In the spirit of today’s progressive and advanced business place of the asset management industry, extensive importance has been placed on all things data. In fact, data becomes the heart of the industry’s development and prosperity.

What exactly is Data Governance?

Data governance is the rules, processes, and accountability around the data. It aims to handle:

• How is the data used by the asset manager’s different teams?
• How to harmonize data providers and sources?
• How to manage access to the data?
• Who has ownership of the data?
• Who is responsible for data accuracy?

What is the difference between Data Governance and Data Management?

When defining data governance, it is interesting to distinguish the difference between data governance and data management.

Data governance outlines the overall structure; it is the organization of structures, policies, rules, metrics, and data owners for the end-to-end cycle of data including, collection, storage, and usage. Whereas data management is the actual implementation of these rules and processes.

What are the Challenges and Potential Solutions of Data Governance?

Since 2008 we have been assessing those classes of problems and trying to find sustainable and accurate solutions.

Our past experiences have led us to conclude that although the challenges of data governance may be many and varied, there is a solution for each challenge. In the following section of the article, we will present four principal challenges which may arise when implementing data governance as well as the potential solutions to overcome them.

1- Limited Resources

Limited resources occur because data is viewed as a product rather than a service. Thus, asset managers are building in-house solutions nowadays, allowing us to view data as a product.

As a result, an asset manager has to evaluate the potential added value this would add to oneself, which includes building strategies, and forecasting profit and performance. When observing the challenges of limited resources, this branches out into two different categories:

The first being the struggle of building a suitable and structured IT team. Issues such as hiring the best profiles for the roles and ensuring that the budget accommodates the new workforce. Along with this issue is the usual, high rate of staff turnover, and retention of an adequate IT team is a challenge. Often leading in having to restart the recruitment process, which translates itself into a larger issue. When referring to a lack of skilled human workforce, it is referring to those who are not knowledgeable or equipped enough to support an ongoing data governance program. Evidently, without these industry-specific workers, challenges often occur which subsequently affect the data governance program.

Second, asset managers have financial constraints and thus, may not have sufficient funds to allocate to data governance. Their allocations for resources do not fit within the budget. The cost of IT resources comes as a financial constraint this encompasses hardware as well as software, data center hosting, licensing, and data storage. Where for data storage the pricing can vary increasingly, for example, 700TB costs 20K€/ month (AWS).

Solutions to Limited Resources

Solutions to overcome this problem include establishing strategic visions for how data governance should be managed. Additionally, when revising the budget for the year, it's important to ensure to allocate sufficient funds for each year or quarter. It's important to refocus scarce resources on higher-value activities to prioritize effectively.

Finally, a reasonable solution to combating this problem would be to analyze project data and base one’s budget and schedule accordingly. Nowadays, data has to be a service, rather than a product. As an on-demand service, asset managers have unlimited resources in terms of support and maintenance, allowing asset managers to “pay as you go”!

2- Siloed Data

An issue that arises with siloed data is most of the time, the databases are duplicated, which is seen to be non-consistent making the challenges with siloed data many and varied. For starters, data silos slow down an asset management firm since they significantly limit communication and collaboration channels between the different teams (from the back to the front office). Another issue which we noticed was the fact that they also reduce efficiency and take up a large amount of storage space, we noticed that 100% of the data is collected, however, only 1% of it is ever utilized. As such, siloed data decreases the quality and credibility of data.

Many struggles with data; different teams within asset management firms collect their data from sources but do not share it with other departments, we are aware that the main reason for this is that datasets are often locked and only available for select teams. Correspondingly, different teams abide by their preferred methodologies and priorities in their operations. This fragmented process contributes to inconsistencies within the data and eventually will lead to complications within the asset management firms (such as mapping the financial instruments codes).

Solution to Siloed Data

As asset managers assemble increasingly more data, it's vital to ensure that it does not become too siloed. It is also important to ensure that different teams collaborate and use data from different channels, this will result in dedicated data governance frameworks. These frameworks will open data sets that will break down data silos. Thus, enabling asset managers to deal with accurate data sets. This in turn will enable consumers of data to focus on their main tasks and abilities without creating siloed data. In a previous article, we mentioned that data scientists are wasting 70% of their precious time collecting and cleaning the data.

3- Lack of Leadership

An asset manager's culture can have an enormous impact on the future success rate of a company's investment products. If leaders view the subject of IT as a non-essential product in their industry and don’t view it as a priority this will lead to a depletion in terms of structure and coordination. Subsequently, we noticed that if the asset managers do not have long-term plans in place, it usually has detrimental effects, resulting in a misunderstanding of challenges from the data management systems, and viewing the IT systems as a cost rather than an important investment.

Furthermore, if a leader does not communicate his visions to ensure that the methods of achieving a goal are established with his fellow asset managers this also has a negative impact. This often poses to be a challenge because the IT teams and asset managers do not talk the same language as one another. This makes it increasingly difficult to build down the barriers and allow for communication within divisions.

Solution to Lack of Leadership

A fundamental solution that needs to be followed is that the business team is required to drive the data governance project, simply since data is deemed as an integral part of an asset management firm, as asset managers make investment decisions based solely on data.

whilst the development of a data governance program is taking place, its effects should be consistently reviewed to ensure policy delivery, structure, and execution are made comprehensible. Implementing these standards can permit the data governance division to employ best practices, whilst remaining agile and confronted with foreseeable changes. Correspondingly, ensuring that data governance officers are aware of how to break down concerns or ideas about data governance through data modeling and presentations is vital in combating the issue of insufficient leadership levels. As such, replacing individual skillsets by utilizing collaborative techniques.

4- Poor Data Quality

Collecting, cleaning, and sharing data is a common issue being faced by the asset management industry leading to the incessant search for high data quality. A large amount of time is wasted assembling data, that has no impact on long-term aims or objectives. Besides, the ever-growing number of sources combined with the increasing amount of instructed data makes it challenging to test whether the data is from an accurate source, and poses the infamous question, “Is my data accurate?”. This can be a result of wrongly collecting data from correct sources coupled with not cross-validating one’s data.

The main reason for poor data quality could be a result of human errors. When manually entering data, this often leads to errors as it is an error-prone task, this includes missing details, typos, or entering data in the wrong fields. Moreover, another compelling reason would be poor data integrations, the data quality varies drastically across one system which one is obtaining data sources.

When data is not up to date this also contributes to poor data quality and makes it unusable and unreliable. Similarly, if data is not cross validated this causes issues with the quality.

Solution to Poor Data Quality

To conquer the challenge of poor data quality, data standardization should take place. This would eradicate the need to do manual manipulation to normalize data again. Moreover, it would increase the quality of the data. Cross-validation should inevitably be of regular occurrence to ensure a high quality of data.

Conclusion

The themes highlighted in the above article, articulate, and support the vitality and importance of data governance being employed within asset management firms for an array of different reasons. From the article, it is clear the benefits outweigh the possible and potential challenges. Thus, reinforcing the need to un-fragmented the view on data governance.