I've spent considerable time and effort to try and understand precisely how banks can currently create money. One issue that arises--if banks have such a good deal, why doesn't everyone with a few million dollars to invest go out and start one? I would appreciate it if those with intimate knowledge of how banks work, would blow holes in the following scenario.
Primary Question: Can a U.S. bank utilize fractional reserve banking to multiply it's own assets and then lend them to the government in the form of 30 year treasury bonds?
Secondary Question: If this is unfeasible for a single bank, could a cartel / keiretsu accomplish it as a group endeavor, or could it spontaneously happen systemically?
Scenario:
Apparently it’s a good time to start a bank (see http://online.wsj.com/article/SB123724564069748443.html) Let's say a group of investors pool their funds and create a holding investors called ABC Holdings. The Holding Company incorporates a new entity ABC Bank and capitalizes it with $10 million. StartABank.com is consulted, and eventually a federal bank charter is applied for and obtained from the Office of the Comptroller of the Currency (OCC).
ABC Holdings now creates 10 new entities, called ABC Investment Trust #1 thru #10. ABC Bank executes the following series of loans (If lending to controlled entities is a problem, then assume that they are affiliated interests that are structured in creative ways):
All investment trusts invests in 30 year treasury bonds, yielding on average 5% interest annually, or roughly $3 million. By lending the newly earned $3 million to the investment trusts in the same manner described above, total assets increase by $20 million to $80 million. Year 2 investments bring in $4 million, which enables $24 million in new loans to be made increasing total assets to about $104,000. Year 3 investments bring in about $5 million, which enables $30 million in new loans to be made, increasing total assets to about $134,000,000.
By year 3, the following situation exists:
Huh, very interesting post. I considered answering it and then saw this:
luftmensch: I would appreciate it if those with intimate knowledge of how banks work, would blow holes in the following scenario.
luftmensch:One issue that arises--if banks have such a good deal, why doesn't everyone with a few million dollars to invest go out and start one?
They already have! In seriousness though, it would take more than a few million dollars. I'm not sure what the capitalization requirements are, but I'm sure they are enormous. The barriers to entry in Banking are extraordinarily high, as they are for any legally sanctioned cartel: if you and your friends had what amounts to a free pass to create money, you'd be very selective about who you allowed in to your little club, wouldn't you?
============================
David Z
"The issue is always the same, the government or the market. There is no third solution."
The scenario is specifically prohibited by regulation W. In particular, section 223.11 states:
"A member bank may not engage in a covered transaction with an affiliate (other than a financial subsidiary of the member bank) if the aggregate amount of the member bank's covered transactions with such affiliate would exceed 10 percent of the capital stock and surplus of the member bank."
You ask us to assume that the entities are structured in creative ways, but finding the creative ways around this regulation that won't attract Fed scrutiny is exactly the problem.
luftmensch:I would appreciate it if those with intimate knowledge of how banks work, would blow holes in the following scenario.
If you don't get an answer here then I suggest you try the forum at The Market Ticker. I can think of nobody more likely to know the answer than Karl Denninger, his entire raison d'etre seems to be "banks behaving badly".
While you're at it, I would also highly recommend taking a look at the little known cargocultist.com, for an up-to date and intelligent analysis of the murky world of FRB.
And if you ever get a good answer to you question - please send me a PM as I'd like to know too.
What Went Wrong with Economics
why thank you, mick.
The main problem with what you've described is that re-depositing the money at Bank ABC doesn't increase the Bank's capital to $19 million, it increases the deposits. As a new bank, created as you've described it wouldn't in fact be able to lend any money at all - equity capital (the original $10 million), is held completely separate to customer deposits.
let's assume that it manages to get $10 million in deposits, it can then create a loan to a trusted confederate of at least $9 million. If the confederate re-deposits that money at the bank, then the bank has $19 million in deposits, but can't create any more loans because it doesn't have the equity capital to support them. (Under Basel regulations, a bank's equity capital reserve must be at least 10% of its total loan book.)
Banks are only allowed to increase their equity capital if they make a profit, so that limits the scope for increasing equity capital holdings.
However, if the bank can sell some or all of its loan book - say in the form of Mortgage Backed Securities - then it can use the profits from that sale to increase its equity capital, and it can then make more loans.
Another way to exploit the same mechanism, in a way that won't attract too much attention, at least not until it's too late - is to make friends with a mortgage broker and a house builder, and find a nice tract of land somewhere.
-- cc
So if the bank had $10 million in equity capital reserve, and $10 million in deposits...would that allow it to, in thoery, loan out up to $100 million? (The $10 million being 10% of the total loan book). Does that work, or I am missing something?
It sounds like to really understand banking, one has to study Basil regulations, OCC guidelines...and probably a few other rule-books.
BTW I just read your blog...I've been trying to figure out a lot of the same things. So I think you are saying:
1. There is a misperception that banks can lend out 10x their deposits.
2. Banks can only lend out 90% of deposits, but through the "redeposit process", the aggregate amount of their loans can come to equal 10x the amount of deposits. (So, the result of #1 is correct, but the way it happens is mis-identified). Also related, total amount of lending can't exceed 10x equity capital.
3. (And this is the scariest thing)--Loan securitization such as MBS, ABS, etc., enables the loans to be sold off, therefore the loan goes off the originators loan books...therefore they can use this mechanism to expand the money supply "indefinitely", while also increasing their equity capital through loan fees, etc. Tell me, does "HFI" have somethnig to do with this? I.e., can the bank "securitize" a loan and not even sell it off, retain ownership of it, yet still give it a value as if it were an asett rather than a liability, by accounting for it now as an investment?
I agree with you that this is a critical topic to understand, very few people seem to understand it intimately...and if we are to have a proper discussion, we should be precise with the facts. Funny thing is, I've been trying to figure this out for years, have had numerous conversations with pretty educated people, and worked at a large bank for awhile working with MBS related data, and still am trying to figure out these details.
A few more thoughts:
I guess the other question that arises is...is it not insanity to give anyone with say $5m-$10m the ability to create $50m in loans from $5m in equity and $5m in deposits? Is that kind of power really appropriate to be handed from the state into private hands to profit from? Is it possible to have a public discussion about this?
luftmensch: So if the bank had $10 million in equity capital reserve, and $10 million in deposits...would that allow it to, in thoery, loan out up to $100 million? (The $10 million being 10% of the total loan book). Does that work, or I am missing something? It sounds like to really understand banking, one has to study Basil regulations, OCC guidelines...and probably a few other rule-books.
I will qualify that a little, I have never actually seen that written down in any of the banking regulations, but I have to presume it's there somewhere. It is generally implied. Also, in data-mining the FDIC call reports I've only noticed one case where loans weren't less than deposits. As far as I could tell, that particular bank had just been through a merger, so I suspect it was in the process of sorting out its loan book.
Yes, pretty much. I find it extraordinary that in the 21st century there is no consolidated book/central bank site, etc. that does pull it all together. Personally, I recommend starting from the American Call Reports at the FDIC. They're the detailed accounting records, so they make it possible to see what's happening at the individual bank level in a real system, and work backwards.
-- c
Because the government restricts it. Banking is one of the most highly regulated industries in the country.
At most, I think only 5% of the adult population would need to stop cooperating to have real change.
luftmensch: BTW I just read your blog...I've been trying to figure out a lot of the same things. So I think you are saying: 1. There is a misperception that banks can lend out 10x their deposits. 2. Banks can only lend out 90% of deposits, but through the "redeposit process", the aggregate amount of their loans can come to equal 10x the amount of deposits. (So, the result of #1 is correct, but the way it happens is mis-identified). Also related, total amount of lending can't exceed 10x equity capital. 3. (And this is the scariest thing)--Loan securitization such as MBS, ABS, etc., enables the loans to be sold off, therefore the loan goes off the originators loan books...therefore they can use this mechanism to expand the money supply "indefinitely", while also increasing their equity capital through loan fees, etc. Tell me, does "HFI" have somethnig to do with this? I.e., can the bank "securitize" a loan and not even sell it off, retain ownership of it, yet still give it a value as if it were an asett rather than a liability, by accounting for it now as an investment? I agree with you that this is a critical topic to understand, very few people seem to understand it intimately...and if we are to have a proper discussion, we should be precise with the facts. Funny thing is, I've been trying to figure this out for years, have had numerous conversations with pretty educated people, and worked at a large bank for awhile working with MBS related data, and still am trying to figure out these details.
2) Well - i believe there are actually multiple problems here. The standard economics textbook explanation is that the banking system multiplies the money and loan supply 10* from an "initial deposit into the system" The trouble with that description is that the initial deposit in some sense took place about 3 centuries ago, so if the textbook model was correct, we would now be at the point where everything is fully expanded. In which case the banking system is just managing a fixed supply of loans, and re-issuing them as they get repaid.
Clearly that's not what's happening, so the problem is to figure out what exactly is - and that of course, is where the various misconceptions about how the thing is implemeted in theory, and then in practice, has increased the noise to signal ratio on this debate to quite an extraordinary degree.
3) Quite possibly. We know that the Banks got caught with what has been described as 'a surprisingly large amount of securitised products', when the credit crisis hit, and they had to recognise some of the underlying losses. In effect, that is exactly what they are doing when the move loans to equity capital, and it wouldn't surprise me at all if they've found other ways to do it. According to BIS figures there's something like $6 trillion dollars of these things outstanding, just for the United states - and it's a very interesting question as to where they actually are. It's also an interesting question of how exactly they generated that large a quantity of loans in such a short period - i have a nasty feeling a lot of it is probably the financial industry lending to itself to buy securitised loans in some shape or form.
And yes, like you i've been very surprised by the ignorance on the details of this thing i've encountered both in the banking industry, and also amongst economists.
Spideynw: luftmensch:One issue that arises--if banks have such a good deal, why doesn't everyone with a few million dollars to invest go out and start one? Because the government restricts it. Banking is one of the most highly regulated industries in the country.
http://www.fedpartnership.gov/bank-life-cycle/start-a-bank/de-novo-bank-application-process.cfm
The paperwork is fairly heinous - but it's straightforward.
To answer the original question. Increasingly they are. It's partly even knowing that it's possible (rather than assuming that it's not), and partly having access to the necessary money. A few years back, $2 million was a lot of money - but a side effect of the money and debt expansion over the last 15 years is that that's no longer the case.