Understanding Bloomberg Data License (part 3) - Controlling cost

A common question for market data analysts and vendor managers is how to control and reduce Bloomberg Data License costs.

In the last decade we have seen a dramatic increase in data usage within the industry. The data is required for many uses such as internal and external (regulatory) reporting but also portfolio & risk management.

To satisfy the business' needs and data budgets creative thinking and deeper understanding are needed.

Internal data request policy

The first step is for companies to craft an internal policy detailing the current lay of the land and working towards an ideal future. This results in having clear internal agreements and insights into the securities and categories requested. An internal policy capturing the scope and requirements of data requests is needed.

Such a policy should also cover internal definitions of the expected lifetime of data, the definition of static and dynamic fields, and the acceptance rate of refresh for data and what would trigger these requests.

Once the policy is in place future developments and usage can be referenced against it to maintain overall control.

Three areas of control

A very simple way to start is to analyze the current and future needs against these 3 areas:

• The amount of securities requested
• The amount of categories requested
• The amount of additional daily access requests

Reduce securities requested

The first area of cost control is the amount of securities being requested. The main rule of requesting data is to only request what is needed. Usually this translates to securities in position and its underlying data (e.g. reference rates).

The reduction of the number of securities depends on the business needs and the internal policy. Some companies choose to request all data daily for quality checks. Other companies sit on the other extreme in which they request data only once and never again until an incident is raised. There is no ideal situation as each companies' risk acceptance level and budget is different.

The ideal compromise, however, lies somewhere in the middle where data is requested once and depending on various triggers such as e.g. corporate actions, rating changes or sudden price spikes could trigger a data request to validate the securities data quality.

Reduce categories

The reduction of whole categories is not always feasible however one of the areas that can yield the most savings. It is worthwhile to analyze the actual need for certain categories as these may only apply to specific asset classes and not the whole universe.

Splitting the requests per asset class is a way to reduce costs. The drawback of such a practice is the added technical complexity that it introduces.

Reduce multiple daily requests

Reducing access requests and thus costs can be a bit easier to achieve.

If the company has two separate systems or environments requiring the same data it will be difficult. The first step to solve this is to unify the needs of the consuming systems into a single request.

Using an Enterprise Data Management system (Markit EDM, AIM GAIN, Golden Source etc.) can help structure this for organizations with a bit more complex architecture.

Bringing it all together

As usual there is no one single answer to how to reduce cost. Each company needs to evaluate both long and short term requirements and have these aligned with the possibilities. Only then a clear picture on the how and where data savings can be achieved.