage]

Oh sorry, I misunderstood your question. I more use the optimiser to test different scenarios.

Like you said with AB I never like that either. Doesn't really make sense to me.

"in this case, 100 random cumulative profit curves were generated by taking random samples (with replacement) from the trade list"

This was quoted from the guide. Which to me seems like it is doing a similar thing.

Thanks...all good. I'm not a member of the RT forum so will do that as I'm sure there will probably more detail there. Thanks again.

 

[Roller_1]

Thanks...all good. I'm not a member of the RT forum so will do that as I'm sure there will probably more detail there. Thanks again.

I just had a quick two minute search of the forum and there is some posts about it on there. Basically it is doing similar to AB, rearranging the trades. So make of this what you will.

I would advise joining up there and asking some more in-depth questions as you will get some better answers than I can give.

I think the more important question is to ask what happens if I took trades that are not included in the back test. Like doing trades skipping and taking trades that may not meet the initial criteria but by tweaking the data or signals then you can take them to see how that affects results.

 

[MovingAverage]

I think the more important question is to ask what happens if I took trades that are not included in the back test. Like doing trades skipping and taking trades that may not meet the initial criteria but by tweaking the data or signals then you can take them to see how that affects results.

This is exactly why I hate AB MC. TradeSim addresses the issue you highlight by generating different sets of trades.

 

[Warr87]

I haven't looked at the MC in RT, but if you have suggestions for Marsten to improve it or want more info then he is very accommodating. If he can't, he'll tell you why. I never liked the AB MC but that's because I found it confusing at times and assumed it was more me than AB. (Probably both! haha.)

 

[MovingAverage]

I haven't looked at the MC in RT, but if you have suggestions for Marsten to improve it or want more info then he is very accommodating. If he can't, he'll tell you why. I never liked the AB MC but that's because I found it confusing at times and assumed it was more me than AB. (Probably both! haha.)

I've been wedded to AB for a long time, but will be biting the bullet and moving across to RT. There is a lot I like about AB so will probably continue to use AB in parallel with RT, but RT's ability to easily test and evaluate systems leaves AB for dead.

 

[Roller_1]

I've been wedded to AB for a long time, but will be biting the bullet and moving across to RT. There is a lot I like about AB so will probably continue to use AB in parallel with RT, but RT's ability to easily test and evaluate systems leaves AB for dead.

i use AB for charting and double checking ideas sometimes

 

[MovingAverage]

i use AB for charting and double checking ideas sometimes

would you say AB is better for charting?