Capital Markets: Managing Intelligence Optimization vs. Big, Diverse Data

The dialogue on big, diverse data has been changing in capital markets during the last few months, and for the better. First, the name is changing and not just due to over usage and hype, but because firms are experimenting and definitions and applications are shifting away from the old definition - if there ever really was one. The V's (velocity, variety, volume) are just some of the characteristics identified years ago about data, so they really fail to match what many firms are currently experiencing.

At the moment firms are seeing all types of data with an advanced analytics platform and the application of findings can drive value and hold the promise to what folks have been clamoring about. It's more a process to harvest and exploit untapped knowledge, than one specific rigid definition -- you might even call it something like "business insights or intelligence optimization -- BIO." The process is building a greater level of information for a fact-based approach to decision making and organizing the firm's capabilities and people with a platform and tools that will add value to the business and customers.

Capital markets firms that are leading adoption of a big, diverse strategy realize having a stated plan and governance in place is a best practice and strong place to start. These companies start in areas with the greatest pain or opportunities to mine first. For capital markets, trading and investment strategies are always top of the list. Risk management and customer insights are also critical areas of exploration, where the application of resources can help improve trading and investment strategies.

Remember, correlation is not causality, so traditional variables used in the discovery process may or may not explain or highlight new insights for trading and investment when put together with new information sets. So investment managers and trading firms are growing an appreciation for how the new process reveals alpha or highlights bias in models and algorithms. The process of quickly spotting these small insights, moving on and iterating is another part of the workflow that leading firms employ. Not "boiling the ocean", so to speak and staying focused is a critical element of the deployment.

An important part of this process is focused on people, the talent pipeline and skill shortage. The competition is fierce for experienced "BIO" talent. Capital markets firms are evaluating how to staff or build teams to manage their big analytics projects. They are working with colleges to promote data science, analytics and math to create the pipeline coming in. Sponsoring workshops and activities with schools to highlight their firms as interesting versus the startup and large tech companies that often attract young talent. Once you get the word out and let graduates know that you have big problems to solve, it makes a creative value proposition for your firm as to not lose qualified candidates to competing industries.

Another approach is to scan internally and find people who already exhibit many of the attributes or skills that are associated with "data scientists." The precise criteria for data scientists is still developing, but is usually based on skills or traits such as statistics, data mining, being analytical, problem-solving (think detective deductive/inductive reasoning), inquisitive (hypothesis formation), collaborative and communicative. It may seem like a tall order, but perhaps it's a team rather than an individual.

With your people in place, you may need to centralize, perhaps even separate, a big, diverse data analytics group to foster ideas and prevent infection from a "poisonous corporate" culture that may kill off projects This team may be organized with the sole objective of uncovering answers to some of the business's tough questions and stumbling upon a finding, no matter the size of the impact, that can move the needle of earnings growth or customer improvement.

The last point to raise is about infrastructure. There are many ways to approach this, but in all of them how the architecture is determined needs to match the above two elements first, rather than the other way around. The reality is that there is no one singular "box" or app that will manage your big, diverse data or BIO project completely. Nor is there a silver bullet that will attract and engage the people across your firm to transform the organization into a data driven culture or create an environment where questions can be generated. Rather it's the thoughtful combination of existing and new solutions that maps to where your firm is on the information maturity curve, and where it wants to go.

Leading capital markets firms now making progress with "BIO" projects see data as a strategic asset and realize the benefits, regardless of how intangible they may initially appear, will ultimately deliver value. If data is the "life blood or DNA" of the organization as we've heard so often, then it would stand to reason the findings and patterns are unique as well. These capital markets firms understand the collaborative gain from democratizing data.

Democratizing data does can come at a cost and it can be confusing to know what the right architecture is as well as paying for it. Conversations about budgeting for new solutions or fashioning older ones to bend to what's needed takes investment. Chief data officers are making efforts to educate the executive suite, leveraging leading big, diverse data vendor partners and using them to present the latest developments in the industry, delivering tangible value every 3-6 months and having projects fund other projects. The requirement is to keep the corporate mind and wallet open.