Monday, September 14, 2009

Loans Create Deposits -- how banks actually work

I used to believe that banks lent out deposits. So, you deposit money in a bank, and the bank turns around and loans that money out to someone. "Fractional reserve accounting" meant that the bank could lend out more money than it had on deposit, which seemed fishy. As the bank usually made long term loans funded by short term deposits (mismatched maturity) they were always at risk of a bank run, which is where FDIC insurance came in.

I was wrong.

At a macro level, banks create deposits by extending loans. At an individual level they do the same thing, but also need to manage reserve accounts. This is how it works:

Loans create deposits

Banks do not loan out deposits. When you deposit money in a bank, you are not putting it "hard at work" in any way. It is not funding factories, houses, or anything else. Saving money in a bank is the same as putting it under the mattress.

When an individual bank (bank A) makes a loan, it creates (credits) an asset (receivable). To make its books balance, it debits its reserve account at the Fed (also an asset) by the same amount. The loan creates a deposit, say in a different bank (bank B). That bank credits the deposit on the liability side, and credits its reserve account at the Fed. The reserve account debit made by bank A is exactly matched by the reserve account credit made by bank B. While the total amount of reserves in the system is unchanged, bank A is now short reserves, and bank B is long (assuming they had the correct amount of reserves in the first place).

Bank lending is not reserve constrained

So, what will bank A now do, since it is short reserves? It can try to win over bank B customers and get them to move their deposits over, this transferring the reserve account assets bank B has to itself. If bank A fails to do this, then it simply borrows the reserves it needs overnight from... bank B. The overnight lending market is designed to do exactly this. Bank B, in this case, happens to have exactly the quantity of reserves bank a needs, and since reserves earn no interest, is happy to lend to bank A at the federal funds rate, which is the overnight interbank lending rate.

Suppose the system as a whole is short of reserves? In this case, the banks who are short on reserves borrow directly from the Fed's discount window.

Suppose the system as a whole has excess reserves? In this case, the Fed sells treasury securities, which enable banks to transfer assets from their reserve accounts (at the Fed) to their treasury accounts (also at the Fed). The Fed sells treasuries not to "finance" the deficit, but to drain the system of excess reserves, so there will be overnight lending, and thus, a non-zero interest rate. If the Fed did not drain extra reserves from the system, banks would not lend, and the overnight lending rate would be zero.

Again, loans create deposits

So, at an individual bank level, banks make loans as they like, and then either borrow the reserves they need at the discount window or overnight market, OR the grow their deposit base to meet their reserve requirements. But the banks make the loans first, and then do whatever to hit their reserve target at the end of the day.

At a macro level, the banking system makes loans, which in turn create deposits.

Note to Austrians: the problem is not maturity mismatching, the problem is private sector credit extension

You'll note that individuals cannot make loans the way banks can. We do not have the facilities they have at the Fed. Austrians think that it is maturity mismatching that enables banks to lend money, but they are wrong (although maturity mismatching has it's own problems), the issue is private sector credit extension. Private sector credit extension expands private sector balance sheets, can it can do this to whatever level it can get away with.

Capital requirements constrain bank lending

So, if reserve requirements don't limit bank lending, what does? Answer: capital requirements. Bank balance sheets must rest on a certain core of paid-in equity, but unfortunately, regulators frequently fudge, ignore, or otherwise enable banks to ignore those rules and leverage themselves to calamitous levels, getting paid obscenely in the process.

I leave you with one of the most remarkably ignorant posts I've seen on this topic, notable only because the individual actually works within the banking system! On a comment forum in Econlog, he stated
The point here is that a particular bank cannot be sure where any disbursed loan funds may end up and can only count on a small proportion returning. This means that they cannot simply inflate their balance sheet by making loans and counting on the funds to come back as deposits. Additionally, if they did this their liquidity ratios (one of the key measures of any bank's soundness) would steadily deteriorate, restricting their ability to make further loans. Their capital ratios would also steadily deteriorate, reducing their ability to raise funds.
Ozrisk does not seem to realize that making loans reduces a bank's ability to make further loans, which is exactly how capital requirements are supposed to work, and their whole point.

12 Comments:

Blogger businesscapital said...

Thanks for sharing with us.Your blog is very informative.keep posting.. capital loans

9:59 PM  
Blogger parviziyi said...

I found the above worth my while. One thing it doesn't mention is that under US bank regulations there are no reserve requirements on customer savings account deposits or any other kind of bank deposits except for "transaction deposits", aka checking account deposits, and of course customers don't keep big sums in checking accounts for long. Under the bank regulations in the UK and Canada, there are no reserve requirements at all.

What Winterspeak is saying here is that the US reserve requirements for transaction deposits doesn't limit bank lending because the regulations allow the bank to borrow money to meet its reserve requirement.

5:48 AM  
Blogger gac said...

"When an individual bank (bank A) makes a loan, it creates (credits) an asset (receivable). To make its books balance, it debits its reserve account at the Fed (also an asset) by the same amount."

The debits and credits above are incorrect. An increase in an asset account is created via a debit. An increase in a liability account is made via a credit. So, when Bank A makes a loan, it debits the Loans Receivable account on its books. To makes its books balance, it creates (credits) a new liability account in the name of the borrower, no differently than if the borrower had made a cash deposit at the bank. Reserves do not come into play, except for the requirement that Bank A have the required reserves at the Fed to back the newly made loan.

8:06 PM  
Blogger gac said...

Further to my earlier post, here's how back reserves come into play when a bank loan is made.

When the borrower proceeds to draw down his new loan account at Bank A by writing cheques on it, those cheques will be cashed at Bank A or any other bank.

If the cheques are cashed at Bank A and cash is disbursed, Bank A will credit (reduce) its holdings of cash and debit (reduce) the borrower's liability account on its own books.

Alternatively if the cheques are cashed at Bank B, Bank B's cash holdings will also be reduced, and its Reserves held at the Fed (an asset account on its books) will be debited (increased) commensurately. Bank B will then send the cheques drawn on Bank A for clearing at the pertinent Fed branch and Bank B's reserves with the Fed (a liability account on the Fed's books) will be credited (increased), while Bank A's reserves will be debited (reduced).

I realize the above can get confusing as debits and credits have a opposite effects depending on whether they are applied to asset or liability accounts.

8:21 AM  
Blogger winterspeak said...

gac:

ack! you are correct -- I believe I got my debits and credits mixed up.

Just read "increase" whenever I say credit, and "decrease" whenever I say debit.

I am ashamed. I'll try and do better next time.

11:03 AM  
OpenID Shalom Jeff said...

Winterspeak,

Do deposits get destroyed if loans are repaid or written off?

Thanks.

12:29 PM  
OpenID Shalom Jeff said...

Winterspeak,

Do deposits get destroyed if loans are repaid or written off?

Thanks.

12:30 PM  
Blogger winterspeak said...

Jeff: Yes.

When you pay down a load, both the asset and liability get destroyed.

1:22 PM  
OpenID Shalom Jeff said...

winter: Thanks, can you provide evidence?

9:54 PM  
Blogger winterspeak said...

Jeff: It's just accounting.

A bank books a loan as a receivable (asset). When the load is paid down, the bank reduces that receivable. When the loan is fully paid off, then the receivable no longer exists (as the bank cannot expect additional payment to come in to service that loan).

The money for the paying the load was sitting in a deposit account somewhere as a liability. As that money left the deposit account, that liability was written down.

When loans are created, balance sheets expand (both sides get bigger). When loans are paid down, balance sheets contract (both sides get smaller).

9:59 AM  
Blogger winterspeak said...

Jeff:

Sorry, I think I misunderstood your question:

"Do deposits get destroyed if loans are repaid or written off?"

If a loan is repaid, then yes, the deposit vanishes and the receivable vanishes as well. When the bank cashes the check, it decreases the money in the checking account and the outstanding loan balance (asset) at the same time.

If the loan is written off, then the bank writes down the receivable (asset) and writes down it's own equity. The bank's deposit base is unchanged.

9:35 AM  
Blogger Tom Brown said...

I enjoyed your post. There is one sentence that I find odd though. You wrote: "If the Fed did not drain extra reserves from the system, banks would not lend, and the overnight lending rate would be zero." When you say "banks would not lend" you mean "banks would not lend reserves to each other," correct? I don't see how the extra reserves would cause them not to lend to customers... it should in fact incentive them to loan to customers since it would be cheaper for them to meet their reserve requirements.

Can you post some references for you description? Your description matches what I've read from Steve Keen and Scott Fullwiler (Google "Krugman's Flashing Neon Sign" for example).

For another example of a blogger that doesn't appear to know what he's talking about, Google "Tamny Ron Paul Fractional Reserve Banking Money Multiplier Myth" on Forbes.com.

7:58 AM  

Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home