Where Credit’s Due Ep. 4: Behind banks’ journey to digitize commercial lending with OakNorth and Numerated

Today we’re talking about how banks are digitizing their commercial lending operations – if consumer lines took the front seat when it comes to innovation and digitization, commercial lines have been waiting quietly in the backseat but are now next to be brought forward with the latest technologies and underwriting efficiencies. 

But it’s not easy to implement such change into age-old processes, and partnerships are key here, which is why we also explore how software vendors are helping banks and credit unions to widen their lending offering to more businesses in the US.

My guests today are Dan O’Malley, CEO of Numerated, a SaaS digital loan origination system, and Peter Grant, the President of OakNorth, which aims to transform commercial lending with its Credit Intelligence software product.

Subscribe: Apple Podcasts I SoundCloud I Spotify I Google Podcasts

The following excerpts were edited for clarity.

How did the digital commercial lending landscape change after the pandemic?

Dan O’Malley: There are two aspects that I would highlight – the engagement model with customers, where everybody was really forced to use digital channels for the better part of a year and a half. And then the expectations for efficiency for commercial lending – when you do a year's worth allowance in about a month or two, because of the PPP program, which is what most banks experienced, your eyes are open to lending a little more efficiently.

Peter Grant: It's certainly highlighted some deficiencies in commercial lending compared to its peers in the consumer market. And it's also a shock event, something we hadn't seen for 100 years. I think what it also highlighted was the ability to look forward and then have the granularity to see the differences between certain businesses that may share the same NAICS code. 

Also, the regulators have also been really stress testing and giving vulnerability ratings to banks’ books at a level of detail they've not seen before. 

The third thing to add is that we haven't seen this in a credit cycle. Given what's coming around the corner, this has been a good training ground for what may come ahead with high inflation rates, interest rates, and everything else as well, for some of the younger underwriters.

How do you think banks can upgrade their commercial lending infrastructure?

Peter Grant: Every conference I’ve attended recently starts with “we need to change, we need to transform, we're going through a Darwinian event”. There were 16,000 banks 30 years ago, there are now just over 4000 banks here in the US, and a lot of that's driven by digitization and alternative types of lending. The fintechs are very competitive out there in this space. 

From our view, we serve as a partner to the banks to help them digitize. But really, it's the data and the prioritization of the projects that banks are looking at currently. Some of our bigger customers, who have gone through transformations already, can get data out of their systems quite quickly. Some of the mid to smaller sized banks struggle -- they're still using XML or paper based processes, which makes it really difficult for unstructured data to get out to start their digitization projects. They're struggling to see which ones they should prioritize. 

Dan O’Malley: There's so many ways that data and processes can flow out of a digital system, you have to keep track of things in multiple places, you've got data over here and data over there. To me, the end result of a full digital transformation process is you've got one set of rails for data, and it all comes into the same place, and now can all be reused. 

There’s steps of digital transformation, it's pretty hard to get all the way to the end and actually eliminate all of those ways where loans can go off the rails – very few banks have done that for all of their loans. If you're a bank going through digital transformation, you have to ask what's most important. We see banks taking different paths and which one is right for a particular bank depends on where they’re at. 

What kind of digital lending tools and capabilities can banks find through vendors? How can these improve underwriting and help them with reaching more customers? 

Dan O’Malley: I do think rethinking underwriting now is getting a fresh look, and amazingly across all credit types, from the tiny ones that should just be fully automated, to a larger commercial loan, where it's going to be a person doing the work, but that person now expects help. An ML based solution like OakNorth is really going to help you dig in there and find your areas of weakness, for example. 

It's kind of amazing to me – when I started working on digitizing business lending, it was almost 7-8 years ago now. I got called the heretic, even unethical once, because I wanted to make the lending process easier for business by using technology -- underwriters were just dug in. And you compare that to last week's RMA conference, an underwriting conference, where there was this openness, everybody was looking for solutions. 

Peter Grant: The fintechs can move really quickly, they're very nimble, they have access to capital, they can improve their customer experience. The big banks now need to make a decision about going to the cloud first of all – that's really the infrastructure layer, whether it be on Google or AWS. Then you've got your middle layer, which is basically your plumbing for your applications. Then you've got your top layer, which is your Numerated, your OakNorths of this world. 

Banks are still grappling with what they need to do and where they fit into the process, because there's a huge change in management that needs to happen as well. So it's not just the investment in technology, it's investment in the people that you're bringing into the organization that understand how to use the technology and get the benefits from it. You need to be attracting the type of people that fintechs have as well. In their own words, they do need to bring some younger blood into the organization who expect to use these types of tools. 

What makes a good partnership between a bank and a software vendor? What are some other factors that can yield a successful outcome?

Peter Grant: Openness is really important – banks historically have always leant towards building their own technology, as opposed to buying or certainly partnering with the fintechs. And not all fintechs are out there to eat their lunch. We're not a lender here in the US, we're a tech company. But banks will sometimes say, “Well, I'm going to build that myself”. Certainly the bigger banks say that. Why is that the mentality, instead of looking to someone who’s already provided a solution, who's using something that's successful, that can actually help them ramp up and be successful? At the end of the day, banks’ product is lending money, and the technology helps them do that. So it's really changing their mindset, going from a build to a buyer type mentality.

Dan O’Malley:  My perspective would be alignment of vision for what the company or the fintech is trying to do and what the bank wants. Our vision is transforming how banks lend to businesses, not making it a little bit better, not tweaking them, not being a platform for getting your data to sit there. We will change how you lend. And so you either hear and go, ‘Man, I got to do that, too’. Or you're like, ‘I don't know, do I need to transform? It sounds big and expensive’. So it’s about alignment of vision, because we're going to work together for a long time. 

What are some of the challenges that might arise along the way? 

Peter Grant: Interestingly, it takes a while before the banks actually make a decision. During that journey, you do get to know your customer pretty well. The important thing is your way of selling. Tech was sold top down – you basically went into the board room, you sold them an amazing ROI and amazing vision, they bought the product. And lo and behold, they tried to deploy it, and lots of the end users wouldn't use that particular product. So you need to make sure during that journey, that when you do have the executive alignment and buying in the vision, you also need to bring the users on the journey with you to make sure that everything's in lockstep. 

The reason why I highlighted it takes a long time is because you need to build trust with the customer. So things are going to go wrong – whether it be getting data out of your systems, whether a certain individual on the project is not pulling their weight, whatever it may be, you need the ability to fix that. Users should be able to share things with you so you can understand what you're developing to make sure you're deploying something that is going to get used, has value, and does have an ROI as a particular product. So sharing best practices is really important. 

How is underwriting changing? 

Dan O’Malley: There's a lot of opinions and maybe a few facts, because so much is changing. My opinion is that we're going to see that banks are more open to real life cash flow underwriting, as opposed to spreading last year's financials and expecting next year to be the same. Seems like next year has not been the same for a couple years now. Maybe we should just accept that. 

The other thing I think we're finally starting to see is a belief that models actually matter. And it's not just art, there's much more science, and not just for $10,000 credits, but for bigger ones. I'm an ex Capital One guy, I built a lot of my stuff, I'm a data scientist by training. Models matter, they really do. Even if you're not always making 100% of the decision with a model, why not let it raise the questions that you should have to answer. I think that's just a beautiful thing. And I'm happy to be sitting here in 2022 to watch that happen.

Peter Grant: I think that one thing we didn't address is actually people and talent .You can't be a complete black box, you need to be a glass box. But every conference I've gone to in the last month, everybody said where do we get these people from? They're really struggling with the Great Resignation, and the pandemic highlighted some of the weaknesses and the ability for banks to change their behaviors. I work from home using Zoom's modern technology and everything else. So that's all a good thing. The difficult thing is, how do you find the people? The good thing is through digitization, you don't necessarily need as many people – you still have the same people, but you can deploy them to do more human oriented tasks to drive better efficiencies in the bank as well. So just put your thinking cap on and really get your head around that, and plot out a vision and what you're trying to do with your vendors.