Brian Loughman:  If you think back 10 or 15 years ago, a lot of the activities in the business were being recorded electronically, but they were often archived in separate systems. The warehouses would have an inventory log; the sales team would have a sales register and accounts receivable. A company might have had all the building blocks for financial systems, but the data stayed separate. And a lot of the incidental transactions were recorded on paper. Today, practically everything is recorded digitally, and it can often be collected easily into a single database. Then it can be analyzed to develop a perspective that is relevant to various stakeholders in the business.

For example, in the past, companies used data analytics around their customer base. They would log every customer service call into a database and then develop insight around this data. It was a fractured process, focused on a very specific function. Today you're seeing organizations analyze broad stores of information about all kinds of routine transactions. They can do this because they're aggregating their 'big data' and because the systems holding this data are more nimble. At the same time, newer technologies have been developed to search and analyze vast amounts of data very quickly. These advancements have led companies to say, 'We have all this data; maybe we should start figuring out new ways to see what it's telling us about our business.'

T&R:  Is this something that you're seeing frequently in corporate treasury, finance, and risk management functions?

BL:  Because treasury and finance are so closely tied to the business cycle, we usually see those functions leveraging data analytics activities that are already happening in the business units. For example, the COO might be very focused on the company's leading performance indicators. Treasury and finance might take those metrics and then extrapolate information such as what's coming up in the accounts receivable cycle.

Scott Keipper:  There's significant potential for using data analytics to benefit treasury and finance—for example, by improving reporting reconciliation and liquidity management. But for the most part, finance teams are in the early stages of using big-data solutions. They're primarily focused on IT cost savings.

We work a lot with treasury and finance organizations in financial services companies. For them, in the current environment, big-data initiatives are driven primarily by increased regulatory requirements or specific cost-savings benefits. However, some are starting to look at more significant opportunities for convergence between risk, treasury, and finance, which will allow for more robust analytics and performance management—for example, in activity-based costing and transfer pricing.

T&R:  When they implement big-data solutions for other purposes, are these treasury and finance functions also gaining new insights into key performance metrics?

SK:  Yes, absolutely. Traditionally, finance functions have had access to subledger data, and they've been able to do a certain amount of analysis on that data to support financial disclosure. They may have also supplemented that analysis with information from the lines of business or risk management, but the data was disparate, challenging to obtain, and even more difficult to reconcile. As they have access to more robust solutions, they can dive deeper into the data.

In a bank, for example, if finance wants to understand what's driving variances in certain capital reserves, being able to actually drill down into the data, all the way back to desk-level or transaction information, enables them to really understand what's driving the changes to the business. Traditionally, that type of analysis would have taken weeks, but it now has the potential to become a standard part of the financial close or financial reporting process.

Banks are also able to use these solutions to uncover and free up 'trapped' capital or liquidity. When banks don't have data readily available, they have to take conservative assumptions around capital and liquidity reserves. But as a result of being able to drill down much deeper into the organization, they can understand more about each transaction or position, and they can classify it more efficiently. They might even use this data to make decisions about changes to their business and more closely align funding costs with the economics of the business unit.

BL:  For treasury functions outside of the financial services sector, financing working capital is a key concern. They want to make sure they're not getting stretched out on receivables, and they need to manage payables. Now companies are starting to have the opportunity to collect all this data in one place and mine it in sophisticated ways. They also have the option to use visualization tools to see correlations that they may not have previously known existed.

T&R:  Are companies also using analytics to dive into data from external sources?

SK:  The ability to look at unstructured data is something we're seeing as an emerging trend. Companies are looking at external data that's very unstructured in nature—think of email, Facebook, phone conversations, things like that—and integrating it into traditional structured transactional data. Then they're running analytics against all that data to help drive an understanding of performance metrics and risk metrics about the organization. This trend is starting to show up in marketing, customer relationship management, and even credit underwriting.

BL:  For companies in some industries, a lot of their customers or competitors might be using blogs and other digital media to talk about their businesses, and some companies are aggregating that information. For example, an organization might be looking at whether its customers are talking about expecting to increase in size. An individual instance might not mean much, but if a company sees that kind of commentary across a particular segment of the industry, it might have an interesting indicator of where its customers are headed and how it can best position itself to serve them.

SK:  That being said, most finance and treasury functions have not yet explored the possibilities of unstructured data analysis.

T&R:  What are some specific performance or risk metrics that companies are pinning down by delving into unstructured data?

BL:  Well, I'm sure you've heard a lot about all the various compliance issues financial services companies have been dealing with, such as allegations of improper trading, insider trading, and interest rate fixing. Big data can actually be used to ferret out these kinds of activities. It's not uncommon for all the traders in a market—trading a particular security or currency—to instant message each other all the time. All of that is saved, and some organizations are using software systems to survey it and identify periods of time or certain traders that are causes for concern.

From a treasury standpoint, to the extent that there's any particular behavior that you're seeking to monitor, the advent of big data and the powerful analytics tools that are now available mean you can do so for everybody in the company at the same time. The reality is that computer-driven forensic data analytics helps you generate more intelligence about your business operations because the computer is better than we are at making connections very quickly.

Everybody has certain red flags they want to keep an eye on, and big-data tools enable you to do that in fairly close to real time. Most companies are already capturing most of this data; the question is whether they're effectively collating and analyzing it. Even if you have a lot of data, a thoughtful application of analytics can help you determine which data you're not interested in, based on your particular objectives, and you can focus in on the most critical data. Then you can create a lot of regular reports that give you close to real-time—so, daily or even hourly—reports on your key indicators. You can achieve a much better sense of your business operations.

T&R:  What is the future of big-data analytics for treasury, finance, and risk management?

SK:  Some organizations are moving toward a convergence in which the risk, treasury, and finance functions will use a common set of data and a common reporting framework. This solves what has historically been a big challenge for some companies: reconciliation between finance and risk data. That said, with big-data solutions, there is also potentially an opportunity to support reconciliation and forensic testing using other types of analyses—approximating how things reconcile or what controls exist—rather than re-architecting the whole infrastructure.

I also think there's a lot of potential for aggregation of data for analysis across companies and across industries. Regulators are in a unique position because they're capturing such a massive amount of data across the various regulatory initiatives. There's an opportunity to look at how they pull the data together as an organization to start to identify industry trends. Historically it's been disjointed because the regulators are not necessarily focused on the technology. But I think there's an opportunity there.

BL:  I think the analytics capabilities will continue to expand, as well. This sounds far-fetched, but our 'know your trader' forensic analytics system can actually determine the emotions of people from their written communications. We have a team of Ph.D. linguists who helped build these taxonomies. And we don't have to read anyone's emails; the machine does that and alerts us when it encounters a red flag.

Many folks in treasury and finance have financial tests, looking for activity that goes over certain preset limits. And of course that's an important use of a big-data solution. But you could also have tests that are more nuanced and more qualitative. You could survey the emotional state of employees in key positions, and if you found extremes on either end, those might be people you would want to take a closer look at.