Thursday, 12 April 2012

FIX Cube - Getting started.

FIX Cube has two input fields: the usual large textarea to enter the log and a second textfield.
This second textfield allows you to group the data by FIX tags.

Its rules are simple:
  1. Enter one or more comma separated FIX tags to define the data-slicing.
  2. Use square brackets to group together FIX tags into oneslice
  3. Use the word 'fix' to drill-down to the very FIX messages.

A few examples:
  • 1,207,fix
    Groups the items by account, then by security exchange before reaching the FIX messages 
  • 167, [55,200]
    Groups the items first by product type (options or futures) followed by the combined value of the ticker and maturity month year (aka contract)

  • 32
    Groups the items by volume (last shares). Sort them and you will know your largest transactions.

We hope you like it.

Monday, 28 November 2011

Basics of the VX Time Slider

When looking at your FIX data is quite common to define a time range smaller than the whole set.
For instance, if you have data for a whole day of activities you might want to measure your trends up until midday or up until 16:00pm.
You achieve this by clicking on the time range bar and dragging the mouse to define your interval.

Your time slider will be different depending on how your data stretches.The picture below shows some common ranges:

It's important to note that, once you apply your time filter, every metric will be recomputed on the newly defined subset. In other words, you are not just filtering out some data you are actually seeing your data within a new time context.

Friday, 25 November 2011

About the VX Demo.

The VX Demo is designed to show you how to solve a common business problem: The discovery of your best and worst traders in a given time.

By discovery we mean: to find out who, when and how.

Order your table by P&L and you will find your worst and best traders. Our best trader is account D865 with a profit of 845,718 $

Now, if you explode p&l into a chart you will see that this account was doing well in the morning (a peak around 9:00 am). The early afternoon is pretty bad (around 14:00 am) is really in a bad state. But then,there is a recovery and by 19:00 pm this account has a profit over 500k.

Our next level drill down shows us the contracts. Account D865 actually traded only 8 contracts all day. He did actually do badly with 4 of them, particularly with CME Japanese Yen. Yet, trading the CME Australian Dollar was the reason he did so well.

Final Words 
Before I conclude, let me remind you that your view can be easily configured to build different pivot tables. For instance, what if you wanted to see the performance of a certain trader by security type ? Just put the security type on top of your view and you will be able to slice your data by options and futures.

In this last picture you can see that account L260 was quite successful in tradinx Eurex Eurostoxx 50 opt.


Wednesday, 23 November 2011

What's happening to our website ?

In the last weeks, you might have spotted a few little changes on our website. In the coming weeks, you might notice a few more. You might wonder what's happening ? Simple. We are trying to make sure that our visitors get to know another one of our tools: VX.

Tuesday, 22 November 2011

Monday, 12 July 2010

Who in the world uses FIX ?

July 2010. I decided to get a snapshot of 3 months (90 days) of data from Google Analytics. I filtered out the occasional visitors and took into account only those who had executed at least one FIX message in either FIX Analyzer or FIX Log Analyzer.

In that period, the 'analyze' buttons on FIX Analyzer and FIX Log Analyzer had been pressed 23,129 times from 6,815 unique visitors in 75 countries.

Yes, I thought I have enough data to give a satisfactory answer about where FIX is getting used today.

Of course, I am assuming that apps which are using FIX to receive/send/execute messages are constantly debugged, diagnosed, studied and that many people turn to the web (specifically our ValidFIX site) to do more of this work...

Anyway, here is the chart and table I produced.

United States214631.49%
United Kingdom120717.71%
Hong Kong1992.92%

With Others* including the following 54 countries: 
  • Andorra
  • Austria
  • Belarus
  • Bolivia
  • Bulgaria
  • Chile
  • China
  • Colombia
  • Croatia
  • Czech Republic
  • Denmark
  • Ecuador
  • Egypt
  • Finland
  • Georgia
  • Greece
  • Hungary
  • Iceland
  • Indonesia
  • Ireland
  • Isle of Man
  • Jersey
  • Jordan
  • Kenya
  • Kuwait
  • Lebanon
  • Macedonia [FYROM]
  • Malaysia
  • Malta
  • Mexico
  • Morocco
  • Netherlands
  • New Zealand
  • Niger
  • Norway
  • Pakistan
  • Poland
  • Portugal
  • Romania
  • Saudi Arabia
  • Serbia
  • Slovakia
  • Slovenia
  • South Africa
  • South Korea
  • Spain
  • Sri Lanka
  • Taiwan
  • Thailand
  • Tunisia
  • Turkey
  • United Arab Emirates
  • Uruguay
  • Vietnam
Hope this helps.

Tuesday, 25 May 2010

Why ValidFIX likes Google (a tale of a pleasant, unexpected discovery.)

There was an interesting post on the official google web master blog in april: Using site speed in web search ranking

In short: speed matters, faster websites get a better ranking in google.

I read the post with interest and then moved on.


In developing our site we did not spend much effort in SEO – we just followed google standard rules – but we did obsess with performance.

People who work with FIX Messages deserve better tools to analyze their FIX messages and since we were going to deliver them through the web we did not want to sacrifice any aspect related to their usability.

To an extent, our obsession dictated our minimal GUI interface and made us work hard to follow strictly yahoo performance rules.

And then the pleasant discovery.

A week or two after google’s post we started seeing our ranking moving up. If a person was googling for ‘fix analyzer ‘ or ‘fix log analyzer’ we were coming as first or second result.

Beating onixs and firstfuturessoftware.

This was great news for us as more people were (and are) likely to find us but it made me wonder what was happening since I could not see new links from reputable websites (such as quickfixengine) linking to us.

Was the google engine just taking its time to update the ranking ?

Maybe, but I believe it’s more likely that we have been rewarded for designing and delivering a fast website.

Thanks for reading this.