Starting a Bank

http://ma.tt/2009/08/starting-a-bank/

by ionfish • 17 years ago

144 65 17 years ago

sharpn2• 17 years ago

There are a whole bunch of reasons why this would be nearly impossible, but it's an interesting thought experiment nonetheless.

stse2• 17 years ago

I don't like this comment. At least give me a hint why it would be "nearly impossible". You won't be able to do it with three guys in a basement, but I can't see how it would be "more impossible" than any other large corporation. If there hasn't been some recent change in regulation, it is probably a lot easier now then 10 years ago.

sharpn2• 17 years ago

It's harder than many other large corporations because banking laws vary from state to state, and there are federal laws restricting banks operating universally across several states. Also, as a 'startup bank' you are initially reliant on your incumbent rivals for clearing and interbank transactions (unless you do literally leave cash in a vault at 0%). I agree technology has made it much easier recently though.

jakarta• 17 years ago

It's possible to start small community banks and there are programs for it. Difficult, but not impossible.

But about his bank. It seems like he wont make loans to local businesses and instead choose what he deems to be the least risky financial instruments. So the interest earned on deposits will be low.

And then as for safety, we all know that deposits in FDIC backed banks are good up to a certain limit. So he's not going to do much better for the customer on that end.

Plus, I don't know why he's complaining about bank bonuses. I believe he's mixing up investment banks with your usual commercial/community banks. As long as incentives are aligned to be towards loan quality rather than quantity, you will get good loans.

kiba• 17 years ago

This will probably not happen when there are too much regulatory barrier to start and too many rules to comply with.

The more improtant something is, the more likely it will be highly regulated, or monopolized.

m_eiman• 17 years ago

That just means that there's less competition. And when you start getting successful, your competitors will need to spend a few years implementing their system before they can start actually competing with you. And I wouldn't worry about the big banks catching on to the "do what you can to help your customer" concept, it's not in their culture.

cookiecaper3• 17 years ago

It also means that unless you're starting out with at least $xx million dollars, you have no hope. It can take _years_ to get a business started if you're unlucky enough to choose a field under heavy regulation (something with it's own Cabinent-level bureau, like the FCC, EPA, or Fed).

The paperwork takes a long, long time to process, the processes are laden with significant fees, you have to pay staff (usually lawyers, not cheap) independently to prepare all of the forms and offer guidance through the process, and you have to be able to afford to wait for everything to go through, and that assumes a completely clean process with no need to contest, appeal, or resubmit anything.

The megacorps already initiated to the oligopolies hope that this is enough to prevent any significant or innovative new competition, and it usually is.

Sadly, those with the most innovative ideas are usually not those with the most $xx millions of dollars.

gwern• 17 years ago

Sounds like he wants to start a full-reserve bank: https://secure.wikimedia.org/wikipedia/en/wiki/Full-reserve_...

Not impossible - after all, banks were not always fractional-reserve, and it admittedly is a very secure form of banking.

tdoggette• 17 years ago

Not full, just a higher fraction.

mseebach• 17 years ago

Since he wants to keep accounts below the FDIC limit, there's really no incentive for him not the hit the markets aggressively, trying to make a nice interest for his customers.

Tichy2• 17 years ago

"We take your money and put it in a vault"

That's not actually the job of a bank, is it? The job is to mediate lending and borrowing of money. By identifying the right people to lend money to, banks can offer the people whose money the take an interest rate.

I still agree that there is a market for a better bank.

I remember an example from the book "Why not?" where a bank (or something like that) offered to automatically adapt their customers to the best current interest rate and making a killing with it. No more "if you open an account now, you get 2%" and one month later "if you open an account now you get 4%" and two months later "if you open an account now, you get only 1%".

steveplace2• 17 years ago

Well why not differentiate the product by allowing the bank customers choose where the money is getting invested, kiva-style. And yes, there is a ton of risk there but it's unique.

Tichy2• 17 years ago

That seems counterproductive. I give my money to the bank precisely because I don't want to bother with estimating borrower credibility myself. Also, as an individual I can not hedge my risk well enough, because I don't have enough money to spread it out effectively.

losvedir2• 17 years ago

Not a "bank", per se, but exactly what you're talking about: http://www.lendingclub.com/

ewjordan• 17 years ago

"That's not actually the job of a bank, is it?" (re: money -> vault)

Well, from a customer point of view, it is, more or less. Most people are not picking a bank because of the interest rate it offers, they're picking it because they need to be able to write checks, use a debit card, make payments, etc. Anyone that's serious about growing their money is not opening a Bank of America savings account, they're using a real brokerage of some sort.

Am I wrong here? I tend to keep money far away from my bank account unless I intend to use it fairly soon, because I can do more with it elsewhere, but maybe I'm abnormal in that respect?

Personally, I'd guess that the single most important factor in choosing a bank is simply which one happens to be closer to your house so that you can make deposits and use the damn ATM without getting charged a fee. Mixed in with a little bit of hatred for banks that have screwed you over in the past, perhaps.

Unfortunately it's real difficult to compete on location until you're huge, so I don't know if something like this could ever gain much traction, especially as a startup without some serious cash invested up-front. I don't know if good web services could really close that gap; you feel the sting of a $3 ATM fee every time you get money, and making deposits is a lot easier if you don't have to send something through the mail, and I don't think a great website is going to lessen either pain point one bit.

shimon• 17 years ago

There are many "internet-only" bank accounts in the US that reimburse you for ATM fees. I use http://www.salemfivedirect.com/ , which reimburses up to $15 of ATM fees per month, and Etrade bank does something similar. Being free to totally ignore ATM fees is really nice, and because internet banks are competing for customers that are likely more mobile and savvy than local branch banks, odds are the customer service is better.

It is also possible to deposit checks by scanning them and uploading the check image; one of my friends does that for her personal account.

frig2• 17 years ago

Yeah we live in interesting times in which the historical role of banks and what banks actually are used for are diverging.

There's two kinds of things here (and I'd lump them separately):

(1) there's "putting money in vaults" and handling long-distance transactions (eg: debit cards, check-clearing, and so on)

(2) there's banks as "so you've got money you want to lend but no time? give your money to us and we'll loan it out, passing some share of the profits back to you" intermediaries

(1) is arguably what most people actually use banks for; (2) is how most consumer banks make their money (in the form of mortgages and cds and so on).

Aside from historical circumstance there's no particular reason why (1) and (2) are handled by the same institutions. Since (1) on its own isn't profitable any institution that does (1) needs to have its (1) activity subsidized; the present situation is that banks subsidize (1) via (2), but there's nothing that intrinsically says institutions that do (2) should do (1) (or that the only way to subsidize (1) is via (2)).

In many countries (Japan, some european countries that escape me now) there's an entire system of "postal banking" which is basically de-facto state-subsidized banking expressly for the purpose of (1).

If you think about it the connection between (1) and the postal system is pretty logical: the postal service exists to streamline communication within the state; it's pretty sensible to piggyback a payments and money transfer system on the existing communications infrastructure (and of course it's much more reliable to transfer money then it is mail an envelope with cash).

The USA never went the route of establishing a postal savings bank (nb: the historical names include "savings" but they always allowed payments / de-facto "checks" to be drawn); this was mainly b/c the idea was invented well after the founding of the republic and the USA is pretty slow on the uptake wrt good ideas invented elsewhere.

One of the takeaways or realizations I see people coming to a lot in response to the current crises is along the following lines:

- at the moment keeping commerce flowing depends on keeping checking and debit and credit-card systems all working smoothly (in other words: keeping the (1) activity going); without that the economy halts

- currently keeping (1) going is dependent upon the health of various private parties (banks + other financial firms); as these entities don't really make any money from (1) directly their ability to perform the services of (1) is contingent on how well they're doing in their other endeavors (activities in category (2), writ large); (1) is not enough

- thus disruption in (2) leads to disruption in (1) leads to disruption in the entire economy, even when (so it seems) there's no direct connection between (2) and (1)

...which leads to a realization that the current arrangement worked until it didn't but it's not that great of a system because it needlessly couples (1) to (2); depending on the kindness (and prosperity) of strangers to keep essential infrastructure running isn't a good plan.

Once people have that reaction they tend to produce one of the following proposals:

(A) set up a system of state-run full-reserve checking+savings accounts + debit cards, etc. (and optionally: phase out the FDIC guarantees; if you want a guarantee go with the full reserve bank); fund operations from taxes

(B) set up a dumb/simple/basic bank that just does the basics and takes almost no risk

I tend to think (A) is the longer-term smarter option but am well aware how against the grain it runs in the USA; (B) is like someone trying to accomplish the same goals but not really understanding how the tools they have at hand work (that's what the original post is like).

The quick case for (A) is that there's a reason the constitution grants the authority to build a post-office: aside from the general betterment of the commonwealth it specifically prevents a situation where your (then-fledgling) government is effectively held hostage by private-sector actors because those private actors own and operate infrastructure essential for the government to function (what good is a federal government if Megamail Inc. won't deliver it's mail?). These days basic savings + checking + long-distance payments are critical infrastructure on par with the post office (and on par with the original grant of authority to coin money and so on), and to avoid being held hostage the right option is to provide a de minimis payment infrastructure.

Good luck with that politically, though.

eru4• 17 years ago

You can find Postbanken in at least Germany and Switzerland. By the way, as far as I know the German Postbank is treated like any other bank and not subsidised. Deutsche Bank recently bought a large chunk of Postbank from Deutsche Post.

steveplace2• 17 years ago

The 3-6-3 model of banking has been thrown off to the wayside in pursuit of faster, quicker gains; but the original model still holds: savings at 3%, loans at 6%, golf course by 3pm.

known• 17 years ago

I got an idea. Take loan from a bank to buy a bank.

trader• 17 years ago

Obviously "matt" has not worked in the finance industry and clearly does not understand how banking works. Sounds decent in theory but in reality the bank would have no investors and would lose any talent when they didnt pay them. They would probably be the cause of the financial crisis by only employing stupid people who gave poor, risky loans.

jonknee• 17 years ago

"Stupid" people couldn't have done any worse than the "smart" "talent" that got us into our current mess.

knightinblue• 17 years ago

Goldman Sachs just paid out 16 billion dollars in bonuses to their employees. If we had an extra 16 billion dollars lying around, we’d put it in the bank for a rainy day. (If Goldman had never paid out bonuses they never would have needed government intervention.)

A flawed comparison. Goldman is an investment bank that needs to pull in incredible returns percentage wise. Safebank is a watchman guarding a vault. Basically, Goldman needs to pay out bonuses to attract top-shelf talent, while Safebank doesn't.

biohacker42• 17 years ago

SafeBank couldn’t raise VC or anything like that...

It also couldn't be a public corporation because share holders could make a very good case that management is failing its fiduciary duty by being hyper conservative.

And that's what I find most interesting about that article. It opens the discussion about public vs privately owned enterprise.

Privately owned business can forgo addition revenue by as much as the ownership feels like. The reasons for that don't have to be altruistic, it could be about longevity, not 20 or 50 years longevity but hundreds of years of longevity.

But we only live once, and our lives are short, so most of us like to maximize profits. Thus most often the risk vs. reward calculation doesn't look at time spans much longer then one human life time.

eru4• 17 years ago

Interestingly public companies should be able to deal with long-term horizons much better than private companies. In theory a share's worth is equal to all future dividends discounted to the present day.

mdasen• 17 years ago

The problem is that his proposals ignore the economics behind banking. If you take someone's money and put it in a vault, how can you provide interest on that money? You can't. You'd, in fact, have to levy fees to cover the costs of staffing, buildings, security, etc. Yes, even an internet bank has to have physical infrastructure somewhere.

Heck, why am I going to spend $3 for a SafeBank iPhone app when I can access my online banking from many banks for free?

He rails against ATM fees. How is he going to get rid of them for me?

I don't notice that any of the large banks require the use of Internet Explorer. BofA, Wells Fargo, Citi all work with non-IE browsers. I'm sure one can find a bank (out of the many thousands in this country) that has a site that doesn't work with Firefox, but it isn't an easy thing to find - which is why he hasn't specified any particular bank! He doesn't know of any!

While ad blocking might be cool, it has nothing to do with my bank. It would be like WalMart telling me that acupuncture was good - it has nothing to do with their business.

People might hate banks, but they hate people mining them for data potentially more. "We noticed that you spend a lot at pornoshop.com. Did you know cheapporn.com usually has better prices?"

And he hasn't satisfied the most important question: how would this be better than a member-owned credit union? A corporation has to try for profits. A member-owned credit union gives those profits back to its members as dividends. Their rules are set up for the benefit of their owners - who are the people depositing money with them.

The problem is that people want a host of expensive things. They want tellers. They want ATMs everywhere - and that means cash physically everywhere or you refunding the ATM fees. They want free checks. They want online banking.

While one can argue that the banking industry got a little out of control, you can't just put money in a safe and expect to be able to offer people all that. You need to lend it out and that incurs risk and even more overhead - you have to deal with all the payments, the legal stuff when something bad happens, foreclosures (because, yes, some people won't end up paying you back). And if you want to be "safe" and only lend to the most credit-worthy people, your margins go down because they're the people who can get money anywhere at really good rates.

I mean, if you lend me money at 4.83% (the overnight average for a 15 year fixed mortgage at bankrate.com) and give your depositors 2% interest on their money, you have a 2.83% margin to operate on. But then you want to refund people's ATM fees of, let's say, $10/mo. On $5,000 (more than the average depositor has), that means you have an effective 3% rate on deposits. If the depositor only has $1,000 deposited, you've given them an effective over 10% rate and you're now way into the red. And that doesn't even take into consideration the around 30% reserve you want to hold earning nothing. And you'd want to give them free checks and not charge overage fees and do all sorts of nice things.

And if you're a "nice" bank, you'll want to provide lower mortgage rates than the average, right?

The fact is that there are already "nice" banks. They're called credit unions. They're owned and controlled by their members (customers). How will you do better than a credit union that's a not-for-profit institution?

By being more efficient? That's the only way. And he's vastly over-estimated the money that rolls in when you play it "safe" and "nice". Banks make a lot of money. They make that money through a combination of managed risk and fees. If you get rid of both of those, you don't make money, you lose it.

jobu• 17 years ago

His head is in the clouds. Although, the idea of starting a bank in the middle of a downturn is not bad. People are saving more money than usual right now (cheap capital), and many businesses are desperate for loans. The trick is to weed out the failing businesses, and loan to the ones strong enough to last the downturn. As always, easier said than done...

If anyone out there is looking for a "nice" bank, you're much better off with a credit union. Credit Unions are by definition customer focused.

Eliezer• 17 years ago

Actually, I don't think he understands how banks work, period. They loan the money out. That is where the money comes from. How much money a bank makes is about what sort of loans it makes. Guy talks about "keeping the money in a vault" so he has, roughly speaking, no clue whatsoever how banking works.

webwright• 17 years ago

"The problem is that his proposals ignore the economics behind banking. If you take someone's money and put it in a vault, how can you provide interest on that money?"

Except that he said this:

SafeBank would maintain a reserve level 2-3x higher than Fed requirements and any other bank.

What he's saying is this. SafeBank would be less risky and could spend a LOT less money. Revenue would be lower, but he's saying that spending at financial institutions is nigh ridiculous and could more than compensate for it.

"And he hasn't satisfied the most important question: how would this be better than a member-owned credit union?"

How is that the more important question? What if it was just as good as a credit union?

How much better is Starbucks? I think 80% of what he's talking about is a marketing opportunity.

cookiecaper3• 17 years ago

>SafeBank would maintain a reserve level 2-3x higher than Fed requirements and any other bank.

Which, for the record, isn't full-reserve (the legal limit is 900% leveraging... i.e., if I deposit $10, the bank can loan $90 off of that). He'd still be loaning out money he doesn't have, and he'd still be able to generate revenue from that.

In the case of full-reserve banking, you can still make money. You may not be able to offer FREE CHECKING and free x x x x blah blah blah, but with the interest accumulated from your loans, a decent amount of working capital, and fees from accounts it's definitely conceivable.

ig12• 17 years ago

It wouldn't even cover your costs. Most bank current accounts are run at a loss for the bank, and just operate as a loss-leader for other products the banks sell. A current account costs in the region of a couple of hundred dollars a year to run (see countries which charge for current account services).

Assuming the bank is making an (adjusted) return of 2% on your money, the account has to have 10k in it even to break even. Very few current accounts do.

pelle• 17 years ago

Everything you say is true for traditional full service banks. I believe however that we need to drop the idea of the one stop full service bank.

Where real innovation and safety comes is through the idea of narrow limited purpose banks. There is a lot of space to innovate if you as a bank can focus on one product only.

The simple way of doing this is to imagine every account type a bank has as a separate mutual fund.

The checking account would be replaced by a 100% reserve 0 interest cash mutual fund, which makes its money from float, membership fees, transaction fees etc. Most checking accounts now are paying an insignificant amount of interest anyway.

Savings accounts can be offered by various levels of money market and similar funds.

If all of these funds follow the same simple standards it is easy to transfer between them, even if they are run by different companies. We are working on creating these new standards on the Agile Banking list:

http://groups.google.com/group/agile-banking

One of these standards we are working on is a dead simple OAuth/REST based transaction API that we're calling OpenTransact. It allows for very simple transfers, but allows more complicated financial transactions to be built on top of it.

http://wiki.github.com/opentransact/opentransact/opentransac...

What about ATM fees, credit cards etc. These should be run by independent companies. On the Agile Banking list we are talking about independent card issuers. These are independent companies that offer regular payment cards that in essence control OAuth Access Tokens to an OpenTransact based cash mutual fund as I described above.

As a card issuers only access to your account is via OAuth, you as a consumer can easily manage limits and even revoke the Access Token. You are in control.

Business model for these card issuers? Advertising on cards, transaction fees, membership fees etc.

There is plenty of money to be made in this business, we don't need to keep thinking about it in the frame of how it has been done until today.

fatdog7892• 17 years ago

They do have an existing, simple standard to transfer money. It's called ACH, and it's been around for decades. It works quite well.

ATM fees should not be run by separate companies -- then you'd have a company that would need to profit off ATM transactions, instead of using them as a loss leader, which means higher fees and more of them.

Credit cards can't be run by non-banks, b/c they are loans, and state and federal law limits the institutions that can issue variable-size loans.

Furthermore, the OAuth system would not work. Period. A customer being able to cut off the card issuer's access to their account? You would never get any card issuer to agree to use the system, because of the enormous potential for fraud (run up the bill, cancel the issuer's access before payment). BTW, you can already manage your card's limits -- just use the online portal, or call your company.

Finally, advertising is an overrated business model. Believe it or not, at some point, you need to make money off an actual valuable service or product. As mentioned above, transaction fees would need to be jacked up (for the institution to make money), eliminating any reduced costs benefit to consumers.

senthil_rajasek• 17 years ago

"People might hate banks, but they hate people mining them for data potentially more."

Matt clearly mentions that selling accurate consumer data as one of the ways to generate revenue.

It's not any worse than the current situation. Aren't banks already mining consumer data?

Safebank could possibly be a more ethical alternative.

BTW, credit unions charge fees like overdraft fees, wire transfer fees etc., only they charge a slightly lower price than banks.

va_coder4• 17 years ago

This is a very interesting topic, but I feel he really doesn't understand banking for a number of reasons and this is one of them.

The banks mine tons of data and use it to make profit off of you and I in more ways than you could imagine.

I find finance and technology fascinating, but the idea of a good citizen bank is very wishful thinking.

dmhouse• 17 years ago

Interestingly, nearly all ATMs (with the exception of privately-owned ATMs in bars, etc.) in the UK don't charge for services (for about five years now). However, I wonder whether an ATM owned by bank A charges bank B itself when bank B's customers withdraw cash at bank A's ATMs.

va_coder4• 17 years ago

He also failed to address fractional reserve banking. Many of the big banks, which are shareholders of the Federal Reserve, are able to loan out money they really don't have due to fractional reserve banking.

Also, the Federal Reserve is neither a federal agency nor does it have any reserves. It is owned and operated by the banks, which are able to game the system and this gives them an advantage over you and I.

jdminhbg• 17 years ago

The problem with the "Bank Safe" slogan is that the FDIC exists.

Dilpil• 17 years ago

This. No one lost money by depositing their money with goldman sachs (they were not a deposit accepting institution). People lost their money speculating on stock and real estate. Having their money in a slightly safer bank would not have helped.

randallsquared2• 17 years ago

It's marketing. That's all.

cookiecaper3• 17 years ago

Except that there've been rumors swirling regarding the FDIC's solvency for a while now. People no longer confident in the government's bailout ability would be comforted to hear there's an option where much more of their money will be physically retained.

Not that the government will let the FDIC fail. They'll just inject new money endlessly to prevent it.

amkimian• 17 years ago

Remember that the FDIC is not funded by the government - it is funded by premiums that banks pay to get their "insurance". In this model SafeBank would have to pay a premium to the FDIC to get that insurance and as many small banks have found recently (when the premiums went up) those premiums are not insignificant. That's not to say that the government would not bail out the FDIC but to date there has been no reason to do so as they are completely funded by their clients...

randomtask• 17 years ago

His idea is ridiculous. If I were to put money in a bank like this it would cost me, because I'm not getting the interest another bank would pay me and inflation would devalue my balance every day it sits there. That's not "safe" for the customer at all.

jacoblyles• 17 years ago

People don't care about the safety of their bank because of the FDIC, but they do care about interest rates. Rival banks will be able to offer higher returns because they will be earning higher, riskier investment returns.

Moreover, with a lower return on equity modeled in than every other bank in the industry, nobody is going to give you money to start with.

Lastly, there are severe limits to how innovative you can be with a bank, again because of the FDIC.

The only way I see this gaining any traction is as a niche bank for people with deposits over the FDIC insured limit who might want a little more safety. However, there are two problems with this. 1) He proposes to turn away deposits of this size at first and 2) people with that much money usually have better options.

bonsaitree• 17 years ago

This reminds me of the wonderful theoretical country discussions we had in my economics classes at college.

Forget a new kind of bank. Go total bluesky. What IF, we could design our own set of sovereign laws and principles.

What would the "modern" country look like?

randallsquared2• 17 years ago

http://seasteading.org/mission/intro ;)

Perceval• 17 years ago

It would separate two forms of money: the common form of money used as means of exchange and a store of value; and, the debt/credit money created out of thin air through a contract and collateral. The two currencies would float against one another. People would be paid in common currency, but could borrow in the debt currency.

hamidp• 17 years ago

The whole point of Safebank seems to be to have it be safe. How safe a is a week-old bank started by some guy with a blog who doesn't know the difference between retail and investment banking?

People don't love their banks? I do. Bank of America gives me ATMs everywhere and pretty awesome online banking, and the credit union gives me unbelievable customer service and great interest rates.

On a sidenote, do compare the HN comments with those on the site.

rgr23• 17 years ago

The problem with his reasoning is that the big banks are evidently quite safe. No matter how bad their investments are, the government will bail them out with taxpayer's money to prevent them from collapsing. I doubt the same would apply to a bank ran out of somebody's basement.

lamby• 17 years ago

Pretty much all of the points I was going to raise have been covered - I just wanted to add my facepalm at the people in the comments who clearly don't understand a hypothetical scenario.

djahng• 17 years ago

"'Bank of America spent $40,000,000 dollars on airplanes last year. We spent $40,000 to develop an iPhone application so you can check your balance from anywhere.' (Hmm, the iPhone app should cost like $2.99.)"

I think he's misunderstanding something...the iPhone app may cost $2.99, but it doesn't cost $2.99 to develop.

alaskamiller• 17 years ago

He's saying he wants to charge the bank app at $2.99 as a way of generating revenue.

va_coder4• 17 years ago

If you were going around selling a business model of a new bank, as soon as you mentioned a 2.99 iphone app you would be shown the door.

losvedir2• 17 years ago

The part that's interesting to me is his idea that it will be essentially a "vault" and super safe. The problem is he'd be targeting people under the FDIC limit, so that safety wouldn't be desirable.

But what if he targeted people OVER 100k. (For that matter, let's say he didn't offer loans at all so he really was a "vault") Sure you couldn't offer interest, and maybe you'd even have to charge a very tiny fee, but are there places that do that? What do people do if they have a lot of money -- put in Treasurys? If I had millions of dollars, I'd pay someone to just deposit it with the Fed. I think the Fed even pays interest now.

nathanwdavis• 17 years ago

This just would not work. The reality of banking is that people want a relationship with their bank that they don't have to maintain - meaning they would only go to their bank or bank website when they absolutely have to. Not to go check out the latest blog post.

Invite only? With the much larger startup cost of a bank this is ridiculous.

What is really amazing is that this guy is actually ... "one of PC World’s Top 50 People on the Web, Inc.com’s 30 under 30, and Business Week’s 25 Most Influential People on the Web"

Starting a Bank

Related Stories