> > Bank do refuse to lend to large segments of the population which can be lent to profitably.

> Why do they refuse to lend if it is profitable?

Good question.

Note I didn't say it is profitable for the banks to lend to those segments. I said those segments can be lent to profitably.

In order to profitably lend to customers who banks turn away (mostly those with a 'bad' credit history based on credit bureau data, or those with little/no data at the credit bureau), you need:

1) Effective credit underwriting (i.e. accurate models/processes to decide whether to accept a particular application, and how much to lend to that applicant at this time), and

2) Efficient processes of acquiring, onboarding and servicing customers.

When a bank lends 10k at 8% for 3 years to a low risk customer, they'll make over 1k in interest over the life of the loan. So they can afford to have inefficient operations/systems.

When you lend 500 to a customer who is likely to pay back, e.g. 600 over the next 6 months, there's only 100 in risk-adjusted margin there, to cover all your costs of underwriting the loan, following up on missed payments etc.

I feel like the Market Segment is so small, and your costs will be so high, and it requires so much expertise and oversight/monitoring that it won't be worth it.

"Worth it" is subjective; it doesn't necessarily have to be about turning a big profit. In fact if it's a nonprofit, then the goal could be to just help under-served communities, and it seems like it could have a positive enough impact to attract some angel investors

For the same reason Google shutters profitable projects, or why oil drilling companies often pack up but sell the rights to someone else to pull out the dregs, or why Annhauser doesn't make high quality craft brews.

They aren't profitable enough for an organization of that size.

Markets get segmented into tiers. With banks that might create a situation where the easiest lending targets get processed by the highest volume, lowest cost institutions, and the small valuable cases become boutique.

The problem with that is, unlike other boutique areas, the wealthier person can spend a little more for a limited run high quality good. But here the boutique has to sell to the poorest groups, so it's hard to compensate the provider through pricing for the lower efficiencies of being smaller. It might shut down services that the industry would provide if less consolidated.

I'm not defending this, just presenting one possible way this could be playing out even with everyone acting more or less rationally.

In the US, a major portion of it is that banks are limited in the data they use. For example, certain lenders highly restricted in how they can use zip code in any automatic algorithm. So they stick to manual processes (which are only profitable on large loan sizes) or very simple automated algorithms.

Non-bank smaller lenders often have many fewer regulations, allowing them to make more targeted automated algorithms, which means they can serve people banks can't.

Are there documented cases of non-bank lenders redlining?

Because of risk. Just because there's 10k people out there that can afford to borrow and repay, doesn't mean you set 100k loans hoping you got some of them.

That's the point the parent is making. If the risk profile is as you say, that isn't a profitable segment for the banks. The question being asked is: why would the risk profile be any different for peer to peer lenders than it is for the banks? If it isn't profitable for the banks, and the risk profile isn't different for the peer to peer lenders, then it won't be profitable for them either.

The profit made by the bank needs to also cover the cost of employees, bureaucracy, huge buildings, etc. For a peer-to-peer lender, even tiny profit is enough.

As an example, imagine that you lend someone $100, and there is a 50% chance they will pay you back, and in that case they will pay you $210. And all this costs you is a mouse click. Some individual would click "yes". In their free time.

Now in the bank, there is an employee doing this as a part of their paid job. They have a manager, that manager also has a manager, plus you need to pay the janitor, etc. You also need to pay diversity training for all of them. And all financial transactions they do must follow all kinds of regulations, which regularly change. Simply, the overhead is not worth it.

$100 > $210 is an ridiculously high interest rate. In a world today where borrowing cost for a lot of people are ~10% APR.

But I do see mobile payment and mobile borrowing could be mounted together.

See points above. Banks, being an integral part of the financial system, are only allowed to take on so much risk. Third party companies have no such restrictions. But they can’t issue loans because they are not a bank. So what happens is a bank issues a loan according to criteria the bank and the third party agree to. Then the bank sells it the next day to the third party. That way the bank is more insulated from risk because they aren’t holding the loan.

> they can’t issue loans because they are not a bank

I have to think there is some misunderstanding here - lots of nonbank companies issue loans. Ford Motor Credit is not a bank.

The thing that banks are uniquely allowed to do is to take deposits, and lend out that money.

Most places cap interest at 25-35%, that’s too low for high default customers.

Also, they need to "predict" whether the loan would be profitable as they have not loaned the money yet. The decision largely depends on the precision of their prediction

Banks are limited in the total amount they can by regulatory controls to prevent uncontrolled expansion of the money supply.