Analytics Don't Solve Problems with Michael Canale

007 - GAT Podcast: Force Multiplier - Analytics Don't Solve Problems

MJ 00:00 Hi, everyone. This is Michael Jelen from the Global Applied Technology podcast. The GAT Team, as we call ourselves, is a globally distributed team of software engineers, data scientists, graphic designers, and industry experts who serve clients through our products built atop the BRG DRIVE(TM) analytics platform. We're helping some of the world's largest and most innovative clients and governments transform raw data into actionable insights, drive efficiency through automation, and empower collaboration to improve business decisions. You can learn more about us, our products, and our team on our website, brggat.com. And if you have any questions or comments, please email us at G-A-T at thinkbrg.com.

Today, I'll be speaking with Michael Canale, the head of BRG's Financial Services Technology team and a founder of BRG's Financial Institution Advisory practice. Mike has over twenty years of experience across consumer finance operations, servicing, and technology. At BRG, he focuses on our financial services technology strategy, products, and strategic relationships. Mike made some pretty interesting points, including the fact that analytics don't solve problems. They just find them. He discusses the crossroads we're at regarding financial regulation, including the impact of crypto and decentralized finance, and the need to automate entire processes through a digital workforce of machines. Please enjoy this conversation with Mike Canale.

Good morning, Mike. How's it going?

MC 01:17 Good. How are you doing?

MJ 01:19 Great, very excited to talk to you today about the role of technology in the financial services industry.

MC 01:24 Same here.

MJ 01:25 And I was hoping you could start off with a little introduction and background on yourself.

MC 01:31 I started my career in consulting in 2009. Coming out of the financial crisis, I got into the bankruptcy and restructuring world, where I got my first taste of consumer finance. My first engagement was winding down a mortgage servicing company in Ocala, Florida. I did that for about five years and then moved over to the operations risk and compliance and technology side of consulting in 2014. Since then, I've built several technology platforms to deliver analytics and customized automation solutions to clients. And I've also worked on several regulatory enforcement matters, where I led teams that performed different assessments to evaluate compliance management systems and then also remediate issues that we found. In 2019, I came over to BRG with three other partners and started the Financial Institution Advisory practice. And today, we're up to over twenty [people]. And we've also partnered with the Global Applied Technology team within BRG to continue building tech and bring it to our clients.

MJ 02:44 Excellent. And as a member of the Global Applied Technology team, I've been very passionate about technology for quite a while. Can you tell me a little bit about how you found technology and how you started to integrate it with financial services in your background?

MC 02:58 When I started in financial services in 2014, one of the first engagement jobs that I had was to actually perform manual mail tracking. I had to take a sheet of about 300 different packages and go on the US Postal Service website and type in the package ID number, track where it was located, update the spreadsheet, and send it back. That that was really inefficient and boring. I immediately started thinking of ways to automate it or reduce the amount of work to finish the job. I met someone in my office that actually had some coding skills. He took a look at what I was doing and came up with a quick solution, and I went from doing 300 a day to doing over a thousand. It was taking me less than thirty minutes to get it done. And that kind of sparked my interest in solving the efficiency issues using technology.

MJ 04:03 Yeah, that is a similar story to the way that I also found the power and magnitude of using code or analytics in order to build faster and more automated solutions. When you look at the way that things have changed over the past ten or fifteen years, I have certainly seen an increase in technology in financial services. But how do you think we got to where we are today? And what room for growth do we still have?

MC 04:28 I think we got to where we are today by—lots of companies have internal technology groups. They're using vendors. Some are starting to form partnerships with fintechs. But it's really been done without a holistic view of the entire organization. So, there's still lots of inefficient processes that result in things like high cost, lower productivity, especially in financial services.

Automation is really focused more on the customer experience and less on back-office operations. Where things like analytics have been deployed, they often don't solve problems. They just find them. They're essentially creating more work. So right now, a script may find a potential issue. But there's still armies of people revealing and validating whether there's actually a problem. I think financial institutions still spend a fortune on compliance. And they generally focus on the highest-risk issues that are out there and do things like sample testing versus full population testing. You're starting to see more startups focus on the regtech side of things, where they're deploying analytics that will do what's called continuous monitoring. And that's full population testing. But again, like the point I just brought up, analytics still just find those problems. They don't necessarily solve them. So, there's lots of opportunities to still reduce the amount of work or even reduce the amount of issues that can be found. And one of the ways I think we can do that is by moving where we're focusing that automation.

Analytics are traditionally focused on after-the-fact mining data to find something as a detective control, where automation really should be focusing on preventing those issues from even happening. And I think that's where you'll start to see some advanced types of technologies deployed.

MJ 06:42 Have you seen any very good examples of that or places that are doing it right in identifying and automatically reconciling or potentially predicting the potential risks involved in any sort of transaction set?

MC 06:57 I've seen it done a little bit. I have not seen this mass adoption of using things like optical character recognition, artificial intelligence, machine learning, or a combination of the above to really streamline the originations process and reduce the number of errors. Some companies have started to do this. There's a mortgage fintech startup. They've really automated a lot of the originations process. But they still have a decent number of people checking documents after they've been filled out and uploaded. So, I think there's still quite a bit of room for improvement.

MJ 07:53 And since the financial services industry is so highly regulated—and, often, a good chunk of this additional work is to appease and ensure that reporting to regulators is effective and thorough—how do the regulators look at this? Do they have most of these processes automated? Or is it still largely a manual effort on that side?

MC 08:15 They don't. So, the Consumer Financial Protection Bureau has started to dip their toe into tech. They do have an innovation group where different vendors can apply to get what's called a no-action letter approval from the CFPB, where if they actually get that, companies that use their tools would actually get almost like a waiver. The thumbs-up from the CFPB would be like, "Hey, we're not going to dig here too much." But in terms of their actual oversight and the way that they evaluate companies, they're still doing things very manually. They'll have a doc request list. They'll do sampling of transactions. They'll have people do testing and verification. And they don't always get it right. That's one of the biggest complaints about the CFPB: "Hey, they don't understand our business. They came in here. They read something. They think it's wrong. And there's really no issue there." I think they've been a little hesitant in the past to embrace tech because they don't truly understand it. But with the rise of fintechs, you're starting to see the CFPB also migrate the way it thinks about tech.

MJ 09:33 Okay. And so, while there may be a migration as a result of fintechs, it seems to me like often, during crises or large events, that is really the impetus for regulation to change and evolve at a quicker pace. Have you seen, or do you see, any nudges occurring in the political landscape or different areas that you think could shift that or change, in a nice way, for regulators to start to adopt more technology?

MC 10:02 So there has been some focus on tech. But it's more from enforcement and less from how regulators plan on using technology. I mean, I think it's a natural—their perspective is naturally going to evolve. As they start to get more involved in enforcing on fintechs and other technology firms, they're going to have to get smart about what this technology actually is, what it does, how it works. And that's going to force them to embrace hiring people with more technology backgrounds, potentially partnering with outside firms to create tools that will make their jobs easier and more efficient, which is not typically something that you would think about a government agency, being focused on efficiency. But they really can step up their surveillance in a significant way if they embrace tech.

MJ 11:05 Gotcha. And do you think that some of the different political changes that have happened recently may drive slightly more scrutiny than a previous administration? I know these things seem to go in waves, where it seems like one administration may take their foot off the gas and then the next may put it back on. Where do you think we are right now in that cycle? And where do you see the next couple of years going?

MC 11:28 Yeah. So, history-wise, with the CFPB, they started coming out of the financial crisis. Dodd-Frank was passed. As part of that, the CFPB was founded. And that really changed the level of scrutiny on banks, mortgage lenders, and credit rating agencies. In the period immediately following that, there was an intense focus on implementing and enhancing compliance management systems. So both banks and mortgage servicers were hit particularly hard. These were areas that they had traditionally ignored, or they just were not prepared for the level of scrutiny that resulted from that legislation. When Trump was elected, there was a significant drop-off in enforcement. So up to 2015, there was a really significant amount of action from the CFPB. And that covered things like credit reporting, fair lending, compliance management system reviews, you name it. After Trump came in, that enforcement activity dropped off by as much as 80 percent from its peak in 2015. Compliance also dropped significantly.

I think what you're going to see now, with Biden and the Democrats having control, is more of a return to the old ways. Enforcement will definitely ramp up with the pandemic, the passage of the CARES Act. There's plenty of things for them to start looking into from a servicing perspective and potential borrower harm. They've really spent a lot of time looking at analytics around call centers, dropped calls. They're looking at how forbearance was treated during the pandemic. But my point is that there's a lot that's happened in the past two years, along with the new administration, that gives the CFPB and other regulatory entities plenty of ammunition.

MJ 13:47 Makes sense. And as we've seen recently in the news, quite a bit of regulation is being discussed in the Senate about blockchain and these new technologies. I can't help but think that the way that fintech is sort of playing a role in the financial system at the moment is kind of. It reminds me of mullet in some ways, where you've got in the front, business in the front, party in the back. It appears on its face to be a very reputable, nice, clean-cut financial system and starting to dip into technology in the back. But it really, truly isn't revolutionary in a lot of ways, the same way that decentralized finance on blockchain and entirely autonomous exchanges and autonomous places, where you can deposit assets and make loans without going through credit checks or other kinds of reporting, or really kind of shaking up the industry quite a lot. I guess I'm curious. What do you think regulators are thinking about blockchain and things that are coming out that are truly very, very different and may use different fundamental technology and financial rails to drive innovation compared to the old system?

MC 14:53 I think they're worried about it. I've been thinking about this for a couple of years. And when we first got to BRG, we started thinking: what is crypto and blockchain going to do to the regulatory landscape? I think you're starting to see, from a financial crime angle, they're starting to be a lot of discussion around what could potentially be done to monitor transactions and crypto, prevent bad actors from doing transactions and laundering money. I think that's one of the benefits of crypto is the anonymity of it. But at the same time, it allows most of the controls of the traditional financial system to be circumvented. From a regulatory perspective, that's not a good thing. I think finding a balance, though, like speeding up the speed at which transactions can happen, giving that control and power more to the users rather than an intermediary—I think those are all really positive things. But there has to be a way to also prevent fraud. What the right way is, I think, that's what the big debate is. I think it's going to take a while to figure that out.

MJ 16:18 And, in the meantime, it seems like every financial firm, whether traditional, fintech, or DeFi, are really sort of guessing how they are going to need to respond to regulators. And that appears to be driving a huge amount of burden to these firms. So, when you think about that regulatory burden that's coming from potentially unclear or unknown future regulation, and you couple that with, generally speaking, kind of an old-school industry that does quite a lot of things manually anyway, it seems like there's just an enormous amount of work that's piling up that isn't really being solved in the most efficient and integrated way.

What sort of solutions would you propose in this industry? Or what are ways that you think, by and large, your average, everyday financial services firm or perhaps a new fintech could start to get ahead of the game and do things a little bit easier for themselves while also appeasing regulators?

MC 17:15 You brought up a really good point earlier that fintechs, while they're viewed as disruptive, they're really just different pieces of the traditional financial system. You could have a company that's starting up a bank charter. And they're going to offer traditional lending services. And they have a cool website to do that through. And maybe they have some advanced technology on the back end. But they're essentially offering the same thing that traditional banks offer, just packaged up in a neater, slicker way.

As these disruptive fintechs mature, they're starting to get that regulatory pressure, right? So, the CFPB has already started to do some oversight and enforcement on fintechs. And when they go to IPO, they have to have certain things, like a compliance management system. They have to have that financial crime, BSA/AML monitoring in place. And a lot of them still do business with traditional banks. So, they have either partner banks or sponsor banks that have a set of requirements that they need to follow. And those requirements are really grounded in the way the traditional financial system works.

So up to now, the solutions that have been put in place are kind of more of the same, right? Fintechs are hiring people from banks. They're bringing what they know. There's certainly opportunities to innovate and be disruptive. But I think in the scramble to meet what a minimum requirement is to be deemed compliant, you're starting to see more of the same be adopted within fintechs. And I think that's a mistake, right? Why hire armies of people to review compliance when you could use technology to at least automate some of it and potentially all of it?

There's different types of tools that can be used. So, analytics we touched on earlier; there's lots of regtech analytics that can be used from a risk-surveillance perspective. Then if you combine that with things like optical character recognition, you can read through digitized documents. You can target specific documents and pieces of data. You can use that to automate the application process, underwriting, QC processes, operations and back-office functions, and compliance. And then, in addition to that, there's robotic process automation, which can take the things that a company finds from deploying those analytics or that OCR, and that can be used to kick off workflow functions, set up next steps for the team. It can do things like sending letters, placing phone calls, entering data into servicing systems. So, there's lots of opportunities to use these different sets of tools to create a really comprehensive, automated operation.

MJ 20:30 So, it sounds like it may be best to start from a completely blank slate and imagine what the best future case could be from the very beginning, implementing things like machine learning to automatically detect issues rather than build dashboards and keep checking on analytics.

And then, once things are identified, you'd be passing them in their totality, not just using a sampling methodology, through some sort of automated process where a computer may be reading them, extracting handwriting, things like that, performing whatever regulatory tests would be required. I know you mentioned earlier, the mortgage industry, where there are hundreds of tests and requirements around when someone can and cannot receive a loan, what's acceptable based on Fannie Mae, Freddie Mac, etc.

And then, of course, the way that that information gets captured, stored, and then used to drive either regulation, and send that over to regulators, or just simply better business processes to drive more efficiency and ensure that firms are doing the right thing. And all along that whole journey, it seems like technology can be plugged into current, existing platforms. Or it might even create things that are brand new. So, it does seem like there's an enormous amount of opportunities there.

But I think there seems to be a lot of value in integrating all of that into a single workflow or single process. You mentioned that there are a lot of different firms that perform a single piece of that. But how would you bring all of those pieces together in order to create something that is truly revolutionary for your clients?

MC 22:05 I was talking earlier about regtech companies. There's lots of regtech companies out there that sell analytics. But that's really all they have. Or they have a very rudimentary workflow system that still requires a human to operate it. There's lots of different tools to do RPA. There's lots of vendors out there pushing OCR.

But there's nobody that's really packaged all of it together to create a digital workforce and really drive that end-to-end solution. I think each one of these tools is good on its own. But if you can combine them together and do the end-to-end process, I think that's really the next step in making it work. And the goal here, really, is [to] reduce the amount of time spent trying to find information, use automation to really move through your institution quickly and efficiently, and solve problems rather than just serve up issues that people need to do more work on.

MJ 23:16 Perfect, and where do you think we are in that journey of moving into this automated process? Do you think COVID and remote work is impacting the industry at all? And where do you see this moving in the next five to ten years, both at smaller firms as well as larger firms?

MC 23:35 Once COVID hit, digitization really took off, along with the rise of fintechs and digital assets. Certainly, there is a trend and a greater reliance on automated routing and processing, more efficient use of imaging technology. But I think we're still at the very beginning of that. Most companies that I go into either don't digitize documents at the start of their origination cycle or their application process, they kind of do it at the end; or they'll go out and purchase a loan portfolio that doesn't necessarily have all the documents digitized. They may spend some money to have that done. They may not until there's an issue where they have to.

I think we're still in early days. Next five to ten years, we're really going to see this stuff take off. COVID certainly pushed us ahead a couple of years. And you have companies a lot more focused on spending on digitization. I think part of that too is this remote workforce. You kind of have to have it in order for there to be efficiency while working from home. I need to be able to pull up documents. I can't go to a warehouse and pull things up, right? Also, is it really the best use of my time to be flipping through documents and doing QC when I could have OCR and a bot basically tell me everything I need to know? And I can do it on 100 percent of the population versus just the handful or sample that I can get through in a day or two.

MJ 25:20 Yeah, it also seems like there may be some trust issues that had to be overcome as people are now finding themselves permanently, more or less, working remotely. It felt like back when everyone was in an office, all of the documents and information could be stored in a single physical location. But now that that's no longer possible and people need to be mobile and working from all different locations, it does become a little bit more difficult for a compliance officer or someone in that team to be able to monitor that and ensure, yeah, that it is actually effective.

So now that we've spoken quite a bit about some of the tools that are out there, what other different firms are supplying and ways that we can improve this for clients? What's your approach? Or how would you approach a client and try to integrate technology into their overall processes?

MC 26:10 I think building trust and starting small is the right approach, right? Lots of firms had been burned in the past by vaporware, pie-in-the-sky promises that a certain company's solution will solve all their problems.

We approach it differently. We will work with you, collaborate, do a pilot on a very small set of problems, prove that our capabilities will actually work for you, and then move on to a much larger engagement.

I think a lot of firms tend to take the shotgun approach. And they'll take lots of different things, try and ram them down their businesses' throat. It's not organized correctly. It's not project-managed the right way.

We try to prevent that from happening. We have a very methodical project management approach. We have a core team that focuses specifically on each client. You're getting a really high-touch, high-quality leadership team that not only has technology expertise but has also worked in industry. We have people who have been chief risk officers, chief financial officers, chief compliance officers. So, we understand, from the business perspective as well as a technology perspective, what the right way to approach these things are.

MJ 27:35 And starting small like that and then scaling up, is there any sort of risk or upper limit of the amount of information that can be processed? Or have you seen that once you do something on a small scale, it does actually work quite well on a large scale as well?

MC 27:51 I think the tools that we're using, we're pretty technology agnostic. So, we can use pretty much any cloud provider. We can use any different set of analytics. We can build custom solutions. In terms of sheer processing power, we can scale up pretty quickly.

We worked on an engagement earlier this year where a client actually had us deploy OCR to find issues in their underwriting process for loss mitigation. We were able to blast through, I think, something like three million pages in a couple of days. It was more than had ever been processed in one pass. We had to come up with some creative solutions to do that. But now we have that expertise. It didn't really derail the project in any way, shape, or form. And, ultimately, it saved that client millions of dollars in manual review.

MJ 28:47 Wow. Well, we've certainly covered a lot today. It's been very exciting talking about the new technology that's coming out, ways that that's revolutionizing, in many ways, an older, more traditional-style industry. And you've done a great job talking about each of the different pieces and how important it is to integrate them into a single process flow to respond to regulation, and just generally beef up and ensure more efficient processes within a business.

I want to ask, is there any final thought or anything you'd want to wrap up for all the listeners today about the topics we've discussed?

MC 29:22 I think it starts small. We're still at the very beginning of deploying these technologies. Over the next decade, they are going to be increasingly more common. In fact, they're probably going to replace a large portion of your workforce. So, start embracing that early, because if you're not one of the first movers, you're going to get left behind. And I think working with an organization like BRG, our group, the Global Applied Technology team, and bringing our institutional expertise and subject-matter expertise, as well as our technology stack, we will be able to get you moving in the direction of implementing technology as part of your ops very quickly.

MJ 30:12 Great. Well, thank you so much, Mike. It's been wonderful chatting with you today. And looking forward to meeting very soon. Yeah.

MC 30:17 Same here. I appreciate the invite. [laughter]

MJ 30:21 Absolutely, have a great rest of your day.

MC 30:23 Yup, you too. Take care.

MJ 30:24 Bye.

MC 30:25 Bye.

MJ 30:26 The views and opinions expressed in this podcast are those of the participants and do not necessarily reflect the opinions, position, or policy of Berkeley Research Group or its other employees and affiliates.