Case Studies

Case Study: BBG Data Storage

Project goal

Description

  • Data granularity – tick (level 1), 1/5/10/60 minutes bars, daily bars.
  • Fields sourced – last price, bid, ask, traded volume, cumulative volume, open
    interest, trading conditions (aka trade qualifiers), expiry date, last trade date,
    first notice date.
  • History goes 10-15 years back in a history (limited by history of data source).
  • Sourcing historical corporate actions.
  • Adjusting Futures data based on research roll schedule.
  • Building filters based on trading conditions (not trivial as they vary between
    exchanges, asset types and change over time) and price action.
  • Sourcing historical calendars.
  • Tracking and reacting on ticker renames (very common for Reuters).
  • Universe of unique symbols consists of around 3000 names.

Technologies

  • C#
  • Java
  • Python (scipy + numpy + matplot)
  • MS SQL, Mongo DB

Scope

  • 35 months, 18 man-years

Results

  • Secured and fail safe solution created.
  • Number of error and production incidents dramatically went down.
  • Cost of sourcing new names/market reduced considerably.
  • Quality of data improved.

Case Study: BBG Data Storage

Project goal

  • To develop technical solution for storing Bloomberg historical data, updating it
    from both historical and real time data.
  • Technically support it

Description

  • Data granularity–tick (level 1), 1/5/10/60 minutes bars, daily bars
  • Field sourced – last price, bid, ask, traded volume by default .
    Support of any data fields available via BBG API
  • History goes 10 + years back (limited by Bloomberg for non – daily bars)
  • Aggregated data (1/5/60 min bars and daily) created from tick data where available.
    Otherwise aggregated data requested from Bloomberg
  • Calendar support
  • Data for current day collected from real time data and represented in the same data formats (ticks, 1/5/60 min bars)

Technologies

  • C#, WPF, BBG API
  • MS SQL

Scope

  • 33 months, 9 man-years

Results

  • Reliable data source for automatic
    trading.
  • Client replaced existing data source with developed
    one.

Case Study: Risk report automation

Project goal

  • Create solution for generation Risk reports automatically.
  • Technically support it

Description

  • DB used as storage for any report related data (both historical and reference).
  • MS Excel add – in created for working with the data. If requested info is missing from DB – it will be requested from Bloomberg and stored in DB.
  • Users can create report templates using MS Excel with minimal set of macros for report API.
  • System supports running reports on schedule (daily, weekly or monthly) with sending results in email, storing it on network location.
  • Reports can be generated as excel files or PDF.
  • Reports can be requested to be generated on demand, using Web – site, which also can be used as admin console.

Technologies

  • C#, ASP.NET, MVC
  • MS SQL
  • VBA

Scope

  • 18 months, 3 man-years

Results

  • Flexible and reliable solution for report generation.
  • Client adopted solution and made it prod critical system.
  • Quality of data improved.