A massively more inclusive credit score, with Charles Wandia
The story of East African fintech is built around the mobile phone. Not the smartphone, mind you, but the humble feature phone. In markets where all but the most significant of locations lacked an established infrastructure of landlines and local bank branches, these phones leapfrogged both: powered by pre-paid sim cards that often served as a secondary currency and cleverly leveraged USSD technologies. But it has been seventeen years since M-PESA arrived on the scenes and yet traditional credit tools have often failed to cross the floor - in Uganda, the credit bureaus are filled with the data of there are just 2.4 million traditionally banked consumers, while a further 14 million sit waiting, with their data-rich mobile banking histories all but ignored.
But no more. In today's epsidoe I'm speaking to Charles Wandia of gnuGRID who, with Airtel, have built a national-level mobile credit score.
gnuGRID is the first and only indigenous credit reference bureau in Uganda, enabling financial inclusion through credit information sharing - and you can find them online at https://gnugridcrb.com/
You can also watch the official launch of that score at: https://www.youtube.com/watch?v=4VbEI_0l57A&t=19s
gnuGRID is also on LinkedIn, of course, as is Charles Wandia: https://www.linkedin.com/in/charles-wandia-983b4a65/
As you'll hear on the show, Charles also offers tailor-made training on credit scorecard development via Credit Tick Consulting at https://www.linkedin.com/company/credit-risk-tick-consulting/
Speaking of LinkedIn, that's where you can also find and connect with me: https://www.linkedin.com/in/brendanlegrange (please do reach out, follow the show's page, and share the content with your networks)
Meanwhile, my action-adventure novels are on Amazon, some versions even for free, and my work with ConfirmU and our gamified psychometric scores is discussed at https://confirmu.com/ and on episode 24 of this show https://www.howtolendmoneytostrangers.show/episodes/episode-24
If you have any feedback or questions, or if you would like to participate in the show, please feel free to reach out to me via the contact page on this site.
Keep well, Brendan
The full written transcript, with timestamps, is below:
Charles Wandia 0:00
Now, we don't want to focus on the 5% but rather on how do we bring the 95% on board?
Because I have a view of how much potential it has to change people's life. What I'm telling you is that in the banking environment in Uganda, there are about 2.4 million borrowers and all the bureau's are targeting the banked population; on the mobile wallet, there are 14 million. So if there is a bank willing to scale, this is the way.
But you know, in credit, you have to come with data. So how do we target? We worked with Airtel to say let's bridge this gap: why don't we utilise that data together with bureau data and come up with a credit score?
Brendan Le Grange 0:49
I love a good adventure story. Vine-draped impenetrable forests filled with wild beasts, lost rivers, and the promise of untold gold. It's what inspired my own books - Drachen is available for free as an ebook on Amazon and Butterfly Hill for just a few bucks - but more to the point, it is what inspired my line of thinking for this intro.
You see, today we're talking about Uganda, a country that in my colonial-educated brain is immediately associated with David Livingstone and Henry Morton Stanley, and the search for the source of the Nile. It turns out, though, that that story is more complicated than I remembered, and that the true source of the Nile is probably in the hills of Burundi, so yeah...
Still, Bwindi Impenetrable Forest with its mountain gorillas and the Kibale forest with his chimpanzees, and well, one of the (disputed) sources of the Nile are calling me, so perhaps I'll visit soon.
Welcome to How to Lend Money to Strangers with Brendan le Grange.
Charles Wandia, Head of Data Science at gnuGRID CRB, the first ever indigenous credit reference bureau in Uganda, welcome to the show.
Charles Wandia 2:13
Thanks, Brendan. It's a pleasure to be here.
Brendan Le Grange 2:16
Charles, you have been working in data science and credit scoring in East Africa for nearly 15 years now. And that's very exciting for me to see, because when I worked for Barclays in Africa, which would have been 2006-2007, there was almost no data there to work with. The bank systems weren't really up for the task, either. And as a result, there wasn't a lot of talk or focus on this idea of data science and credit scoring and data-driven strategies.
But just a few years after I finished working in that market, you started working there, and it looks like things have already changed. So if you'll indulge me a bit, could we start by discussing your early career, but in that context of what's been happening over the last decade in East Africa in terms of data science and credit scoring?
Charles Wandia 3:05
So I was born and raised in Kenya and I went University of Nairobi, to study statistics. And I believe that's what piqued my interest in data science.
Then I went on to start my career as a risk professional, I specialised in credit risk at Equity Bank. And then I moved on to join NCB that, at that point, had partnered with Safaricom to give digital loans. Safaricom is also the biggest telco in Kenya. So the telco will provide the data, and then the bank will do the data science, analytics and provide customer with limits.
So I came in there to help in them decisioning that's the product that was called M-Shwari, became very popular. And then, as I was still working in that bank, there was another partnership again with Safaricom to introduce an overdraft facility on the mobile wallet. Again, very popular at the time I was leaving, the word is buzzing around 12 million US dollars per day, and then 80% will get paid the same day. So it was really remarkable in terms of widening the financial inclusion.
Brendan Le Grange 4:21
Yeah, and in there already, it's clear the role that mobile data plays in banking in East Africa. Safaricom, as you said, big telco company, but also a primary driver of financial services in the region, at least in terms of reaching the majority of the population in East Africa.
In other markets around the world. There's so people talking about how can we use telco data or will we get telco data, you've grown up and it's been a pillar of the data environment.
Charles Wandia 4:49
If you look at Africa, it's just now that the smartphone penetration is increasing. Most of the population especially in the rural areas, they have a feature phone and that's their primary bank account, they will never probably go to a bank and open an account, because they are doing some financial transactions, buying air time, saving on their phone.
People no longer here in East Africa required to open a bank account as long as they can put money in their phone.
So I believe that has really contributed to the growth. In my career, I also went to work for one of the largest banks to start digital lending. And I pioneered a product that would give about 30,000 US dollars without collateral. So at that point, people thought a you taking too much risk. But using data, it turns out to be better than a same decision done by a credit officer.
Yeah, you know, you remove the, the manual emotions, and you standardise it. So the portfolio that was digitally assessed, that portfolio was performing much, much better than portfolios that were still being managed by credit professionals.
Kenya has really grown, if you compare the period you are in, in Barclays, for instance, credit information sharing was just starting. It was until 2013 that, you know, credit bureau framework was published and became legal. So most bureaus really started operations there. And that has grown significantly. And that I believe, is also what really has helped banks reach these previously targeted markets.
Now, if I look at the same period, if you compare 2006, there wasn't M-PESA. So again, in Kenya, right now, we moving away from cash into a mobile wallet, which m PESA provides. If you look at smartphone penetration, it's increasing fintechs have gone in in numbers to use smartphone data, which they can scrub to make some credit decision.
So yeah, I think the industry has moved quite a lot since you left.
Brendan Le Grange 7:02
You've had the chance to work at a remarkable period, a real concentration of the sort of innovation that's happened over 60 years, you know, the markets all happening within 10 years in this very small region, that sudden burst of East African fintech. And we'll talk a little bit more about that as we go on.
But Charles, in the middle of last year, you took on a new challenge, or actually two new challenges, but we'll we'll get to that later. For now. You jumped across the border and went to help build Uganda's first ever credit bureau.
So talk to me about that decision, but also about Uganda, what does the lending landscape look like? They're what sort of players are involved? What are the borrowers look like? What's the basic lay of the land if we want to picture it in our heads?
Charles Wandia 7:52
So in the middle of last year, I crossed the border, and the landscape is more or less the same. But if you look at the East African countries, Kenya is always leading in terms of development. And therefore, when an opportunity came knocking to kind of replicate the same successes we've had in Kenya in terms of digital lending, I couldn't say no, because I have a view of how much potential has into changing people's life. So I moved to Uganda.
It's not really a new bureau, credit bureau. There are other bureaus, which have been here they are multinationals coming into this market, but new grid is an indigenous credit bureau. So it started by locals. You know, the whole intention is to say, we understand the market much better. We know there's still some populations which have been left behind. So gnuGRID is mostly targeting the last mile population or really the financially excluded population.
So we are always asking ourselves, How do we bridge this gap? They are invisible to the lenders because they don't have data. So we collect a lot of alternative data. So only 5% of the population is banked. And all the Bureau's are targeting banked population. Yeah. So what about the 95% who are unbanked? who borrow in informal systems like Sacco's village? You know, groups, what about them? So that's really the reason why I came here.
Brendan Le Grange 9:27
5% Banks is a tiny number. Yeah, it'd be really interesting to hear how you're reaching these these parts. I want to start though, with the one project I do know that you're doing and that's the work you're doing with air tail.
You've just launched the first ever mobile credit score in Uganda. It's how I came across your profile on LinkedIn. Tell me more about that. What is this mobile credit score, and what role is it going to be play in lending to this last mile population? Yeah,
Charles Wandia 9:55
Excellent. So gnyGRID is focusing on providing solutions over the last mile for excluded customer base.
So there are about 22 million borrowers here who don't have any credit history, they've never borrowed in a bank, and they are living life. So they have working capital needs or consumption credit. And really, if you actually coming from a very poor background, it's very hard to get such credit.
Brendan Le Grange 10:23
And expensive.
Charles Wandia 10:23
Now, we have banks who are saying, Okay, we know you need that but in credit, you have to come with data, we have all these digital lenders who are trying to target them. But again, they also can't afford to do data science, or they don't have the skills.
So that's why we work with Airtel to say, let's bridge this gap. Let's try to be the boundary between the lenders and the borrowers. And the only way you can do that is having standardised credit score.
So Airtel agreed they have a lot of transactions, most of the last mile transacting on their phone. So we came together and said, why don't we utilise that data together with bureau data and came up with a credit score, which the lenders can then use to target these last mile population.
So Airtel provides all the transactions when you buy airtime on your phone, when you buy data, pay a bill, you know, that information tells us probably you have some responsibility in your house, you're moving around, probably you have some kind of mobility, are you having so many people different one sending to you, tells you you're making sales.
But if you're just receiving from one person, we can can infer probably your law student getting some update from the parents.
So when you look at that data, in that sense, it start making some sense, when you look at the balances with mobile money is actually your main bank account, studying the balances can tell us something.
So we looked at a lot of data and built a credit score using machine learning.
And it can be relied on by lenders, but lenders don't just need the credit score, they need a view of how much you probably earning so that they can able to tailor make your limit. So we went ahead and provided income estimates. So it's not just a credit score.
And basically, we've got a lot of excitement in the market, we have very many lenders coming to us and saying, okay, when can we start? In
Brendan Le Grange 12:25
Almost every market in the world now we'll talk about, oh, we should get telco data on it. In the UK telco data's on the credit bureau, but often when we talk about telco data, we simply mean are you paying the bill at the end of the month is one data set that looks a lot like a small loan. But the data you're talking about is far richer than this. So this is prepaid mobile, I'm imagining, you can see how often people are topping up where they topping up where money is coming. And going all through the month getting insights, which is, is fascinating.
As I said, not long ago, we had no data. And now you're able to take this big data that's structured but but very different, and put it in a machine learning and build a credit score.
And not just a credit score, as you mentioned, but a score that includes affordability. So this is a massive leap forward in terms of what a lender has access to when they're making that decision. And I know it's very early days, because you launched the score just weeks ago. But I'm interested to know how much of that 22 million unbanked population you can now reach?
Charles Wandia 13:34
Yeah, so if you look at the numbers, the active borrowers, the people who can really qualify, they're about 14 million from the 22. What I'm telling you is that in the banking environment in Uganda, there are about 2.4 million borrowers. Now, on the mobile wallet, there are 14 million. So if there is a bank willing to scale, this is a way.
Brendan Le Grange 14:01
Yeah. So at the moment, about 11% of the market has been served that you're opening up and represent. And of course, then if we think about the people that are still left out of that, they're more likely to enter via the telco route as well.
And if we go a little bit more technical, I'm quite interested in your experience, you'd build credit scores, the old fashioned way, like we all know, in the banks. How was it different working with this data?
Charles Wandia 14:28
Yeah, basically, the credit scoring methodology is more or less the same. We use logistic regression, as well as machine learning, but when you're dealing with alternative data, you have to go a step farther and try to figure out what can predict default.
Traditionally, banks rely on bureau data, which is limited, but financial footsteps, which is what we getting from the telco is more richer, in terms of you know, telling you whether someone can afford to pay short term loan. Again, you want to make it very easy for lenders to understand your score.
Because you know, credit, you must make credit using informed decisions, it has to be explainable to the borrower. Yeah, that's probably one of the things we've been trying to navigate in terms of being explainable to the borrower
Brendan Le Grange 15:19
Around the world, I think every market is thinking about how they could use telco data. And as I hear you talk through the types of data that make up telco data in Uganda or in East Africa, it also speaks something towards open banking and other markets where the transaction might be happening on a different sort of product, they may not have a situation where only 5% of the people sit on the bureau with a bank account, but they still interested in this sort of data because we can add so much nuance.
Charles Wandia 15:49
So I think if you look at the whole mobile credit score, and this telco versus bureau partnership, I think that's something that should be studied further, even in other markets, you know, we are allowing the telco to monetize their data, they have a lot of data sitting just there. Again, look at what we're doing in terms of deepening financial inclusion. So you monetize your data, as well as playing a part in the financial inclusion landscape. Yeah,
Brendan Le Grange 16:17
quite possibly give that person the ability to buy a another phone, to expand their usage of telco services, to build a business and make more money, which allows them to borrow more again. So it really is a symbiotic relationship we're creating here where we're saying, hey, yeah, we'll help the customer together. Because all of a sudden, people have access to a much broader marketplace of lenders. So really a powerful story. And yeah, anyone listening should be following the story just to see how this shapes an economy. And then as a second level of self interest. We all have customers with phones, we all have alternative data, we all have this question of affordability and risk. And what can transaction data tell us? And that we can also learn from what you're doing with new credit? So if people wanted to learn more and follow that story online, what are the socials? What What's the web page they could go to, to stay on top of it.
Charles Wandia 17:12
So we launched the score like two weeks ago. So the story is widely published. But if you go to new greed LinkedIn page, you will find the story there. If you go to Airtel money, Uganda pages, you will find the story there. We have a new grid website, g n Ucrete. If you look at that website, there is there is a story there. Again, you can always reach me, through my email just.kimmitt@gmail.com, I'll be more than happy to take you through. But basically, through the social media platforms, YouTube, we had, you know, major TV stations in Uganda cover their story, and TV, UBC, you can find the story and follow.
Brendan Le Grange 17:59
Yeah, I'll link all of those in the show notes as well. Charles is not the only product that you do get new credit. So before we move on to other topics, what other products are you serving to lenders,
Charles Wandia 18:10
So we have the traditional bureau score, which we extend to lenders, we have quite a number of banks who rely on acid digital lending, so they use a score for digital lending.
And if you're a bank, and you're trying to do acquisition campaign, you know, we can identify profitable segments for you.
We are doing collections for for banks. So we have also collections models. And then if a bank needs to do account management, maybe you need to increase limits for for your borrowers. Tell you the scores of the borrowers. There are other exposure that you probably haven't seen. So again, we still continue providing the traditional Bureau services.
However, what sets us aside is now we focused on not just banks, the bureau is allowed to collect data even from other credit service providers, as long as they have been accredited. And that's why we focused on where most of the last mile burrowed in circle, and all these digital lenders, so you're allowing them to share data with us providing the technology so that then through that information sharing, we can then create a system where the lenders can provide credit in a risk free environment and then the borrowers can benefit.
Brendan Le Grange 19:31
A while ago on the show, I had Dillon Harandiran on the show, who is now looking at solving different problems in the world, but at that time, he was trying to address this in the UK, you know, the fact is fintechs and alternative lenders struggled to participate on the credit bureaus because the technology to share their data was so clunky and hard to use and slow and everybody loses when it's hard for new players to join. So it's great to hear that you're already a step ahead here bringing your new lenders and providing them with tech to do that.
Obviously new groups not afraid of innovation, you're doing some pretty cool stuff. As you look forward. Is there anything on the horizon that we should be keeping an eye out for?
Charles Wandia 20:12
Of course, we continue commercialising the mobile credit score in Uganda. And beyond Uganda, we've had a lot of interest in other markets, where the telcos, they're telling us, okay, how can we replicate this?
Brendan Le Grange 20:26
Well, I think it was MasterCard that just took a big share of mtN mobile banking business. So a lot of interest from from further abroad as well into this larger ecosystem. So I'm sure your phones and your emails are going to be busy as people try and see what you're doing and watch it grow.
But Charles, I said earlier on, this is not the only new challenge you personally took on this year or last year, you also offering tailor made training on credit scorecard development. So talk to me a little bit about that. Yeah, so
Charles Wandia 20:57
My experience is quite diverse, and mostly targeting digital lending. So I thought I should start a consulting firm to do that, for very many financial institutions, who have no idea how to go about it. So my consulting company is called credit risk tick.
And mostly is to offer digital lending consulting, if you want to start digital lending and your financial institution, you have no idea how to use this course how to build this cause how to digitise your lending processes, I do offer that consulting.
Now I also do training on credit scoring. So people reach out to me, they wondering how do we do credit scoring using these telco data or alternative data and really grown in a very short time.
Brendan Le Grange 21:52
I mean, it's often easier for the data to spread and for those skills to be there to analyse it. So it's great to hear that you are there to to help alongside that when you're talking about going from serving 2 million, 4 million people to go into serving 16 - 18 million people, you can certainly see why there's going to be a vacuum and a rush for people to start building scorecards. Again, I think a project that's going to be very interesting to people is saying, is there a separate website for that? Yeah,
Charles Wandia 22:18
For now, you can reach me on LinkedIn, Credit Risk Tick Consulting, or you can search for me, Charles, one year, you can send me a message and then we can get in touch. You can also reach me on my Gmail charles.committee@gmail.com Yeah, and I'd always respond to you.
Brendan Le Grange 22:39
Perfect. And once again, I'll include those in the show notes. Charles, thank you so much for making the time as I say, I love the market, because I had a little bit of experience working there. But I do feel like I missed out on so much. Because as I was leaving was when the first and M-PESA was rolling out. And I never really got to experience on the ground, that sudden wave of mobile banking of mobile lending of FinTech in general that's come out of the region.
So wonderful to hear what's happening. So yeah, Charles, thank you again for your time. It's been wonderful learning from you.
Charles Wandia 23:10
Thank you. Thank you so much, Brendan. was a pleasure to be here. And
Brendan Le Grange 23:14
thank you all for listening. Please do look for and follow the show on your favourite podcast platform and share the updates widely on LinkedIn. Plus, send me a connection request while you're there. This show is written and recorded by myself Brendan Lagrange in Brighton England show music is by I am weak and you can find show notes and written transcripts at WWW dot How to Lend Money to Strangers dot show. And I'll see you again next Thursday.
And thank you all for listening.
Please do look for and follow the show on your favourite podcast platform and share the updates widely on LinkedIn where lending nerds are found in our largest concentration. Plus, send me a connection request while you're there.
This show is written and recorded by myself Brendan le Grange in Brighton, England and edited by Fina Charleson of FC Productions.
Show music is by Iam_wake, and you can find show notes and written transcripts at www.HowtoLendMoneytoStrangers.show and I'll see you again next Thursday.