> there’s no reason why a narrow bank couldn’t take deposits, stick them in with the fed providing 4% interest, and keep 25% of that for itself to cover its costs, passing 3% onto the customer.

The problem with this business model is that this 4% interest either (A) requires a certain hold period or (B) allows on-demand withdrawal.

(A) is exactly the model used today. It did not stop the collapse SVB and FRC. Assets that require a certain hold, such as Treasuries held by SVB and mortgages held by FRC, fluctuate in price. If those assets drop in price (as recently) and bank's customers withdraw money, the bank is in a pickle.

(B) moves banking into one real bank -- the government. This can be done (e.g. in the Soviet Union), but comes with a lot of limitations and challenges. It is also not something you can morph the current system into; this is a "break, then rebuild" path and is very painful. I would personally move money away if I see a whiff of this in the air. My 2c.

The 100% reserve bank people will at some point discover why Silvio Gesell imagined the concept of demurrage currencies.

Because at some point the market interest rate is 0% and the cash rate is 0%. There is no reason to lend money via longer term deposits. The amount of deposits available to banks dries up. The circulation of money stagnates and ultimately causes a recession with high unemployment.

The demurrage fee doesn't change the market interest rate, it just moves the guaranteed interest rate into the negative range so that the market interest rate can be negative, if it has to be, instead of being stuck at 0%.

That's exactly what Silicon Valley Bank did. While federal bonds are nearly risk free in terms of default, they still suffer from interest rate risk -- that is, they may lose significant value if interest rates rise. In the case of SVB, US treasury bonds lost enough value to make the bank insolvent.

Yes, but because this poster thought this arbitrage was obvious and would solve all our problems, the political issue of banking regulation remains. It is a Dunning Kruger trap.

Money market accounts use the same vehicle (treasuries) and still manage to be ostensibly safer than smaller banks at the moment. These funds manage their investments by continuously buying new bonds and selling off old thus limiting the risk exposure to all but the most sudden interest hikes.

I think money market accounts don’t sell off old, but rather they hold short dated stuff to maturity. And constantly reset the rate they pay holders.

Of course, they can have counterparty risks. The counterparty is usually unprepared to actually pay up and expects to rollover its debt.

Sometimes that melts down and doesn’t happen:

https://globalnews.ca/news/160176/coventree-executives-faile...

(Scroll down to Canada in 08-09 here):

https://en.m.wikipedia.org/wiki/Asset-backed_commercial_pape...

They also, AFAIK, don’t make loans like mortgages, auto loans, credit card lines, etc. All those loans are subject to the same interest rate risk. They’re not nearly as liquid, either.

> stick them in with the fed providing 4% interest

Where is this risk free return coming from? You are just saying in more words that the government should give the people what they want.

That article says that the Fed chooses interest on reserves to be below the market rate of interest...

You left out “zero risk”

Banks have always been able to fail, then you lose the deposits. What people are now asking for is unlimited insurance, which encourages risk takings amongst banks

> What people are now asking for is unlimited insurance, which encourages risk takings amongst banks

Except that when a bank is liquidated, the shareholders get zeroed out, and the executives get their money clawed back, which encourages banks to not risk-take.

As a depositor, this sounds fine to me. I don't benefit from banks doing stupid, risky things, and I shouldn't suffer from it either. Raise the capital requirements. Zero out the shareholders when things go pear-shaped.

The problem is not what happens to the shareholders (they get zeroed out the same way that shareholders of other failing companies would be), but what happens to the depositors. The rules, as written, say that deposits above $250k can be wiped out. The rules, as the government applied to the SVB, made large depositors whole.

So we have the worst of both worlds: no depositor guarantee de jure (so depositors bailed and brought SVB down), but a de facto bailout which has significant actual costs and encourages risky depositor behavior. My 2c.

I think the priority at the federal level is still in at least maintaining a facade of banking system stability. It’s very rare that the bank not only becomes insolvent but also loses all its value. Risky investments to a bank are not really considered super risky elsewhere and the investments still hold value, albeit less if they are trying to liquidate. The problem is when banks are unable to maintain enough cash which happens when they over-invest or when their returns just barely keep up with inflation. There is usually plenty of money that is recoverable, with time, to make depositors whole. The “bail out” concern is a red herring for the actual problem: market consolidation. Larger banks effectively write IOUs to each other depending on who owns who after each day of ACH, wires, deposits, and withdrawals, while smaller banks pay in cash. There is additional burden on, usually, medium-sized banks who need to both maintain large cash reserves but also cover the cost of operations.

You’re conflating liquidity with solvency. The problem isn’t a lack of cash on hand, it’s assets that have fallen in value. Saying that they money will be there, it just takes a bunch of time, denies the reality of the time value of money. A stream of payments of $10/year for ten years is not worth $100. Never has been.

Shareholders are routinely zero’d out in bank failures. Are you proposing some other mechanism?

No, because the mechanism we currently have seems to be working. SMBs are currently scared shitless that they will be reaping the consequences of their actions. Depositors are not.

We have that, that's called a (federal) money market mutual fund.

Edit: sorry, you said the Fed, which MMMFs don't put money in. But they buy short-term Treasurys at comparable or better yields, which is ultimately what you want, modulo the risk of US government default, which is minimal.

You are exactly describing what all of these banks did.

What happens to the value of 4% bonds if interest rates go to 6%? What about if people want their money back before the bond duration is up, so you have to sell the 4% bonds in a 6% environment?

Note that in this interest rate environment banks don’t need to buy bonds in order to get 4%. Currently the Fed’s reverse repo facility provides more than that.

The issue is it takes years (decades) to stand up a new bank due to the extensive regulations, processes, etc. that need to happen - and who knows what is going to happen next year - let alone in that timeframe.

The reason why we have such consolidation is that merely existing in the Byzantine world of banking regulations is nearly impossible, and only the largest entities can afford the legal, IT, trading/financial mojo, capital, etc. resources necessary for it to occur.

The big 5 have not only such resources, but enough economic impact if they fail that they can force regulators to stop doing a regulation that would kill them.

Which is ultimately where this is all going. Either nationalization (unlikely) or the big banks writing their own regs (in practice).

The answer to your question is prices would go down until the yield on the bonds equals the market rate.

Separate nerd note: I think you mean maturity not duration.

The maturity of a bond is the point in the future when holders get back the principal.

The duration of a bond is pretty much the first derivative of price with respect to rates expressed in years. Intuitively it's the number of years (on average) an investor in that bond would need to remain invested for the future value of cashflows from the bond to be equal to their investment in the bond.

https://www.investopedia.com/terms/m/macaulayduration.asp

That’s not how the Fed interest rate works. Yo can’t just “deposit” your money with the Fed and have it pay you interest. What you can do is buy bonds. These bonds have a price and maturity date. When the interest rates keep increasing the bonds which pay less interest become cheaper. So you can’t just “withdraw” what you deposited. You have to sell what you “bought” but for much less than you paid, so you go under.

You can't. Banks can.

https://www.newyorkfed.org/markets/reference-rates/obfr and things like that.

The OBFR is the rate banks pay to borrow reserves from each other. You’re thinking of interest on reserves (IORB) [1], which is interest the Fed pays to interfere in that market.

[1] https://www.federalreserve.gov/monetarypolicy/reserve-balanc...

> stick them in with the fed providing 4% interest

The Fed pays interest on those reserves to keep them from being lent. A narrow bank wouldn’t be able to lend its reserves. As such, it makes no sense for the Fed to pay interest on them.

Increasingly banks are only lending to a different parts of the federal government anyway-—deploying their cash on treasuries and agency debt. They are being driven to this by the risk weighted capital adequacy requirements.

The whole thing resembles an ever more intricate Rube Goldberg machine.

> increasingly banks are only lending to a different parts of the federal government anyway-—deploying their cash on treasuries and agency debt

This is incorrect. In 2022, the largest increases were in credit cards (+17.4%), commercial and industrial loans (+14.5%), consumer loans (+11.6%), commercial real estate loans (+11.3%) and real estate loans (+10.1%) [1]. Treasury and agency security holdings across American banks were majorly added between 2019 and 2021, and then spent down as balance sheets swapped. (They remain, of course, a commanding fraction of bank assets for obvious reasons.)

[1] https://www.federalreserve.gov/releases/h8/current/

What‘s the decade by decade trend?

The problem is that bank lending is the main form of money creation in the economy.

If retail banks switched en mass to backing deposits 1-to-1 with Fed reserves then the Fed would quickly take strong action to disincentivise this strongly contractionary effect.

They have plenty of legal tools to do this (e.g. IORB rate). They don’t need to resort to not giving out licenses.