How to Avoid Invalid Conclusions from Data

The ability to harness the power of data is paramount in today's environment where asset managers are under severe cost pressures, operating in a highly competitive environment and navigating a tricky flight to passive investments.

Sensing opportunity, well-positioned financial services firms have made significant investments in internal and external data collection and analytics. With every tick in the markets, vast amounts of information are captured, logged and analyzed for business intelligence.

So why are so many firms struggling to get the insights and intelligence they need to make the transformative change they are looking for?

The reality for firms is that data and analytics-inspired conclusions are only as valuable as the integrity of the data itself and its timeliness.

STATIC & STALE VS. FRESH & ACTIVE

At Broadridge, we hear it all the time — managers have terabytes of data, and many spend hundreds of thousands of dollars on data services or on teams of data scientists to make sense of it all — but they are still drowning in data and lacking in "actionable" information? Before you can make sense of the data, you need to be sure you are analyzing the right data. Making business decisions based on static, stale and limited data is a waste of resources and can yield negative consequences.

For manufacturers and distributors, that means having up-to-date information across hundreds of different internal and external data sources. In the case of mutual fund wholesalers, having current information about which advisors use products like theirs and which of their products are superior to those currently used by advisors is invaluable. It is also helpful to know an adviser's contact preferences, whether they work alone or as part of a team, whether they have particular expertise managing retirement assets, or if they specialize in supporting high-net-worth individuals versus new investors.

Without that level of real-time detail and accuracy, your analysis will not yield meaningful intelligence. It may produce a list of long shot possibilities, but it will not provide near term opportunities to grow assets under management. Simply put, making decisions on bad data yields bad results and worthless outcomes.

Here are five questions to ask to ensure the right data is used:

1. How timely is it? Managers need to understand money in motion. Data that is old, stale and static is not worth much in a rapidly changing world. If data is primarily sourced from public filings, chances are it is out of date. Information from a year ago, six months ago, or even sourced quarterly is not going to reveal the insights that can be applied in the moment. Most consortium-based approaches publish results infrequently and on a delayed basis - as much as three months behind the current market environment. Always talk with data providers about how frequently their core data sources are updated. Data refreshed monthly will highlight current trends and spot near term opportunities.

2. What is the sample size? Millions of buy-sell transactions occur daily. Data should provide a significant representation of the market across a wide spectrum of distributors, channels, and asset classes to be truly representative of what is happening in the markets. Small sample sizes are easier and cheaper to collect, but they do not produce information that connects the dots to boost business. Samples are susceptible to sampling bias, leading to erroneous and contrary conclusions. Coalition-based approaches have an exclusion bias that can result in inaccurate profiling of target firms. If the sample set is not at least 85% of all packaged products (e.g., open-end and closed-end funds and ETFs), across all distribution channels, gaining a real view of market flows and a firm's progress is next to impossible.

3. How is it sourced? Data sourced from public filings is not always relevant and is often out of date. The same can be said for information that is contributed voluntarily by a small consortium of member firms. These types of indirect data sources raise questions about transparency and accuracy, and they should be reviewed carefully to determine if they are representative of underlying events and behaviors of interest.

4. Can it be integrated? Having valuable data spread across many systems makes it challenging to discover meaningful insights about your business. Combining different data sets is not easy, but it is critically important work. A best practice is having timely intelligence pulled directly into one system for a seamless, targeted and holistic view of the data.

5. Is it actionable?Capturing the right data is the start to a strategic framework that ultimately produces a roadmap and an outline for next steps. Quality data and analysis delivers detailed outcomes that tell you who to call on, informs your message, details differentiators, and compares your product set to peers. Analyzing the data and information with one goal — making it actionable — is the key to unlocking unique intelligence to drive your business.

Data integrity matters. Having the right data can lead to a treasure trove of actionable intelligence that streamlines operations, minimizes regulatory risk and supports business growth. In an ever-changing sea of prospects, using robust, timely and active data sets to support predictive and prescriptive analytics is the key to success.

Dan Cwenar is president of Broadridge Data & Analytics.

For reprint and licensing requests for this article, click here.
Law and regulation Money Management Executive
MORE FROM FINANCIAL PLANNING