Sea Group Stock Price 29% Drop, Digital Bank Strategy Synergy & Fintech Underwriting Quality - E305

· Fintech,Southeast Asia,Start-up


“With fintech, people over-index on the math of why someone should do something. They forget about the psychology of someone buying that financial product. Most financial products don't get bought very often if you think about it. It's not a daily decision you're making. There are a lot of considerations that go into purchasing a financial product, and they're not all math. Most people actually are emotion-driven, not math-driven. They forget to talk to their customer about their considerations. They go straight to the numbers and the math, and a lot of people don't enjoy that because it doesn't resonate. They don't engage with it well.” - Shiyan Koh

“Neobanks were a huge category that got funded, but the report indicates only about 5% of them are profitable. This is where they run into that problem, which is you can go out and get a lot of users, but if you're not disciplined about how you underwrite and monetize them, it can be a money-losing proposition. It's on the neobanks and incumbent banks to improve the product experience and make customers stay even if other places offer slightly more interest.” - Shiyan Koh 

“People tend to focus a lot on the cost of acquisition because that seems to be a more straightforward go-to-market. They want to buy and sell. A lot of aspiring fintech founders end up getting shocked by the cost of capital, especially since there's been a lot of change over the past three years. So a lot of models that seem to work, don’t work at different levels of cost of capital. A huge amount of humility around the accuracy of the underwriting is going to be key. When you build a company in your deck, you're going to know your cost of capital pretty fast. Your cost of acquisition is doable if you do a fit test, and you understand the vertical well. I've seen some fintech founders fail when they don’t think about how each of these factors could change over time.” - Jeremy Au

Jeremy Au and Shiyan Koh, Managing Partner of Hustle Fund, discuss three pivotal topics:

1. Sea Group's Strategy: They talk about the Sea Group’s 30% stock drop despite its profitability since Q4 of the previous year. The contraction is attributed to increasing competition from Alibaba (Lazada) and Bytedance (TikTok) across Southeast Asia, decision to focus on growth to compete and the US public markets pricing in their assessment of the situation. Time will tell about whether this strategy (and corresponding execution) works out. Public markets are a voting machine in the short term and a weighing machine in the long term.

2. Digital Banks and Underbanked Market: They discuss the under-banked segment in emerging markets, where incumbent banks often focus on high-end customers. New players like Sea, Grab-Singtel, and Standard Chartered with NTUC Fairprice are entering the banking scene to target this underbanked segment and leverage their vast distribution networks. They also share the challenge of identifying the right financial services to offer to ensure profitable operations. They also discuss how revenue and cost synergies apply differently to various players and their business models.

3. Southeast Asia's Fintech Models: They talk about the importance of founders having a clear understanding of their business model, especially on net interest margin. They underscore the value of knowing the cost of capital, acquisition costs, and the quality of underwriting early on to allow the business to be proactive rather than reactive. They also discussed the significance of understanding customers’ buying psychology and reminded founders not to overly focus on mathematical aspects.

They also delve into the rise of neobanks, benchmarking in the fintech realm, the essence of humility in founders, the crucial aspect of aligning with market realities, and how understanding underwriting accurately can shape the success trajectory of a FinTech startup.

Supported by Baskit

Baskit is a company focused on digitizing Indonesia's supply chains. With a focus on empowering local communities, Baskit recognizes the immense potential within the 200,000 distributors and wholesalers scattered throughout the country's vast landscape. Unlike disruptive approaches, Baskit takes a collaborative stance, working hand in hand with these traditional businesses. By infusing technology and providing access to financing, they modernize operations and create efficient supply chains, benefiting manufacturers and consumers along the way. For more details about Baskit and their journey towards revolutionizing digital commerce and supply chains in Indonesia, please visit their website:


Jeremy Au: (01:17)

Hey, morning Shiyan. Good to see you.

Shiyan Koh: (01:20)

Morning, Jeremy. Jeremy was quickly buttoning his shirt so he wouldn't have too much of a deep V on camera. For all of you listeners out there.

Jeremy Au: (01:28)

I know.

Jeremy Au: (01:29)

I think it's the difference between one button and two buttons and this polo tee. got to keep it conservative here. So talking about conservative, and the news, there has been, a lot of news recently about law, talking about Sea Group, and lots of different things. So, let's talk about it, the big news that you wanted to mention, Shiyan.

Shiyan Koh: (01:47)

I think Sea’s stock is down by like 30%. Huge drop. I think probably the biggest since they've been public, and it's interesting. I think it's all about expectations because they actually did, they were profitable this quarter. They've been profitable since the Q4 of last year, but you're seeing a contraction in the gaming business. And I think it's more that the CEO, Forrest, had said they're basically planning to spend more to try to gain share on the e-commerce side for their Shopee business, which I think makes people nervous. And it sort of indicated also a willingness to go back into loss in order to pursue that market share lead.

So I think investors punish the stock on that, but I think there's some really good spots in the earnings report, right? The financial services business is growing, and that's what basically they've been taking their profits from the gaming business, funding the e-commerce, the financial services business.

And so, perhaps it's a matter of more communicating how the investments are going to pay off and over what timeframe, to try to give people a little bit more color and certainty around that, versus feeling that, hey, we're going to be competing against the bottomless pits of Ali and JD and everyone else in the region or in the markets where we're present.

Jeremy Au: (02:59)

Yeah, I think what comes to mind straight away is that the markets may not necessarily be right. I was having lunch with somebody and I was like, oh, see, are they doing something wrong? And I was like, well, the market is basically saying, Hey, what's our expectation of the profit that we're able to book over the medium to long term? So, we still don't know how it's done. A lot of it is dependent on the quality of execution and the velocity. I think Sea has really outperformed, over one year ago. I think a lot of folks are just very skeptical about the ability to kind of turn a profit and be able to strengthen on that basis.

And I think they were able to execute above expectations, actually, above analyst expectations, above many people's expectations. So I think this is, to some extent also like a communication and strategy shift that wasn't communicated in advance. But I think, taking a step back is also, like you said, it is a function of the competition. A lot of competition has come in. Obviously, Lazada which has been owned by the Alibaba group has continued to be aggressive, but more importantly, ByteDance, with TikTok Shop has been super aggressive and they've added a lot. In our WhatsApp, BRAVE WhatsApp group, we also recently shared an article about tikTok's rise and how many users they've been adding, but also a lot of the volume and G M V they've been growing. So I think it's a fair discussion, which is saying like, Hey, we need to compete, so let's focus on growth. and it also is fair for investors to be like, okay, this is going to be a very intense e-commerce battle again. I don't know, is it chapter two, chapter three, or chapter four? You know, it's a tough fight.

Shiyan Koh: (04:27)

Yeah, I think this is what the market in the short term is, a voting machine, and in the long term is a weighing scale.And so I think it takes a while for this to shake out and I think if you want evidence of that, you can look at VinFast SPAC IPO. Now, VinFast is worth more than GM, which is clearly absurd, given that VinFast shipped, like 18,000 vehicles last year, and I think GM shipped like 5 million.

Jeremy Au: (04:52)

Whoa. You're sure you're not missing a couple of zeroes there? You didn't see 18 million.

Shiyan Koh: (04:56)

Nope, I double-checked. I double-checked this. They just broke ground on a new EV plant in North Carolina, think, but it was an incredibly small offering. I think the offering only raised $30 million and like 95% of the shares are controlled by VinGroup-related entities. So, incredibly small float and a lot of hype on EV. So, yeah, anyway, whoever got to play the pop, good for you. But I think we can't take these things too seriously in the long term. Ultimately, value has to come back to normal.

Jeremy Au: (05:29)

And the value is that it will get up to 5 million vehicles. Is that the bold case that people are claiming in the SPAC prospectus?

Shiyan Koh: (05:38)

I am sure you can claim anything in a SPAC perspective, right?

Jeremy Au: (05:45)

Well, there is a lot. Okay.

There are two topics we have here. There's one is, EV vehicle and the bull and bear side. Oh, boy. Let's close out the Sea group. At least before we go there, we should talk about EVs after this, but I think at least, one interesting part is that every player is also not only doing that, but it also opens up banks, right?

Sea is pumped in parallel, about $173 million into Maribank. So that was the third digital native bank. So it's a function of the regulators opening up banking licenses. And then previously we had Grab work with Singtel to put in a hundred million dollars into GXS bank. And of course, I have a friend who works at Trust Bank, which was a partnership between Standard Chartered and Fair Price Group.

And for them the first out of the gate, they've done about 24 million transactions and they have about, 650,000 weekly active users. There are all fresh launch numbers, and of course, you take a step back even further. We see, like this time, it has launched in the Philippines. Sea's also present through Sea Bank, and Sea Banks's also available in Indonesia. All this comes from Tech in Asia. So thanks, Tech Indonesia for, the article, but I thought it was just an interesting dynamic where I think we see there's an e-commerce side that's obviously getting very aggressive.

And then also you see this, I don't want to call, it's a big bet obviously in terms of capital, but also in terms of management focus on banking. And it always makes me wonder what the actual end game is going to be. Because when you, and I think historically, obviously banking licenses have been lucrative, right?

It's a form of licensing that allows for a certain number of players. So you kind of get to have some level of monopoly profit because the regulators want to work with a closed group of banks, and I think we've seen that in the US where you don't see new banks being launched anymore, but you see all of that getting consolidated. And so that's how your economies of scale and economies of scope kick in really well at that scale. I think it's an interesting dynamic where in Southeast Asia, there's a whole bunch of new banks being built, and so I'm just not sure personally how that's going to play out. What do you think, Shiyan?

Shiyan Koh: (07:41)

I mean, historically, I think in emerging markets, you have a huge under-banked segment. So there are not that many banks and they are pretty comfortable serving the upper tier of customers. And there's less competition because banking tends to be more protected, licensed as you mentioned. There's not a lot of incentive for them to go serve these mid and low-income customers. And you end up having sort of informal market, financial services, loan sharks, all that kind of stuff. And so I think the story we're telling here, as these digital bank services get launched is, Hey, we already have distribution.

That's one of the most expensive things in financial services, whether it's NTUC, whether it's Singtel or Grab, or whether it's Sea, through their e-commerce and gaming properties. We already have distribution, we already touch millions of consumers, and this means that we have a really low-cost channel to get them banked. Go from unbanked to banked or underbanked to more properly banked. And because it's all digital, we're going to be able to serve them in a more cost effective way. So I think that's the story. And then the question is, can you actually do it? And so, here's where the rubber hits the road, which is like, what are you trying to get them to do? Are you trying to get them to save? Do you want to get them to borrow from you? Do you want to get them to invest? What is the financial service that you kind of want these under or unbanked people to engage in and can you deliver it cost-effectively?

I think the challenge historically for credit services has been, part of the reason incumbent banks don't bank these people is because they're not great credits. There could be a segment that is, they're actually good credits, but they're underbanked because the incumbents are too busy dealing with high-end customers. And hey, if you can find those people, like salaried civil servants in the Philippines, underbanked, but good credits, right? Then that's worth it. But if it is someone who is underbanked for a reason because they are not good credit, extending credit to them generally is not a moneymaking proposition.

And so, I think if there is a way to close the loop, like if you're a Grab, you basically can garnish your Grab driver's wages. So you actually are holding collateral in a sense. You get paid first. Then there's a way to close that underwriting loop, right?

Jeremy Au: (10:01)


Shiyan Koh: (10:02)

You have to think about that in each of your segments, which is like, how are you better able to deliver? Savings or credit or insurance or investing products to basically people who have really small dollar amount accounts.

Jeremy Au: (10:15)


Shiyan Koh: (10:17)

And then that's interesting, I think it's an interesting problem, but it's not as easy as being like, Hey, I have distribution. They're using my app anyway. Like, why don't I just do this thing? There are actually a lot more considerations.

Jeremy Au: (10:30)

Yeah, I think that was very much how the initial reports of several years ago during the bull market, everybody was trying to build a bank and very much, I think press releases flagged up that story. And I think where I'll add to that color is that I think first of all, incumbent banks are not slow either I mean, they also have the cash, the profitability. And so for them, they're very much looking at saying, Hey, if you are educating these consumers and you're finding the best ones of the lot, then we can push them over time into the services. Obviously, this is for banks that are more focused on innovating and competing as well.

But at some level, they say Goliath wakes up right then you have to watch out. I think the other thing also I think is that there's been a thesis around building like Neobank in several countries, but the truth of the matter is that. Another level is not every country is as permissive as Singapore in allowing licenses to be given out. And if you don't have that license, then effectively, if you're not regulated by a government and you're trying to work with regulators in parallel, now what happens is that the regulators can have conversations with incumbents and incumbents may very much say, Hey, we recommend that either they're not granted a license or that they should actually have the exact amount, the same amount of compliance as we normally have to do, which effectively kills the Challenger because no challenger can ever, deal with all of the requirements that incumbent banks already deal with. And so I think we've definitely seen the collapse of that FinTech story, or at least the Neobank story in several countries where regulators have ended up saying, no, we don't like what this is headed towards in the banking sector.

Shiyan Koh: (12:05)

Yeah, and I think neobanks were a huge category that got funded, but I think the report is only about 5% of them are profitable. And so I think this is where they run into that problem, which is like you can go out and get lots of users, but if you're not disciplined about how you underwrite and monetize them, it can be a money losing proposition.

Jeremy Au: (12:29)

Yeah, because you can run out and get a lot of users, but then, a lot of users could always be not worth a lot from an economic perspective, right? Because the accounts that they're holding are relatively small. So you're spending a lot of time managing, maintaining, and thankfully your digital in terms of cost, but still, the balance.

Not huge. That's one. And then two is that for your better customers. The truth is they could have accounts in multiple accounts, so they could be using them. I think this is Auntie that I knew and she was very happy about all the digital banks, right? Because she had a main account that she trusted very much, and then she would shop around all digital banks and take all their various offerings, right?

For coupons, promotions an extra bump in interest rates. But at the end of the day, is this person actually a good customer? Well, from a perspective of a digital bank this person looks like, demographically, everything's right, but the share of the wallet or the share of the banking account is not there. And more importantly that you said loyalty at some level is not there. So that's not an insurmountable problem, but I think that's the race between incumbents and challenges.

Shiyan Koh: (13:32)

Yeah, but I mean, I think loyalty. What is loyalty? Like it's a bank, it has a job to do. So, , I think then, it's on the neobanks and incumbent banks to really improve the product experience and make it like, Hey, I can't imagine switching, even if like some places offering me slightly more interest or whatever, because. It's just so easy to deal with my bank versus, anybody else. And I feel like people don't spend enough time on most banks’ product experience is terrible.

Jeremy Au: (13:59)


Shiyan Koh: (13:59)

So that many banks won't give you statements beyond 12 months, and if you want a statement, which I just want to download honestly, they won't, and then you have to write into them and they will mail you paper statements and they will charge you $30 per statement.

Jeremy Au: (14:21)

Just how to make the money. That's the profitability is right there.

Shiyan Koh: (14:24)

Outrageous. It is outrageous.

Jeremy Au: (14:29)

It is outrageous. I mean, by the point is that your perfect customer, which customer needs a statement from two years ago? That means trying to make bank baby.

Shiyan Koh: (14:37)

Oh no. I'm closing that account after this happened. I was like, that's it, I'm out.

Jeremy Au: (14:46)

That's such is life, right? But I think also there's a strategic angle as well for these e-commerce players and so forth. So if you're a Standard Chartered at Fair Price Group, you're building a bank. Feels like you're building a bank, but I think if you're Sea in you're building a bank, or your Grab, you're building a bank that's slightly different because I think there's a strategic side of it. So what do you think is the synergy here?

Shiyan Koh: (15:07)

You don’t think that StanChart and NUTC building a bank is not strategic?

Jeremy Au: (15:14)

Well I think it's tricky.

Shiyan Koh: (15:14)

I mean, StanChart is a bank, I guess.

Jeremy Au: (15:16)

Yeah, exactly. So that's what I'm trying to say is like you're building a bank and now you're building one that's also consumer, much more consumer-oriented. And obviously, be Fairprice if all your users are shopping in Singapore NTUC, fair price.

And I was just there. I was trying to buy shoe polish, which apparently no longer exists in supermarkets. Apparently, they've been out of stock for three years. So shoe polish is a really bad SKU great fun fact. Totally divergent. But back to is that you have millions of people visiting fair price at a time, right?

So I think it's a great Go to Market channel and Snapchat obviously already has the banking expertise. So strategic obviously, but I think it's slightly different in flavor from strategy, from how Sea and Grab are thinking about it, because first of all, they're not banks yet. So what do you think is that strategic synergy from your perspective?

Shiyan Koh: (16:10)

I mean, if I were Grab, I think the drivers are the easiest people to the bank because you own them and they're income-making and you have the best sort of view. I think the consumers are hard. It's like when Grab pops up, the thing is, do you want ride insurance on your upcoming Grab ride? I'm like, no. What is this? In what world? What are you trying to tell me? I could get in a car accident when I get into this Grab, so I need to buy? That's not the right thing to tell me. No, I don't.

Jeremy Au: (16:41)

Do you need us to listen to your phone during the entire car ride? So they could protect you in case of a safety incident?

Shiyan Koh: (16:47)

Yeah, right. It's just like, no, I don't want that. I think for Sea Group, it's interesting. This commerce thing, there's a very clear, like e-comm, Escrow payments, all that sort of stuff. Especially if also, you are banking the shop owners. I sell stuff on Shopee. I could get advances because I can predict what this shop is going to make. I already have a lot of transaction history on them. And then on the gaming side, I think digital goods, digital marketplaces, making it really easy for people to transact, having a wallet, things like that.

I think they're pretty natural places to do some of these things and then maybe over time, people trust you more, and then you can start to sell more sophisticated services, but I feel like for Grab and Sea Group, it's really to start with the people who are making money on their platform because that's where they're going to have the best underwriting data.

And if you're making money, you're already oriented to see the platform as a place that is a reliable business partner versus a consumer is much fickle. And I think the cost of acquiring a consumer into becoming a financial services customer is just way higher.

Jeremy Au: (17:55)

I think that what was interesting as well is that there's also a cost synergy. So if you're a Sea and to some extent Grab, but you're already paying Visa, MasterCard, all these credit cards, a whole bunch of tolls. So it is very much like, if I'm my own product, I'm going to be saving money on all of the tools that are being charged.

Shiyan Koh: (18:12)

Yeah, like Square, right? You're writing a basic closed-loop ecosystem between all the people who are part of your network and you don't have to use Visa, or MasterCard rails. Yeah. Makes sense.

Jeremy Au: (18:24)

Yeah. And I think that's an interesting thing. I mean, visa, MasterCard continues to record like what.

Shiyan Koh: (18:29)

Record profits.

Jeremy Au: (18:30)

Record profits. because I was thinking that every FinTech is basically a slight one we hate Visa, and MasterCard, right? That's effectively it. And then, the end slide is we will become Visa, MasterCard. So I'm just curious how you're thinking about that from your perspective.

Shiyan Koh: (18:44)

I mean, I own Visa, a MasterCard stock. I think they're great businesses. What's the question? Are they great businesses? Yes. Would it be great to have your own Visa, MasterCard? Sure. Is it really hard to do? Yes. Yeah. I think trying to figure out ways to do it in vertical slices where you can own that loop is important.

Jeremy Au: (19:04)


Shiyan Koh: (19:06)

But trying to take them head-on, I think is generally pretty foolish.

Jeremy Au: (19:09)

Yeah, I think what's the definition of success, actually? And because when people say neobank, of course, I think people are like, define success as you have to be as big as DBS, you have to be as big as BNI. I mean, obviously, that's one definition of success, but when you think about these banks, what do you think success looks like? Maybe is it from a VC perspective, an operator perspective? Is it okay just to own the loop, for example, in this one vertical?

Shiyan Koh: (19:34)

I mean, I think you need to have a strong return on equity. If I'm investing a hundred million dollars into this bank, at some point, I need to have sort of a path to seeing how I'm throwing off free cash flow. If not at the individual level, I definitely need to understand it at the group level.

But yeah, why should I do a bank over any other thing that I could invest in my business? Like I have to believe that there is a pass to a viable business or some strategic blocking moat here. Otherwise, it's not right. I'm standing chart. Why do I need to put a hundred million dollars into a digital bank versus a hundred million dollars in any other part of my bank? I have to believe, right? Either I'm accessing customers I would be able to do, or the work that I do on the digital bank actually is going to lower costs in my main bank because, and it's a sort of like a scoped way for me to try stuff out with a population that isn't as wedded to some existing set of ways of doing things.

Jeremy Au: (20:32)


Shiyan Koh: (20:32)

But I opened a bank account recently and I was shocked that I had to sign so many pieces of paper. Today when you can open a TransferWise account just with SingPass, why doesn't every bank do that? And I asked the person and she was like, oh, my bank is very conservative, and like, oh, it only works for Singaporeans and there are many non-Singaporeans opening accounts with us, or something like that. And it is shocking to me that in this day and age and when it's so easy with the validation and you get my tax return, you get my validation on my income. You get all these things and all you do is just click through, and actually, you wanted me, you preferred for me to sign like 10 pieces of paper.

Jeremy Au: (21:14)

Not as if Wise isn't a good spot either. I'm currently stuck in customer service hell with my TransferWise account. It's just like you have an email, then you're like, push me to a call after a bunch of times, the call, like, you got to call another call at a certain time, and I'm just in that circular loop.

Shiyan Koh: (21:31)

Oh no. I'm so sorry. I've always had good experiences with Wise. It's fast.

Jeremy Au: (21:37)

Well, Wise is customer service, if you're hearing this, help. This is like waving a white flag here, but I think there's an interesting dynamic. Obviously, so I guess we're broadening this out to a little bit like the FinTech side. Obviously, we're seeing banks, but still just the function of the bank, also a function of the products that they have. Think one interesting way that I've seen it play out now is that instead of trying to underwrite, for example, the whole bank and you're trying to do, like the savings and a checking account, I think people are trying to move into some of that banking by going to financial products like mortgages.

So trying to deploy capital via mortgages or various forms of property financing. Because the thesis, it was again, like making a slide deck would be like, Hey, the biggest chunk of financing is, for example, buying property, and that's a secured loan. So you are relatively more stable in terms of segments. They have clear requirements. There are good go-to-market channels like developers and other agents who will be there. So it's a little bit clearer. So I think property financing, rent to own, but at some levels about property technology, but other levels also about, financial tech underwriting. Yeah.

Shiyan Koh: (22:43)

Yeah, I think this is like, Hey, don't try to change people's behavior because it's really expensive. Instead, find out what people are already doing, buying houses, renting houses, whatever it is. Make that easier, finance that. So I think in the US, Divvy is like a rent-to-own platform that has grown pretty big. Even I think it's a really interesting business, essentially. It's a home equity line of credit in the form of a credit card, but they do instant underwriting online. So home equity line of credit is not a new product, but if you went to your bank and you're like, Hey, I want to get a home equity line of credit, it would take like multiple weeks and a lot of paperwork to get it going.

And you go to Ava, you go to a website, you like, type in your address, fill out a bunch of stuff, you can get underwritten in 15 minutes, and then you have really cheap financing because it's collateralized and securitized against your house.

Jeremy Au: (23:35)

Yeah. Yeah.

Shiyan Koh: (23:36)

This is huge for people. And so, I think there's this sort of, why try to reinvent the wheel when people already seek credit or in very specific ways, can you just make the whole process less painful?

Jeremy Au: (23:49)

Right. I think there's something to be said about speed, and so I think the trade-off is really about accuracy right after that. Because what you want to do is you want to underwrite accurately and quickly, because then, you know you have the best customers, but then you don't have that pain. But obviously, I'm not saying traditional banks are understanding, they can speed it up and just be just as accurate. But I think that's where a lot of the stumbles can be from FinTech Ops is that working to be fast, obviously to get that customer in but there's a big underwriting experiment that happens, which is about, it's easy to lend up money, it's hard to get it back, right? And so that's the hard part.

Shiyan Koh: (24:26)


Jeremy Au: (24:26)

Interest margin.

Shiyan Koh: (24:26)

Just because you lent a lot of money doesn't mean you're a product market fit. It just means you're giving people something really cheap or free and they're loving it. So yes, I agree with that. Fraud will always be an issue at a big enough scale. The fraudsters are very clever and they will find you so it's for sure it is not an easy proposition.

Jeremy Au: (24:47)

And this is circular back to the banks again, but if you're a fintech and you're doing a lot of lending, and you're talking about lending companies, then it goes back to your cost of capital, and then you can get the cost of capital all the way from equity, from raising from VCs. You have venture debt, then you have deadlines, and then you know all the way down to wherever it is. But still, the cost of capital is a huge part of your net interest margin. So your ability to land a lot, but also your ability to land accurately so that you don't have fraud and non-payment.

Of course, there's the macroeconomic shocks, like a recession, for example, like the impact debt, and then of course your cost of capital. And then suddenly, you're back to this piece. Now suddenly even you're lending and you're like, oh no., The best cost of capital was to set up that checking and savings account in the first place, right? So I think there's a bit of strategic inversion or loop de loop here, which is like all roads will be back to the lowest cost of capital.

Shiyan Koh: (25:35)

Yeah, I mean, lending is not complicated. There are three variables. Lending is cost capital, cost of acquisition, and quality of underwriting. Those are your three variables and unless you are a bank, there's not much you could do about the cost of capital,

Jeremy Au: (25:50)


Shiyan Koh: (25:50)

You don't have a lot of control. CAC is generally where most people die.

(25:53) Jeremy Au:


Shiyan Koh: (25:54)

And then the quality of your underwriting, right? What proprietary data do you have that makes you better able to underwrite than anybody else?

Jeremy Au: (26:02)


(26:05) Shiyan Koh:

That's it. There's nothing really rocket science about it. You just have to answer those three questions.

Jeremy Au: (26:09)

Yeah. It's just rocket science, so easy. Just got to do it, huh?

Shiyan Koh: (26:16)

Yeah. But I think that's what happens. People are always like, oh, I have this, like Social media data, I have all this and you're like, okay, but by how much does it improve your underwriting? How do we know?

Jeremy Au: (26:27)

Yeah, Yeah.

Shiyan Koh: (26:28)

Because at the end of the day, there's hard factors and soft factors. If somebody has a working capital problem, it doesn't matter that the sound of their voice is very trustworthy, or you have the names of five of their friends that you can call and harass. At the end of the day, they don't have the money.

Jeremy Au: (26:43)

Right. Right.

Shiyan Koh: (26:44)

So willingness to pay and ability to pay, right? Those are the hard and soft factors. So it's not an unknown thing.

Jeremy Au: (26:53)

How does someone have, "bring in the right executive or talent to understand all of this"? So I think I met a few finTech founders, but they're probably more focused on the go-to market, the sales, but they don't really understand the finance, why it's the sale, right? The ability to model out and show what we have to do in terms of experiments or what the end state net interest margin will look like for this category. So how do you think about where to find them? Are they at a bank? Are they at another FinTech company? Where are they?

Shiyan Koh: (27:24)

There are lots of people with credit experience, but I would say, I don't think you should start at a FinTech company that involves lending unless you've at least built a simple bottle to prove to yourself that it's possible.

Jeremy Au: (27:33)


Shiyan Koh: (27:33)

You could just be like, okay, here's my cost of capital. They could put up 10%, right? Here's what my average loan value is going to be. Here's what I think I can charge for it. Here's what my CAC is going to be, and here's what percent I think I'm going to have for NPLs, non-performing loans. And you could build a very simple model that does not take crazy math to do that. And you can just be like, does this make sense?

Jeremy Au: (27:58)


Shiyan Koh: (27:59)

And you can benchmark it against other fintechs.

Jeremy Au: (28:03)


Shiyan Koh: (28:04)

And be like, well, I'm probably not going to be industry-leading on day one. It's going to take me time to get up to that. So how much capital is going to be required for me to get to that level? I think at least, you don't need to know for sure, but you need to have a theory of how you think it's going to go.

Jeremy Au: (28:19)

Right. Any parting words of advice for all the FinTech founders out there? I think I was just reading halfway. I didn't get to finish it. Jeff Bussgang at Flybridge just wrote a thesis on FinTech first, and SaaS second. So it was like saying, Hey, vertical SaaS companies, then they often add a financing module later, but now he's saying that other way around, which is, can you start out with lending first and then you build the SaaS modules afterward?

So I thought it was an interesting thesis. I think we've definitely seen that be attempted in Southeast Asia and also seen that fail in Southeast Asia as well. But I'm just curious if you have any advice that you have or thoughts here.

Shiyan Koh: (29:01)

Yeah, I mean I think you've seen the FinTech for SaaS later in Southeast Asia because in general, it's hard to sell SaaS in Southeast Asia. hard sell to selling.

Jeremy Au: (29:09)

Well, America just discovered it recently, so because they have so many banks.

Shiyan Koh: (29:12)

It's just that in general, American banks aren't good at building software. So I think that's a different challenge.

Jeremy Au: (29:17)


Shiyan Koh: (29:18)

No, I mean, FinTech, I don't know if I'd have very different advice for FinTech founders than I would for any other founder other than like, I think, okay. I have one piece of advice for tech founders. I think sometimes with FinTech, people overindex on the math of why someone should do something. Like, oh, it's so obvious, I'm a cheaper way to do this, and they actually forget about the psychology of someone buying that financial product.

Jeremy Au: (29:50)


Shiyan Koh: (29:54)

Most financial products don't get bought very often, if you think about it, it's not a daily decision you're making. And there are a lot of considerations that go into purchasing a financial product. And they're not all math, and most people actually are emotion-driven. They're not math driven. And so sometimes I think people forget about that.

And so when you encounter these products, they're like, why isn't this thing converting? Don't they get it? It's such a better product and they forget to talk to their customer to be like, Hey, when you are buying a mortgage or an insurance policy or whatever, it's like, what are you thinking about? What are your considerations? And then addressing that through their product experience. They go straight to the numbers and the math, and a lot of people don't enjoy that and it doesn't resonate. They don't engage with it well.

Jeremy Au: (30:42)

On my end, I think my piece of advice would be what you said earlier, which is you really got to understand the model, what it lands up in terms of net interest margin by making sure you understand your cost of capital, your cost acquisition, and your quality of underwriting.

I think people tend to focus a lot on the cost of acquisition because that seems to be a more straightforward go-to-market. Dozens of people want to buy and sell and I think a lot of FinTech aspiring founders end up getting very shocked by the cost of capital, especially since there's a lot of change over the past three years. So a lot of models that seem to work, this doesn't work at different levels of cost of capital.

But also I think the huge amount of humility around the accuracy of the underwriting is going to be super duper key because that takes time. Like when you build a company in your deck, you're going to know your cost of capital pretty fast. You just got to walk into a bunch of rooms and then you walk out and you're like, shake your head at what you have and you're like, ah, shit. And then, your cost acquisition is quite doable because you do your Facebook test, you put a product like a fit test, and maybe you understand the vertical pretty well. So yeah, those two pieces, and then your quality of underwriting, you can make a certain set of assumptions about how good your underwriting is versus standards, but you don't actually know.

And so I think you've got to have that final model at the end that ties it all together. Just like you said, it tells you, green light versus yellow light versus red light. And I think where I've seen some fintech founders fail is like, they just didn't really think about how each of these factors could change over time. That's one.

But I also kind of avoided the model's answer at the end of the day about, Hey, what does this mean? I have to do it aggressively or change, right? So yeah.

Shiyan Koh: (32:14)

I mean, humility, I would recommend to everyone. And living in reality, as the model's answers suggest this, living in reality. Super useful. So yes, I fully endorse those recommendations, Jeremy.

Jeremy Au: (32:27)

Same. I fully endorse yours as well. On that note, see you next week. Shiyan.

Shiyan Koh: (32:32)

All right. Take it easy.