Tick Size, Tick Value, Currency Conversion, Data Management, Multiple Time Frame Analysis > Members Forums


Monday, January 05, 2009 Login \| Register

TradersStudio 2.5 Now!

TradersStudio Forums

Forums

Subject: Tick Size, Tick Value, Currency Conversion, Data Management, Multiple Time Frame Analysis

You are not authorized to post a reply.

Author

Messages

bernard
Posts:3

06-01-2006 11:42 PM	Alert
I'm still in the process of going through the manual. Just by glancing the table of content, I'm unable to find the answer for some issues... NQ's minimum tick size has recently changed from 0.5 to 0.25, and the tick value has also cut in half. A product can have multiple changes like the above in the last 20 years. How do I prepare my data to deal with such issue? Can you provide examples for building models with more than 2 time frames to determine the trading signal? Can you provide examples for currency conversions? I have access to lots of data (tick, daily, tick size, currency, news, etc) for a short period of time, but I don't know what to get and how to manage them to take advantage of TradersStudio's functions. This is very important to me, I want to get all the data before my data access is taken away. Thanks!

murray
Posts:435

06-02-2006 12:54 AM	Alert
Currently we do not handle multiple tick sizes. I do not know of any product which currently does. Do you have any ideas on how to do that in terms of the interface you would want to see You can change the values for tick size in the PreDefinedCommodity.mdb. This Database includes all of this plus margins. You can zip the file and upload it if you want to share it with others. In terms of the currency conversion files look at the sample data in C:\Program Files\TradersStudio\TradersStudio20\SampleData\exchange We also have the directory with non-us futures there.

bernard
Posts:3

06-02-2006 4:54 PM	Alert
I guess it would be better if you guys can update the program to include the options for dynamic tick size and tick value, by letting the users to plug in the date range, tick size, and tick value. And/Or you guys can host a central file, with all the tick size and tick value changes for all futures products obtained from CSI. The users can just download it and use the updated program to run the true simulation. I think it's a very important feature (some markets perform a lot differetnly with the new tick size and tick value), and you will own the bragging right to be the most realistic simulation platform on Earth. Just by changing the profit calculation logic a bit, it shouldn't be hard.

murray
Posts:435

06-02-2006 5:05 PM	Alert
We could do this but , the problem is I don't know if all the tick value changes are available. In addition if we really wanted to do it right then we should also use real margin values as they have changed over time. This leads to the question who's margin values.? and can we get histortical values. If someone out there know where to get this data, I will be happy to address this issue.

bernard (guest)

06-03-2006 12:33 PM	Alert
As far as I know, CSI has all the tick value for each product everyday, but they do not give the historical changes in one file (maybe I'm wrong on this matter). Since you are an important business partner with them or other data vendors, I think you can get them to compile you the historical tick value data. In a more tedious way, maybe you guys can call each exchange one by one and ask them for such information? About the margin issue, the system can just be evaluated with no margin? Then systems can still be compared side by side to see which one is better? I hope I understand your margin issue correctly...

bryanbonner
Posts:18

06-12-2007 5:36 PM	Alert
This is a very old topic, but with respect, I do not think having all the tick changes is useful. The reason is that what we are trying to do with a backtest is test how a system might behave if we experience price behaviour in the future similar to the price behaviour we have seen in the past. Introducing precise historical tick values introduces an element of precision that is at odds with the prospective nature of what we are trying to do. We are not going back in time to trade the actual markets as they existed, we are using historical prices to test them as they exist today. If you are going to do this, then you also need a lot more precise testing. For example: Given the actual volume on the historical day, could you have executed the size you are testing with at the price you are assuming in your test? Many markets we test with were extremely illiquid when they first began trading. Executing reasonable size would not have been possible at the time. Was the market locked limit so that you could not have executed at all? We don't check against historical limit moves. Did the market gap during the day? I took a significant hit in t-bonds once because of this. But you'll never see it on the end-of-day charts. Etc. Basically, there is a lot of historical precision that we are already ignoring so getting picky about tick sizes seems inconsistent to me.

You are not authorized to post a reply.

Forums > TradersStudio 2.x > General TradersStudio Forum > Tick Size, Tick Value, Currency Conversion, Data Management, Multiple Time Frame Analysis