Hacker News Reader

Comment by chadash | Hacker News Reader

Except the parts that don’t scale. For example, some banks specialize in servicing specific communities (e.g., Chinese immigrants). Other banks understand local geographic areas well. For example, they might be able to give you a better rate on financing a gut remodel because they know your contractor does good work and your plans for the remodel will be desirable in a specific neighborhood. It’s hard to do that when you are behemoth national bank.

View on HN

I hear that argument often, but I’m not convinced.

Why should efficient lending decisions somehow be uniquely hard to make in a large corporation with many local branch offices?

We’ve seen massive consolidation in mobile phone networks (where local knowledge should arguably matter just as much for where to build the next tower), supermarkets (I’d assume preferences and spending habits to also vary widely), media, and many other areas of business.

What makes banking so special? And why in the US, but not in most other countries in the world?

And if it’s a question of not having too many “too big to fail” entities: Wouldn’t it then be better to have 20-50 Chases rather than 5 Chases that still concentrate a huge chunk of customers and a long tail of thousands of tiny banks?

>Why should efficient lending decisions somehow be uniquely hard to make in a large corporation with many local branch offices?

Once a bank gets big enough, I think the main issue is risk. The long tail of tiny banks represents a certain customer type that is harder to model in an Excel model and is vetted through personal relationships. Consider one the "big" banks in that long talk - SVB. For a lot of startups getting banking services such ACH, or lending mortgages to founders backed by "paper gains" would be a nightmare for your classically trained big-four bank manager. Conversely, all the work thank bank manager would have to do in winning those deals, likely wouldn't get any benefit from the parent bank except for the bank charter and maybe a small bonus. It's better to be the CEO of a small bank with a highly differentiated order flow, then to fight your exec who lives 1000 miles away that won't let you lend to founder of DropBox because cloud storage isn't a dropdown in their excel spreadsheet.

> Why should efficient lending decisions somehow be uniquely hard to make in a large corporation with many local branch offices?

Easy. When you only exist in for example Manhattan, you don't have to deal with divergence laws against Calif, because you just not in Calif.

But when you have branches in NY and in Calif, laws just prohibited you to make different decisions for NY and Calif - all MUST be SAME.

I think the network issues you mentioned are relatively simple engineering issue, you can easily hire and delegate.

Banks have the issue with delegating power to give you money. From a distance I'd guess there is a real advantage to businesses of local banks and being able to work on personal relationships.

You never have a relationship with a big company, you are just an entry in a database interacting with an algorithm.

Lending decisions might simply, often but not always, be much more complex than how to build nationwide networks.

> Why should efficient lending decisions somehow be uniquely hard to make in a large corporation with many local branch offices?

The trouble is that the decisions can't be entrusted to the branch offices, so you wind up being a little stiff and inflexible and slow.

Reply on news.ycombinator.com