To lend or not to lend . . .

 . . . that is the question for digital challenger banks.

If you look at the revenue breakdown of a traditional retail or commercial bank, 80%+ of the revenue comes from the net interest margin (NIM) on loans and deposits. And in Europe with the NIM on deposits being close to zero, this means that the vast majority of revenue comes from lending (particularly if you include fees from overdrafts). Digital neo banks are doing a decent job of acquiring new customers, but they are not doing a good job of extracting revenues from these customers. And that revenue potential is going to remain severely limited unless they find a way to break into lending. And it is possible they never will.

A system of cross-subsidies

In the UK it cost banks in the region of £200 to operate a current account which means that the vast majority of customers are loss making. The cost is even higher for those customers who are heavy users of cheques and ATMs. Broadly speaking, it is the customers that borrow through loans, credit cards, mortgages, overdrafts and unsecured loans which are profitable and which cross subsidise the loss making customers who don’t borrow.

Barriers to entry in lending

But building a profitable lending franchise isn’t as easy as it looks. If you make a 2% margin on your lending then it only takes one borrower not to pay you back to wipe out the entire profit on 50 other borrowers. Lending also brings with it a great deal of extra regulation and cost. Prudential regulations look to ensure that your credit losses are not going to wipe through capital and put depositor money at risk. Conduct regulations look to ensure you are not lending to vulnerable people or over charging them for their borrowing. This creates a major compliance and cost burden.

As much as they might tell you otherwise, neo banks generally don’t have the data or relationships to lend profitably in mainstream banking. They might be able to pull some interesting aspects of my data from open banking which will give them access to my income and lifestyle choices. But Barclays, who holds my main banking relationship, knows that when I have borrowed many times during the last 20 years I have always repaid the loan in full. This valuable credit data information isn’t available via open banking. And it is the very customers that the traditional banks turn away because they are bad credits that show up looking for a loan from alternative lenders. Lending is the ultimate market for lemons and the traditional banks already know who the lemons are and don’t lend to them.

Perhaps it’s not surprising then that most of the digital banking start ups have steered clear of lending and focussed on fee-based businesses. But the digital challengers are the ones responsible for driving fees in retail banking down to the point where they are heading for zero.  Most fee-based businesses in retail banking don’t rely on proprietary data so are very easy to clone. So the limited revenue pools currently available to digital banks are going to shrink over time if they fail to break into lending.

Can new lenders survive a full business cycle?

There are of course a few neo banks that have made a foray into lending and point to their recent track record of low credit losses as an indication of their superior credit models. This is extremely misleading however. There are many reasons why an unseasoned credit portfolio can show low or zero losses in the early stages. And when the first default happens as the business cycle turns you often see the floodgates open. I recall a commercial lending fund which prior to the previous crisis had had 0% credit losses for its first 3 years based on its “superior data-driven underwriting models”. By the time the crisis played out more than 20% of their borrowers had defaulted on their debts. That’s enough to wipe out even most well-capitalised bank.

Traditional banks are also susceptible to the credit cycle but because they are on the right side of the market for lemons it is much easier to make accurate predictions about what will happen when the crisis turns. The traditional banks have also built up a war chest of capital since the last crisis and have tightened their credit standards. It is almost inconceivable that we would see a large bank having its capital wiped by credit losses in the next recession. The only curve ball is whether the pro-cyclical nature of the new IFRS9 accounting rules will make the ride a bit more bumpy.

For a start-up bank that has recently gotten into lending I would suggest that they run a material risk of finding themselves with a big hole blown through their capital in the next recession. The greenfield challengers launched by incumbent banks might be an exception to this rule since they have privileged access to the credit data and models of their parent bank.

What about loan brokerage?

There is one other business model which I haven’t covered above whereby a digital broker acts as an Amazon-like market place to bring together borrowers and lenders leading to a brokerage commission for the digital broker. But this brokerage market already exists for things like mortgages and is likely to be another shrinking market in terms of fees and margins. Crucially, the traditional banks that underwrite these brokered loans still get to keep the data on whether the customer repaid the loan or not. So as with the insurance market, the underwriters not the brokers will continue to take the lion’s share of the revenue because they have all the valuable risk data. And this will be true even if it is one of the Big Tech giants taking on this brokerage/marketplace role.

A digital shakeout

This then paints an interesting scenario for the digital landscape after the next recession has run its course. Those digital challengers that decide not to get into lending have a big problem. The fee-based revenue pools they are fishing in are going to shrink and will be threatened by more competition entering due to the low barriers to entry. The private equity and capital markets that are funding these players are then going to be less keen to pour more money into them. And because these players didn’t enter the credit market they won’t collect any valuable loss data during the recession that could have been used to improve their models for the next cycle.

But the ones that have chosen to lend also have a big problem. As I have said, I don’t see many of the standalone challengers that have gotten into lending emerging from the next recession with their capital and reputation in tact.

Good news for the traditional banks?

This all spells good news for the traditional banks who I predict will emerge from the next crisis in a strong position from a capital and hopefully reputational perspective. The benefits from their own efforts to digitise and modernise their own operations should also be coming to the fore at this time and the threat from external digital attackers will have largely passed. As mentioned, the greenfield challenger banks being launched by the incumbents have a data advantage so are better positioned to survive the business cycle than the neo players.

But a word of caution

All of this optimism for traditional banks rests on the premise that these banks maintain an information advantage with regard to credit risk data. It is vital that the traditional banks do not give away their crown jewels by allowing fintechs, credit brokers and credit agencies to get their hands on the bank’s proprietary credit risk databases. These should be kept under lock and key in the bank’s vault.

Most textbooks define the main purpose of the banking system as the intermediation and mutualisation of credit risk between depositors and borrowers. Traditional banks are well positioned to dominate this area of banking. Luckily that’s the largest revenue pool in banking which also looks to be safe from the threat of digital attackers.