Deciding on a Backtesting and Trading Platform

Quote from SuperCruz:

A lot of discussions regarding the trading platform has been done in this forum. I have always wondered that to be a successful trader:
“Is it the person trading or the platform being used that makes the difference?”

Since when has a platform crafted, tested & analyzed, funded and executed a trading strategy?

In fact, I would say that most platforms as delivered are not up to the job as I would specify it. That's why I look for open extensability so that I can implement additional infrastructure to improve variosu stages of the process, ie. the platform is a tool and I prefer the ability to sharpen it!


Thx
D
 
Many of you gave valuable and useful comments regarding the query “Is it the person trading or the platform being used that makes the difference?”. Thanks for the same. :)

However, if software is to be considered, then it should be able to run on different platforms (operating systems), at least the most commonly used one like Win98, Win XP etc. Now, Windows Vista is there and rumors are abound that Win XP shall soon be withdrawn from the market.

My question is “Will TradersStudio run on Windows Vista?” Had read at the site that TradersStudio does not work with Vista, so it shall be a problem for people like us who are planning to go for Vista based computers. What is your stand on this issue, Mr. Murray?
 
Quote from SuperCruz:

Many of you gave valuable and useful comments regarding the query “Is it the person trading or the platform being used that makes the difference?”. Thanks for the same. :)

However, if software is to be considered, then it should be able to run on different platforms (operating systems), at least the most commonly used one like Win98, Win XP etc. Now, Windows Vista is there and rumors are abound that Win XP shall soon be withdrawn from the market.

My question is “Will TradersStudio run on Windows Vista?” Had read at the site that TradersStudio does not work with Vista, so it shall be a problem for people like us who are planning to go for Vista based computers. What is your stand on this issue, Mr. Murray?

The current version of TradersStudio is working fine with Vista. There were some problems reported with Vista beta's but since the release of the first Vista service packs, everything has worked perfect.

I am using TradersStudio with Vista on two different machines currently one a Intel Duo and the other an AMD Quad core. I realized because a few people pointed out we did not update the web site relating to this issue. We took care of that and now have added Vista as a supported operating system.
 
Indeed, OpenQuat's greatest strength is the event-driven model. It's too bad the rest of the product is uninspiring. It reminds me a lot of TradingSolutions in how slow it is.

As with OpenQuant, NinjaTrader unfortunately relies on C# which means huge lines of code to do even the simplest tasks. I feel that if they would provide an interpreted trading-specific language like EasyLanguage, it would become a top contender for automated trading in no time.

I reviewed TradeStation 8.3 yesterday. Sadly, I see that very little has changed from ProSuite2000i which is back when I quit using the product. RadarScreen is limited to only 1000 tickers. There's still no portfolio-level, database-level backtesting. And it still seems to be prone to crashes. It seems resource heavy too, running multiple executables. The redudancy between Strategies, Indicators, Signals, etc. makes me long for the simplicity of TradeStation v4. I will give a slight nod to RadarScreen for making real-time scanning of stocks a little more practical. The trading automation is basic.

The latest MultiCharts beta seems to be a poor man's clone of TradeStation 8.3, but without as many features implemented. I suppose the advantage is not being restricted to a single brokerage and data provider. However, the data server has incredibly annoying limitations. There is no way to add en masse all stock tickers except from OpenTick. Even if you do manage to add 7500+ tickers, who's going to sit there and individually select 7500+ tickers to add to the separate backtester product just to be able to backtest on more than one symbol? As with TradeStation, I doubt it's portolio-level backtesting either. The automation is basic so far, but I think they'll work on making it better than TradeStation's over time. Aside from the above quibbles, the greatest strength of MultiCharts is its data server and numerous vendor plugins.

TradersStudio, even if its weak in features vs the competition, seriously needs to offer a downloadable trial. It can become a serious contender if it can catch up on real time and trading automation. It also needs to support free Yahoo, MSN, Google which CSI provides with historical data (instead of earning commissions by recommending CSI?). And I have to say that CSI's data quality has gone downhill unless the fault is in the intermediaries.

CoolTrades is the buggiest, slowest and most useless trading software I have ever had the misfortunate to experience. It makes a lot of sense to me now why there was press release hype and relies on a hyperbole website. They are promoters, not programmers.

Again, I see nothing compelling to switch from AmiBroker yet. The latest version has 2x faster portfolio-level backtesting and walk forward optimization, among other new features. The only drag is I rather not have to code up my own trading automation and portfolio tracking routines, but given the limitations of the built-in basic automation in other products that you cannot customize, that's not necessarily a negative. It is no NinjaTrader on automation with its controller interface though.

Kudos to BulletTrader for supporting multiple system signals on the same contract/ticker with IB.

Quote from maxpi:
Openquant: if you are not a good C# programmer, forget it. They do not provide examples of code except on the forums, sometimes. The event driven coding is the greatest thing since sliced bread, it affords clarity of coding and should be the model for all other software. Having said that, you can make sure your coding in any other package is done in modules and approach the same level of clarity but it's not the same......... the backtester is not real speedy and the charts are just an add-on to display results of automated trades or backtests. The latest version of the software crashed my computer when I ran one of their packaged strategies on range bars, range bars are a new addition to the software and apparently aren't entirely working. They do respond to bug reports and requests for features astonishingly fast, apparently they are good enough programmers to have written maintainable software. Some data downloading features of the software I have never gotten to work properly... I had to use Ninjatrader to download the data, export it to a Microsoft format and then import it to Openquant. They have Openquant only as a way to demo their stuff and hook higher end pro traders into using their Smartquant stuff, they are not very user friendly for the little guy that is struggling with programming, I am dumping that stuff until I can take a few classes in programming, by that time they might have all the features in there and working.

Ninjatrader is coming on very strong as a market leading software. Their charting is on a par with Tradestation's imo and the chart trader feature is great for trading right off the chart, which I love, I hate typing in an order, always did, always will... I had problems with what appeared to be ambiguity between how bars were interpreted in backtesting and realtime, maybe it was my inability to understand what was going on with that but I dropped them as a backtester, I like Multicharts for backtesting. I'm not much of a C# programmer but with the examples of code provided by Ninjatrader I could do it. I can code up an indicator much faster in Ninjatrader than with "Easy Language"... The negative comments about Ninjatraders customer service seem ludicrous to me. They have, by far, the best forum for customer service. They answer questions on the forums within the hour typically. That stops kooks that would like to be on the forums and giving bad answers.

Tradestation 2000 is very prone to giving me fits, it's day has come and gone for sure. I was able to use the Radarscreen as a backtester of sorts however, I don't recall the details of how I did that, but I could test strategies on hundreds of stocks over month's worth of intraday data in fairly quick time. Managing that much data was a nightmare.

Tradestation 8 is fine really but the interface to the brokerage is poor. I abandoned it because of that. With Multicharts, Openquant, Ninjatrader, etc. you can query TWS and have account information, position information, etc. The only way I would use TS8 is if I went from trading futures to stocks. I could use the Radarscreen as an automated stock trading front end for TWS but they frown on trading with their software with another broker. They contractually disallow it in fact but I doubt they enforce it......

The latest version of Multicharts has recaptured my attention, I have a license for the software but was waiting for a better brokerage interface, which seems to have arrived. I can do everything I ever need to do in Easy Language so there is no programming limit for me in Multicharts, the charts look good, etc.

Just my two cents, overviewish look at the softwares I have played with, enjoy........ [/B]
 
Given how similar and innovative-less most platforms are and have been, it's not going to be the platform that makes the difference for non-portfolio level trading. Only a small minority of platforms drive innovation that the rest slowly adopt, if ever.

In fairness, 10 years ago comimssions were not .005 a share, so maybe the demand for innovation just wasn't there.

Quote from SuperCruz:

A lot of discussions regarding the trading platform has been done in this forum. I have always wondered that to be a successful trader:
“Is it the person trading or the platform being used that makes the difference?”
 
Yeah, where is the Open Source charting and trading automation application? It's time is sorely overdue to clean out all the hucksters and promoters in the industry.

Quote from fundjunkie:

Since when has a platform crafted, tested & analyzed, funded and executed a trading strategy?

In fact, I would say that most platforms as delivered are not up to the job as I would specify it. That's why I look for open extensability so that I can implement additional infrastructure to improve variosu stages of the process, ie. the platform is a tool and I prefer the ability to sharpen it!


Thx
D
 
Quote from MachineGhost:

TradersStudio, even if its weak in features vs the competition, seriously needs to offer a downloadable trial. It can become a serious contender if it can catch up on real time and trading automation. It also needs to support free Yahoo, MSN, Google which CSI provides with historical data (instead of earning commissions by recommending CSI?). And I have to say that CSI's data quality has gone downhill unless the fault is in the intermediaries.

What do you mean weakest feature set ? Not all products can have features everyone wants but on a point /counterpoint basis I will say TradersStudio has as strong of a feature set as products costing 10 times as much. I agree about real time and we are working on that along with a multi-core version.

In terms of data. You can use any vendor who supports, CSI,ASC (CSV,TAB,SPACE seperated) or Metastock formats. In our new version due out this summer we support the new metastock format instead of just the older ones. The one issue is for stock analysis where you need to have split only, unadjusted and dividend only adjusted to do our advanced stock analysis.
Most vendors roll dividends into splits and because of percision issues this creates errors in long backtests and they why we seperate out the streams. If other vendors allowed us to get these three streams , you could use them for our advance analysis. Of course you can use split only and use a standard stock session in TradersStudio.

In terms of free data. I don't care if people use it. There are tools for creating files and updating them from free sources. I just don't think people trading real money should count on free data. It's ok to play with but to trade real money it scares me.

I have said before why we don't offer a free demo. I don't want to offer a demo without support. See free demo's mean you can offer quality tech support to these people when they are trying to learn your program because they did not pay for it This is not in the best interest of the serious people because they don't have the resources to judge your product. Also you will have 1000's of people who would never buy it using it.

Because of this we decided to offer a return policy with an $75.00 restocking charge + postage paid to send you the product. This covers our costs to a point and allows us to offer you top quality support and printed documentation. You buy the product for $599.00 list price or $499.00 with a sale we currently have on plus shipping and you will get back $425.00 as long as you do not activate the product and return it within 30 days.

We also plan on a paid trial /tutoralware version once our new version is released in the $75.00 to $99.00 price range with credit offer toward full version.
 
This is the weakest excuse, or reason, for not offering a free demo.

Traders want to make sure the time and money spent on software is worth it - not just the cost, but the time involved in implementing a new system, or changing over to another system.

You can't expect traders to buy your product without evaluating it, can you?

If you can't offer me a demo so that I can evaluate and assess your software then I am not interested.

Don'tbe surprised if the majority of traders - your potential customers - are also not interested.
 
Quote from Jahajee:

This is the weakest excuse, or reason, for not offering a free demo.

Traders want to make sure the time and money spent on software is worth it - not just the cost, but the time involved in implementing a new system, or changing over to another system.

You can't expect traders to buy your product without evaluating it, can you?

If you can't offer me a demo so that I can evaluate and assess your software then I am not interested.

Don'tbe surprised if the majority of traders - your potential customers - are also not interested.

You have a point but don't think this is an "excuse" as you choose to call it. Anyway, I wrestled with this and decided to go with Tradersstudio's terms based on this reasoning:

"Is the value of a genuinely good product worth more to me than the cost of return should I not like it?"

Of course, the answer for me was yes and so I set aside my objections. This was helped by the fact that I already knew Tradersstudio was " significantly good" from forums like this. So the question then was to find out "how good". Passing it by wasn't option because the role it fills is critical to any retail trader with serious intent.

So, if you follow that logic the question then becomes one of how useful you believe Tradersstudio might be to you vs. the cost you're objecting to. Seeing as your primary concern is one of whether a demo is notionally free or not I assume it would not be that useful. In that case I'd suggest you may be better served by saving your money.


Thx
D

p.s. I do think Tradersstudio should revisit this policy but don't think it's worth getting on a high horse over.
 
I use both TS and Multicharts now. First, I got the TS demo and looked at it a few weeks while STILL trading with QCharts. I decided afer a few weeks that I preferred TS.

A few months later, I repeated same process with Multicharts. I am fairly satisfied with both, but I am always looking for better. This is the case with most traders.

How do you convince us to buy TradersStudio, spend weeks even 2 or 3 months running in parallel and testing, then if we don't want it ... too bad, we can't return it.

You can keep your TradersStudio.

Like I said, this is one of the very trading software that does not provide a demo. Red flag to me, don't trust it, they want to sell you and you are stuck with it. If you can't put your product out there for customers to evaluate then you are not really engendering confidence in it, so why should the customers?
 
Back
Top