I think there are many possible answers. One route is to use MySQL. It's free. Design a database to store your bought data. Create your queries, and run them on the database. You could also use SQL Server, Oracle, and numerous other databases. I've done a similar thing using just the language C#. C# includes a thing called Linq, which allows you to run SQL like queries on objects (read database) created within C#. I haven't used Python.

I bought End of Day data for LSE over 10 years. After a lot of work I've learned there are periods when the stock market is predictable, and periods where it's not predictable (at least with my noddy logic). I am only using EoD prices, and I have no trading experience. I suspect the key to better progress is access to more market dimensions (not just prices), and the experience to know how they interact.

Hi

My first recommendation is use mysql or R

Send me a message with an example I'll show you how to do.it if I can

I've done what you are asking about, and went from exploratory analysis to being a few weeks away from a fully automated system.

At this point you are really looking to maximize the return on your time investment. I find SQL database solutions very inflexible and extremely cumbersome for time series data analysis. A database is fine for storing data but not for analyzing it.

The real answer heavily depends on the amount of data you are dealing with. Since there are only a handful of futures in FI (compared to hundreds of equities), it comes down to the frequency of your data.

If you have daily data, Excel is an excellent option. You can do a lot with functions like SUMIFS, COUNTIFS, AVERAGEIFS, SUMPRODUCT and with pivot tables. However, once your spreadsheet grows to 200k rows x 1000 columns, you'll need something different.

Next, I would consider R/Matlab. They are easy to use and won't require a huge learning curve. Actually, since you're currently not looking to build a trading system, I would stop the search here. Theres's a heated debate on R vs Matlab - personally I find Matlab more intuitive and having the benefits of a commercial product.

While you'll get to where you want to with Python/Java/C#, it's a steeper learning curve and you'll definitely spend more time programming than analyzing data and coming up with new ideas. I would use these only after you've figured out your indicators and are ready to put them into a live application.

R vs Matlab

I do R but I agree to what has been written

For EOD data you can go with platforms like easy language or ninja trader

My advice would be get a quant/dev in to help you out. You should use your trading knowledge to guide their research and explore your ideas.

If I understood you right, you don't have already a concept or idea of an algorithmic trade system, but just want to start learning the basics. At this point it does not really matter with which language or system you will finally end up. Neither R nor Matlab or Excel is suited as a stand alone trade system, so you will need some serious test and trade platform anyway. Such a platform usually contains a script language and an introductional tutorial. With it you can set up a working system very quick, maybe in a week when you start as a complete beginner.

But this will not yet be a profitable system. After you got the basics, the next step will be inceed developing a system that really has an edge. This requires some serious math and coding knowledge. Trading experience is not enough, automated systems are very different to discretionary trading.

It sounds to me like you have a few ideas that you want to investigate to verify if your ideas are correct or not. To achieve that a simple database set up with some code may well achieve what you want. For time series data a SQL database isn't the best solution but if you have little development knowledge then going for something like Open Tsdb is going to be a huge learning curve.

You if you just want to keep it simple and try some ideas out then SQL and python will be fine. I'm happy to help out if you like.

There are platforms available that could get you going quicker such as ninja trader if you haven't explored that yet

Take a look here: https://www.quantstart.com/

Think carefully about how frequently you are planning on trading. If you are trading a couple times each day you may be fine with Excel and purchased minute bars. If you plan to make markets for edge you are going to need a professional programmer's help.

only real trade

Hi Dick.......I agree with Johann's comments. You need to have a concept of what your "solution"/"method" is trying to do. You need to be very clear about that. Your platform has to allow you to write your own formulae, and later add (or integrate) your own software (but usually this is later on).

From there, you start developing using basic tools, and as you develop, it is possible your tools will need to change (more flexibility and capabilities).

You will always need data, just make sure it is stored in a manner that can easily be changed/added to, and you are not paying through the nose for it. Usually, it's a long way from start to getting some results, and having to pay for data or extra data, all this time is just an overhead you can do without.

The attempt to be a trader, a quant and a quant dev at the same time may be a bit difficult to achieve.

For faster results I would recommend partnering with a quant dev that knows KDB and R and describing to him the patterns that you know to exist in the data and let him search for them.

It is possible to do everything by yourself, but will take longer, probably lot longer

@John. . . Or .. you take multi-charts .net (30 day trial free) ..and make your own strategy :)

The best thing to do is to purchase a simple back-testing package such as Ninja trader or MultiCharts and experiment with your ideas. Their use is pretty simple and the script language they use is very understandable. I think that MultiCharts is using "Easy Language" script that's really easy - as the name says - and Ninja is using C# like script. Good luck.

Where is the best, place to buy a months worth of fixed income data? (Eurodollars and Treasuries) Time and sales, depth of book...pretty granular stuff...