Grieving for Glass-Steagall

Glass-Steagall is dead. Rather like Soviet-era Communist leaders, it has been officially dead since 1999, and actually dead for much longer. Though there was no state funeral, the body was not embalmed or put on display and few people mourned its passing.

Well, not at the time. But fast forward to 2008 and suddenly the Western world - not just the US - exploded in a paroxysm of grief over the demise of Glass-Steagall. "If only Glass-Steagall hadn't been repealed!" people cried. "Glass-Steagall would have prevented all these banks failing. Glass-Steagall would have stopped all these derivatives being created. Glass-Steagall would have protected everyone's money". Glass-Steagall, it seems, could have prevented the entire financial crisis. Never mind that Lehman, Bear Sterns and Merrill Lynch were pure investment banks, with no retail deposits to put at risk. Never mind that AIG was an insurance company and Fannie and Freddie were government-sponsored enterprises - none of which were subject to Glass-Steagall's provisions. Never mind that Countrywide and most of the other mortgage originators that together created the largest fraud in US corporate history were pure retail lenders and therefore ALSO not subject to Glass-Steagall's provisions. And never mind that the large universal banks in the US survived the financial crisis relatively unscathed - which might not have been the case if Glass-Steagall had prevented them diversifying.

Ever since, there have been calls for its resurrection in some form or another. The UK is adopting a baby Glass-Steagall. The Eurozone is thinking about a gauzy curtain. And the US has brought in a distant relative - the Volcker Rule. But people are not happy. "WE WANT GLASS-STEAGALL! BRING BACK GLASS-STEAGALL!" they cry. Bizarrely, even people in the UK cry this. The UK never had Glass-Steagall. I wonder sometimes if people really understand what it is they are demanding.

Four years on, there is still an immense amount of nostalgia for the age of Glass-Steagall. And there are repeated attempts to bring it back in some form. The latest attempt is by Senators Warren, Cantwell, McCain and King. Warren admits that Glass-Steagall would not have prevented the financial crisis, and would not end "too big to fail", but none-the-less wants the large universal banks to be forced to divest their investment banking activities, apparently in order to improve the "culture" in retail banking. But as I've noted before, the cultural shift in retail banking away from service and towards aggressive product selling had nothing to do with investment banking: it long pre-dated the repeal of Glass-Steagall and was due to low margins and cut-throat competition in the retail banking sector.

Warren - like many others - wants separation of investment banking from retail to prevent "insured deposits" being used to fund risky activities. Sadly that demonstrates a total lack of understanding of the real funding problems in the financial crisis, or indeed of the nature of bank lending. The problem was actually that wholesale funding - which is prone to runs, as I shall explain - was excessively relied upon to fund risky activities, including retail lending. Retail lending may be "boring", as Warren claims, but it isn't "safe". Deposits that back retail loans on bank balance sheets are at risk. If they are insured, that puts taxpayers' money at risk just as much as if those deposits were used for securities purchase.

Nevertheless, the calls for separation of "boring" retail banking from "risky" investment banking continue. And there is an important issue here. It concerns the extent to which we are prepared to provide public funds to support the activities of banks. In essence, Glass-Steagall and all its relatives are an attempt to limit the activities that can be supported by public funding - by which what people really mean is liquidity support from central banks.

The trouble is that this is actually impossible. When investment banks are separated from commercial banks they become their customers. Their cash balances sit in commercial banks, because ALL money that isn't in the form of physical notes and coins sits in commercial banks, one way or another. All financial market trading is intermediated through commercial banks. All new stock issues are intermediated through commercial banks. Payment fails don't just affect the recipients, they affect the liquidity of the banks through which they are intermediated. In practice it is simply impossible to remove liquidity support from payments arising from market trading, and very dangerous to attempt it - as the Fed discovered when Lehman fell. Remove central bank liquidity support for market trading activities and the entire market collapses like a house of cards.

There is a prevalent belief that if a Glass-Steagall separation were imposed, financial markets could be allowed to collapse without commercial banking being affected. This is completely wrong. When major customers of commercial banks fail, the banks themselves are at risk - and by extension so are their retail customers. The Lehman collapse and consequent market freeze very nearly caused the failure of the global payments network, which would have had catastrophic effects on retail customers. Central banks have to provide liquidity support for ALL payments, whatever their source, not to protect investment banks but to protect the retail customers of commercial banks.

It's not enough to provide liquidity support after the event, either. Perception of liquidity support is really important. It is the perception of illiquidity that causes bank runs. Institutional investors are nervous creatures. If they believe they might lose access to their funds, they will pull them. Note that this has nothing to do with whether or not there is a real risk of actual loss: even if lending is collateralised with good quality assets there can still be runs. When funding stresses make headline news and there is no apparent liquidity support, people pull their money from the banks affected. So when investment banks are denied liquidity support, institutional investors will pull their funds if they smell trouble - and the knock-on effect is a run on commercial banks, since investment banks are customers of commercial banks. Admittedly, all that really happens in a modern bank run is that money moves from one commercial bank to another - but large unexpected flows of money are destabilising for the banking system as a whole and can be fatal for individual commercial banks.

Liquidity management for commercial banks is a huge issue. Regulatory pressure at the moment is pushing them to increase their holdings of liquid safe assets such as government debt and to fund themselves more with retail deposits than wholesale funds. Retail deposits have traditionally been slower to run than wholesale funds: in the past this has been partly due to retail customer apathy, and partly because retail depositors have had to wait for banks to open in order to withdraw their money, whereas institutional investors can move their money by trading overnight on Far Eastern markets. But this difference is disappearing fast as internet banking makes it possible for retail customers to make payments and move money around at any hour of the day or night. If you make it easier for people to move their money, they are more likely to do so.

Matthew Klein thinks that separating deposits from lending would solve the problem, because it would mean that deposits were never used to fund risky lending. Most bank deposits are moving balances. There are payments in and out of transaction accounts of all kinds all the time: some of those are retail transactions and others come from institutional or market sources. It is not realistic to say "we will provide liquidity support for transactions from retail sources but not from wholesale ones". The same banks are at risk from payment fails whatever the sources of the transactions, and therefore so are their customers. Klein wants to protect all deposits, irrespective of their size or origin, by backing them with central bank reserves - which neatly solves the transactions problem, but falls foul of popular abhorrence of public support for large institutional, corporate or high-net-worth individual deposits.

Of course many market trades are never cash settled. And it might be assumed therefore that they do not affect commercial banks. This would be wrong. Many of them have cash margin. And cash margin is held in banks. One way or another, all money is held in commercial banks, and it isn't in practice possible to distinguish between different "pots" of money in the provision of liquidity support, as proponents of structural reform (of all kinds) seem to think. Money is money, whatever its source: if it disappears in transit - whether because of customer default, market freeze or payment system failure - its absence causes problems for both the recipient customers and their banks.

Ensuring the smooth operation of payments has become the primary purpose of central bank liquidity, because advanced economies have allowed themselves to become completely dependent on commercial banks to facilitate the vast majority of cash transactions. Separating out different bits of banking would not eliminate this need: all it would do is create the impression that some types of transaction would not be supported, which would increase the likelihood of highly damaging market freezes or runs even though post-Lehman no central bank would dare withdraw liquidity support for non-retail transactions.

The problem it seems to me is that we have universal access to payments through the same set of commercial banks. There might indeed be a case for completely separating retail (insured) deposit-taking from everything else, perhaps in the form of a new public utility, but it would have to have its own payment systems directly supported by central banks. Would this eliminate the need for central bank liquidity support for non-retail transactions? I'm not sure, frankly. I suspect that even if retail payments were separated from wholesale in this way, there would still be untoward economic effects from financial market failure, which would hurt Main Street just as much if not more than losing their deposits. Hatred of all things non-retail is not sufficient justification for systematic dismantling of central bank support. The economic impact of such an act needs to be fully understood - which at the moment it is not.

What is clear though is that a 1930s-style separation of retail and investment banking is meaningless while all payments go through the same set of commercial banks. We forget that there were no payments systems in the 1930s. Everything was settled in physical cash. Do we really want to go back to that?

Glass-Steagall is dead. It is not appropriate for our modern banking system. Grieve for a bygone age, and move on.

Related links:

The 21st Century Glass-Steagall Act - James Pethokoukis
Daniel Tarullo questions wisdom of return to Glass-Steagall - Politico
Five facts about the new Glass-Steagall - Simon Johnson
What Glass-Steagall 2 gets wrong: Everything - Matthew Klein
Supermarket banking - Coppola Comment
How to lose 3 million dollars in one second - Chris Arnade


  1. Frances,

    Quite agree with your attack on the “bring back GS” simpletons. GS is better than nothing, but that’s all that can be said for it.

    As to the UK’s “baby Glass-Steagall” I assume you’re referring to Vickers. For the benefit of anyone interested, Laurence Kotlikoff wrote a scathing attack on Vickers here:

  2. Frances,

    I more or less agree with your claim that “The problem was actually that wholesale funding - which is prone to runs….”. So the solution is to forbid banks which take any significant risk to fund themselves with run-prone liabilities, as suggested by John Cochrane in this WSJ article, is it not? See:

    As Cochrane puts it “Don't let financial institutions issue run-prone liabilities.”

    (Incidentally it’s not just wholesale funding that causes runs. The same problem applies to retail funding: witness Northern Rock.)

    And Cochrane’s solution is very much the solution proposed by Positive Money, Richard Werner and the New Economics Foundation here:

    Which in turn is the same as Matthew Klein proposes in his last paragraph.

    Klein does not propose, as you put it “providing liquidity support for transactions from retail sources but not from wholesale ones". The wholesale / retail distinction is irrelevant to what he is saying. He is saying that where a depositor puts $X into a bank and wants $X back and not one cent less, then the bank cannot do anything remotely risky with that money. Eminently logical I’d say. And in contrast, where a bank wants to do something risky, that can only be funded by shareholders or quasi-shareholders.

    That solves pretty much EVERY PROBLEM associated with banks. Certainly sudden bank failures become impossible. Plus the TBTF subsidy disappears (something which the 10,000 pages of Dodd-Frank and Vickers have failed to achieve). Plus there is no need for the GS split into investment and retail banks, or for Vickers’s hopelessly vague “ring fence”.

    1. 1) I did discuss retail runs in the post. They are much less frequent than wholesale runs. NR's retail run occurred after the wholesale run (on ABCP following the closure of the Bear Sterns hedge funds) that forced it to seek emergency liquidity from the Bank of England.

      2) As I understand it Klein is advocating full reserve banking. I may have implied something else - he thinks so. I am waiting for him to come back to me and if I have misconstrued what he said I will correct.

    2. Klein doesn’t overtly advocate full reserve, far as I can see, but full reserve is an almost inevitable consequence of his proposals.

      Advocates of full reserve have actually been aware of that close linkage between fail safe banking systems (like Klein’s) and full reserve for some time. I.e. set up a full reserve system and you almost inevitably end up with a fail safe system. Conversely, set up a fail safe system, and it’s hard to avoid ending up with full reserve.

    3. I've read Klein's other pieces. It is full reserve banking - he's a supporter of the IMF's revised Chicago plan.

    4. > Ralph Musgrave

      If by full reserve banking you mean deposits with a risk free value in themselves then that is not possible.

      Every deposit is a claim on a the value of a loan , and that loan has allready been made. You cant have deposits that have a risk free value in themselves, you can only underwrite the value with another source of funds

  3. "We forget that there were no payments systems in the 1930s. Everything was settled in physical cash. Do we really want to go back to that?"

    I'm not sure what you mean by this. But settlement has not been in cash for centuries. Bank payment and clearing systems developed out of the medieval fairs of the 12th century -- and have a much longer history than central banking. Physical cash hasn't been the primary means of settling commercial debt in major urban centers for more than half a millenium.

    1. But everything was settled by some physical representation of the transaction - such as a cheque or a bank draft, which are really only forms of physical cash with more limited fungibility. There were no computers and no automated payments. Transactions did not generally go through banks or if they did they took days to clear. It was a totally different world. That's the point.

    2. Settlement has been via bank clearing for centuries, because bank money (i.e. IOUs with bank guarantees or starting in the mid-18th century bank deposit accounts accessed by checks) has been the primary form of commercial currency for centuries. Even before computers clearing took place in major urban centers on a daily basis. Certainly clearing was slower than it is today, but the basic structure of exposure and potential contagion due to the payments/clearing system did not arise in this century.

    3. Oops. That should be "starting in the mid-19th century" ...

    4. Yes, for large transactions bank clearing - cheques/IOUs - has existed for a long time, as you say. But that is not true for retail transactions. Until the 1980s the vast majority of retail transactions took place outside banks in the form of physical cash exchanges. Indeed most people did not use banks much until the 1960s/70s. That is what has changed.

      I would also argue that the standard settlement periods T+3 and T+5 took a lot of risk out of the system. There simply wasn't the constant need for liquidity that there is today with same-day transfers and faster payments - yes, I know that even now we are still only down to T+2 for securities transactions, but they are only a small part of the total, and most transactions through banks now are much faster than that. So although the basic structure was the same, the slower pace due to lack of automation made a great deal of difference.

    5. Ahh. I think I understand now why I find your approach in this post so confusing. "Retail" banking for the purposes of Glass-Steagall does not refer to consumer banking, but rather to commercial and consumer banking (and covers the bulk of economic activity related to the purchase and sale of goods and non-financial services). Your focus on consumer banking appears to me to significantly confuse the actual purpose of G-S which is to separate commercial banking from investment banking.

      Historically bank clearing was used precisely for commercial transaction (consumer transactions are relatively recent phenomenon, as you observe). But such commercial transactions were not necessarily large, instead almost all commerce that passed through urban centers has taken place on the basis of bank money for centuries.

      The history of bank liquidity crises and contagion (often from one country to another, cf. 1763) dating back at least to 14th century Venice makes me doubt your claim that the fundamental nature of the demand for liquidity has changed due to modern technology -- though it does seem likely that technology changes the speed with which crises appear.

    6. Really you have just made my case for me. It is the inclusion of what you call "consumer" banking in commercial banking that renders Glass-Steagall worthless.

      Consumer transactions now dominate what in the UK we call "retail" banking and I would guess they do in the US too. What were once "clearing" banks are now "retail" banks providing depository and payments services to ordinary working people and small businesses as well as selling lending products. And all of those transactions, whatever their origins, go through the same banks and the same payments systems.

      Mind you, from what you have said I am not convinced that Glass-Steagall can ever really have had the effect that people claim for it. It was of course only one of a number of structural and regulatory changes at that time. I suspect many of the stabilising effects claimed for G-S actually come from other sources - such as the extensive system of Federal support for mortgage lending.

    7. "Consumer transactions now dominate what in the UK we call "retail" banking and I would guess they do in the US too."

      Are you saying that the value of individual deposits in the U.K. banking system is greater than the value of non-financial business deposits? That would very much surprise me. Do you have a link to your data?

    8. Also, the whole point of banking systems of payments and clearing has always been to support bank lending. I don't understand why you appear to think bank lending to small businesses is a new phenomenon rather than a profession with a pedigree of centuries.

    9. It is volume of transactions not their value that is the issue. The volume of consumer transactions is simply huge now.

      I never suggested or indicated that I thought lending to small businesses was a new phenomenon. I was talking about deposit-taking and payment services, not lending. And those - particularly the latter - are relatively recent.

  4. It doesn't make sense to separate out the volume of consumer transactions from the volume of business transactions, because most consumer transactions are with businesses, so almost all "consumer" transactions will also be "business" transactions.

    It's clear that the finance of consumer transactions is relatively new (unlike the finance of small business transactions) if that's your point. But I don't understand how consumer finance can be a destablizing seachange in the nature of banking or in particular of the payments system.

    Deposit banks -- which of course also offer payment services -- are the oldest form of banks dating back at least to 13th c Venice. Reinhold Mueller wrote the book on Venetian banking.

    1. I don't think you realise just how huge the volume of retail transactions is. Almost all transactions that used to be cash - not involving banks - now go through bank accounts. And I also don't think you know how fragile the payments systems are: they were never designed to handle the volumes of transactions that they do. It's all very well pointing to how things were done hundreds of years ago and say "well it worked ok then". It does not work the same way now: it is all far faster, there is much less human intervention and the volumes are HUGE.

      However, I'm really not at all sure what all of this has to do with the subject of the post, which is that Glass-Steagall separation would not eliminate the need for central bank liquidity support for investment banking operations. Investment banks are customers of commercial banks and settlement of their transactions goes through the same set of payments systems as commercial and what you call "consumer" transactions. Separating them out and providing liquidity support to some transactions and not others is simply unrealistic.

      I think in your zeal to show that I don't know what I'm talking about you've rather drifted away from the point.

  5. "I think in your zeal to show that I don't know what I'm talking about you've rather drifted away from the point."

    I apologize if it came off that way. Inaccurate descriptions of where we came from seem to me to seriously bias the quality of the discussion of where we're going. So I do what I can to make people more cautious about their claims about our history.

    I have to admit when it comes to the volume of transactions, I always thought that bank to bank transactions dwarf almost everything else in the financial world (e.g. the infamous derivatives, but also foreign exchange, etc.), so it surprises me that you bring consumers into the picture.

    I certainly agree that we have evolved an investment banking system that is heavily dependent on central bank finance. But I don't see how that fact addresses the question of whether we should have an investment banking system that is heavily dependent on central bank finance. Arguments based on "realism" appear to be efforts to avoid the underlying question: Is this the kind of financial system we want to have? And how do we get from what we have to what we want?

    1. Bank to bank transactions dwarf everything else in value but not in volume. These days retail transactions are the highest volume as far as I know.

      I had no intention in this post of addressing the question of whether we "should" have an investment banking system that is heavily dependent on central bank finance. The point I was making is that if we wish to end investment bank dependence on central bank liquidity, Glass-Steagall 2 will not achieve that.

      In the penultimate paragraph I did suggest a way in which ending investment bank funding by central banks might be achieved. And then I went on to question whether it was actually desirable, or whether allowing investment banking payments to fail would have costly economic effects. I concluded that we don't actually know and more work is needed to understand the likely impact of such a change.

      It seems that you really wanted me to have written a different post.

    2. Fair enough. You're one of a small number of bloggers, whom I actually find worth reading. That was one of the reasons your penultimate paragraph set me off.

    3. Why, thank you!

      A couple of small points from the previous comment:

      - I would not regard FX transactions as necessarily "bank to bank". There is a huge number of small FX traders out there - it's an incredibly diverse OTC marketplace.
      - It is also worth bearing in mind that most securities and derivatives trades are not settled, so there are far fewer payment transactions than the volume of trades would suggest - though margin calls do have to be settled. And of course we do bulked payments for high-volume transactions both in investment banking and in retail.

  6. Isn't there a question of scale of the liquidity support required under each system?

    In the Lehman case, there was a direct link to commercial banks (European and US) who owned in their own account, via their investment banking arms, AIG-insured subprime paper in large enough quantities to have become insolvent had AIG and the subprime paper failed at the same time, and then they'd of course would have then taken all the others down with them if those with the exposure had been taken into administration.

    Had worldwide separation been in place at the time of Lehman (OK, it's a bit of improbable but as a thought exercise...), it seems conceivable that you could have let AIG and the pure investment banks linked to the mess fail, and needed only a modicum of liquidity support to support payment-system banks which would not have been directly affected other than a temporary panic that is cheap to fix (liquidity can be paid back after the panic subsides). And in the first place there would have been much less reason for a panic without commercial banks having own account exposure to the activities of the non-commercial bank institutions.

    The result would have still been a credit crunch (for the investment banks are in the credit circuit), but with a payment system still functioning with a possibly much smaller scale of support required with taxpayer money.

    Arguably having to resolve more investment banks/insurers all at the same time may have caused an even worse credit crunch, but hard to say how that would balance against a lesser fiscal load in public bailout funds.

    It's certainly not a panacea but still seems good to have, and going in a general good direction towards disentangling credit from payment processing.

    1. I'm actually not arguing against separation of investment banking, necessarily - although you are rather assuming that commercial banks never use wholesale funding themselves so would not suffer in a wholesale bank run or freeze. My conclusion was that we don't actually know what effect removing central bank liquidity support from investment banks would have.

      I agree with you about the need to disentangle credit from payment processing. Indeed that was at least partly the point of this post, and the reason for quoting Matthew Klein. Separating out bits of the "business" of banking is all very well but once they get into the payments systems they reconstitute into one large plate of spaghetti. And it is payments that create the greatest systemic risk - flows of money, not stocks.

    2. What I'm trying to say is that a bank run is different from a bank failure. I'm not disagreeing with your point on needing to help everyone with liquidity when there's a bank run, just saying that providing liquidity with separation is (essentially) free to taxpayers, even that bit that is of use to non-bank investment entities, while before separation you needed net capital injections.

      If you see the system in aggregate and look at the position of the world's taxpayers taken together, you can imagine they face the world's private banks as one consolidated entity, let's call it the World Consolidated Bank. A bank run is then just zero sum reshuffling between subsidiaries of the World Consolidated Bank. If you further aggregate all central banks into one, the liquidity provided to the victims of the run is matched to the penny by the excess liquidity appearing in the sub-entities people run to.

      Basically it's a question of where failing assets are. If they are inside the World Consolidate Bank's balance sheet, it requires taxpayer-funded recapitalisation when enough fail to make the World Consolidated Bank insolvent. If failing assets are outside, failing customers of the world's consolidated bank cannot as such bankrupt it.

      So in so far as separation moves stuff out of the World Consolidated Bank's balance sheet, it's valuable. The smaller the core system is, the better, and every little helps.

    3. This is not even remotely about bank recapitalisation. It is about liquidity. The G-S proposal does not suggest that failing commercial banks should be recapitalised by taxpayers. And no-one, absolutely no-one is arguing that taxpayers should have to recapitalise investment banks. The G-S proposal is about deposit insurance and and central bank liquidity. Basically its supporters want to restrict central bank funding to FDIC-insured institutions only. My point is that because investment banks and shadow banks are customers of FDIC-insured insitutions they are inevitably indirectly supported by central bank funding. When there is a shadow bank run, as we saw after Lehman, the central bank has to provide liquidity because commercial banks are at risk. G-S separation makes no difference to this. That's the point I'm trying to make. I don't see why people find this so difficult to understand.

  7. Frances,

    You say “we don't actually know what effect removing central bank liquidity support from investment banks would have.” Strikes me you answered that yourself in the above article. You said: “….as the Fed discovered when Lehman fell. Remove central bank liquidity support for market trading activities and the entire market collapses like a house of cards.” I agree, as I suspect would most people interested in this subject.

    However, liquidity support (i.e. taxpayer support) CAN BE removed under a Klein/Kotlikoff/Cochrane system. Under that system, if a bank makes big mistakes, it won’t collapse: all that happens is that the value of the stake in the bank held by loss absorbers (e.g. shareholders) will decline.

    Mervyn King actually alluded to the latter point in his Bagehot to Basel speech. He said, “And we saw in 1987 and again in the early 2000s, that a sharp fall in equity values did not cause the same damage as did the banking crisis. Equity markets provide a natural safety valve, and when they suffer sharp falls, economic policy can respond. But when the banking system failed in September 2008, not even massive injections of both liquidity and capital by the state could prevent a devastating collapse of confidence and output around the world.”

    Next, I don’t understand your “spaghetti” point. You seem to be saying (and may have got this wrong) that the fact that investment banks have to make money payments somehow stops the Klein/Kotlikoff/Cochrane proposal achieving it’s objective. I don’t see why.

    The main objective is to force people who want their bank to invest their money to carry the risk in exactly the same way as when such people invest DIRECT (e.g. in a small business or the stock exchange). The fact that the bank division or subsidiary that makes those investments or loans makes the odd cash payment does not prevent the latter objective from being achieved.

    Indeed, the fact that unit trusts (“mutual funds” in the US) make cash payments does not alter the fact that given poor performance by the trust, those holding the trust units lose out. (Kotlikoff actually specifically advocates that investments and loans made by a bank should be handled by a unit trust which is a separate legal entity from the bank itself.)

    1. Ralph, do you understand what "liquidity" is? This is all about money flows, not about stock valuation. Falling equity values have absolutely nothing to do with this.

      When money stops moving, or moves far too much (as in a bank run), the financial system collapses. That's why central bank liquidity was used to support absolutely everything in the financial crisis. It wasn't realistic to distinguish between transactions from different sources - and it still isn't.

      I did not say that Klein's proposal wouldn't work. I just said that there had to be completely separate payments systems too, otherwise central bank liquidity would still be used to support settlement of investment banking transactions.

    2. Frances, I’m well aware that “central bank liquidity was used to support everything” in the recent crisis. I’m also well aware that it “wasn't realistic to distinguish between transactions from different sources..”

      Indeed, that’s the basic problem with the existing system: grandma’s savings get used to bet on dodgy derivatives so to speak. And it’s precisely that problem which Klein, Kotlikoff, etc solve. I.e. I disagree with your claim that that it “still isn't” realistic to distinguish between safe/transaction money, and money which depositors want their bank to lend on or invest.

      Under KK system, money which depositors want to be near 100% safe is near 100% safe. So there’s never a need to provide “liquidity support” for it.

      As to money that depositors want their bank to lend on or invest, government would refuse to provide any liquidity support. And the effect of that refusal wouldn’t be nearly as catastrophic as refusal to support the system in the recent crises, for reasons given by Mervy King. (Though a bit of stimulus for the economy as a whole would doubtless be needed given a spate of daft bank decisions like we’ve seen recently).

      If you still think that it isn't realistic to distinguish between the above two types of deposit, can you provide a DETAILED explanation? Reason is that there are plenty well qualified people who disagree with you and who would be interested to see your explanation. Those well qualified individuals include:

      Klein, Kotlikoff, Cochrane, Positive Money & Co, Richard Werner, the New Economics Foundation, and Benes of Kumhoff of the IMF.

    3. Ralph,

      Oh for goodness' sake. Do you think you could please stop promoting full reserve banking for one second and actually read what I say? And do you think you could resist the urge to patronise me by listing all sorts of people you think know more about this than I do?

      Let me explain AGAIN. This is not about deposit-taking and lending. It is about how payments work - the plumbing, if you like. The fact is that ALL payments, irrespective of their source, go through payments systems, and the gateway to those payments systems in Western economies is commercial banks. Investment banks, shadow banks and non-banks are customers of commercial banks because they need them to gain access to settlement services - which they do by having transaction accounts with commercial banks, just as corporations do in order to pay suppliers and employees and receive payments from customers. Therefore when you provide liquidity to support payment services offered by commercial banks, you indirectly support the activities of investment banks etc. You could split out investment bank transaction accounts on commercial bank balance sheets and say "no central bank liquidity for these" (sort of "nil by mouth", I suppose) but unless you ALSO make that change in payments systems it would make no difference. And problem is that even if you made that distinction between investment banks' transaction accounts and everyone else's, if large payments to/from some commercial banks' customers are allowed to fail it can bring down the commercial banks themselves and indeed the entire payments architecture. That is why investment banks, shadow banks and non-banks were provided with liquidity in the financial crisis. It was to protect commercial banks and payments systems. There is no easy way of resolving this that I can see, and Glass-Steagall certainly doesn't do it. Nor does full reserve banking, which in this context is really something of an irrelevance.

    4. Frances,

      You yourself favour a “payments system” which is a “public utility”. Positive Money, Kotlikoff, etc favour the same thing IN THAT they want money that depositors use for day to day transactions to be backed by deposits at central banks. But PM, K, etc claim that that payments system can be handled by commercial banks. In contrast, you seem to claim the two cannot be disentangled. I support PM & K on that one.

      Perhaps it would help if you explain how and why a commercial bank would go about nicking money from a safe or current account and using it to bolster the investment department of the bank that handles loans and investments? Or in the case of Kotlikoff’s regime where loans and investments are handled by unit trusts that are separate entities from the bank itself, how would the bank nick someone’s “safe” money and put it into a unit trust without the relevant owner of the safe money noticing?

      Banks ALREADY run unit trusts, as you doubtless know. If my bank nicked £500 from my current account and put it into one of its unit trusts without my permission, it’s almost inconceivable I wouldn’t notice. Not the mention the penalties the bank would have to pay for misappropriating my money.

    5. Investment banks put money into "safe" deposit accounts. They use FDIC-insured deposit accounts in commercial banks. In the UK, the deposit insurance limits are much lower, but even here the first £85K of an investment bank's deposit with a commercial bank will be insured. That's what NONE OF YOU seem to understand. Investment banks' money in commercial bank deposit accounts is as insured as retail customers' money.

      I am not talking about money being taken from "safe" deposit accounts and used to fund "risky" activities. I am talking about money from institutions whose activity is not supposed to be subject to central bank support being put into insured bank accounts in order to gain access to payments systems - and therefore de facto being supported by central bank liquidity despite the prohibition.

    6. This comment has been removed by the author.

  8. The only issue worth discussing is the appropriate level of equity buffer held by banks as capital.

    Anything in between is an intellectual exercise of dubious utility.

    The best work by miles dealing with "Banking Regulations" is Anat Admat and Martin Hellwig's book "The Bankers New Clothes".



  9. Frances Coppola

    Hi ,

    Every deposit in the banking system is a claim on the value of a loan in the system and those loans have allready been made. Deposits can’t be made to have risk free value in themselves, the value can only be underwritten with another source of funds

  10. I'm confused. You seem to be implying that because banks hold the deposits of and settle the payments between non-bank financial companies, banks are at risk for everything those non-bank financial companies do. If that's what you mean, it's wrong. Banks are only at risk to the extent they have open exposure to those non-bank financials, mainly through bank credit to them. Bank risk arises from how banks invest the money they take in from deposits and loans. Banks' payment-handling function only becomes risky if other banks are failing; in other words, only if at least one bank has greatly misinvested the money it took in.

    Lehman's failure created a crisis for banks through many channels none of which have much to do Glass-Steagall, nor with banks' payments-handling function. I think the biggest problem was the lack of a resolution mechanism for systemically important non-banks. That meant a hugely complex and achingly slow liquidation through the courts in multiple jurisdictions. As money and collateral were frozen indefinitely pending that process, people began to position themselves for possible further similar freezes.

    There was no tradition of central bank liquidity provision to investment banks prior to Lehman and there's still no consensus on when or how central banks should provide such. The problem has been largely swept away in the US by forcing the last few big independent IBs to become banks. Maybe the most important thing Warren's bill would do is reverse that and clarify that systemically important IBs in crisis should be resolved and not provided with emergency liquidity.

    As for Klein's I think tongue-in-cheek suggestion that if we really want boring banking we should switch to full reserve banking, he's right that would get rid of all risk. Banks could not be brought down by failures to non-banks because no bank would have any credit exposure to any non-bank, or to anybody at all. Banks' only risk would be theft of notes, which they could presumably cover without ever needing deposit insurance or central bank liquidity support. But they would have to charge high fees to depositors to cover costs, which would drive people out of banks and into presumably much riskier alternatives. Indeed somebody would have to fund investment somehow. I wouldn't be surprised if non-bank lenders invented something practically equivalent to fractionally reserved deposits. And back to square one.

    1. You've completely missed the point. Let me explain.

      House A is social housing occupied by a hard-working family on a low income. House B is a mansion occupied by a very rich family. We are happy to subsidise water and sewerage services to House A but not to House B because the family in House B can well afford to pay for them. But the trouble is that House A and House B have shared services. You can't work out which water is House A's and which is House B's, and you don't know which house the contents of the sewer came from. In order to subsidise services to House A but not House B, therefore, you have to split the services - which means major infrastructure changes. Glass-Steagall does not split the services. All it does is build a fence between the two houses, so they look separate - but underneath, they are still fundamentally joined.

      Stop looking at what banks "do". Look at the plumbing.

  11. Frances
    Interesting blog posting.
    You quoted Mr. M. Klein: “separating deposits from lending”.
    This is nothing else than the abolishing of fractional reserve banking.
    And it’s a pretty much Austrian argument.
    Would you really want to ban the fractional reserve banking?
    What about repo and money market funds?
    How would you deal with them? Banning them too?
    Ad) intermediation:
    Actually, since financial markets became deregulated in the 1980s, we have a disintermediation. Large companies are bypassing banks (key words: repo, securitization etc.) Disintermediation makes funding cheaper, but riskier.

    1. I quoted Mr Klein but that doesn't mean I agree with him. I agree with him to the extent that I think transaction banking (payments services) should be separated from everything else and managed as a public utility. But as far as lending is concerned I have no problem with fractional reserve banking, I see no reason to change the present system whereby banks create debt money when they lend and I don't agree with Mr Klein that all lending should be from own equity. Does that help?

      Nevertheless, I think the future is disintermediation. I have written about this elsewhere.

  12. "Deposits that back retail loans on bank balance sheets are at risk. If they are insured, that puts taxpayers' money at risk just as much as if those deposits were used for securities purchase. "

    I don't understand this comment. Lending to the real economy is more productive and less risky then lending to a hedgefund to speculate in the stock market...

    1. Lending to SMEs is about the highest risk form of lending that there is.

  13. This comment has been removed by the author.

  14. In reference to your discussion with Tom.

    1. G-S (if properly drafted) prevents the banks from financing investment banking activity, thus the banks under G-S don't have balance sheet exposure to the IBs. (Yeah, it sucks for IBs that are used to central bank liquidity, but it would just mean that financial market prices would actually begin to reflect investors interests in the assets again.)

    2. You appear to be arguing that payments system exposure is sufficient. Are you saying that when a bank moves $1M from corp A account to corp B account, the bank somehow becomes liable for the $1M? Is this a claim that our payments system isn't sufficiently robust for banks to make sure that $1M is actually in corp A account before transferring, and thus the bank ends up with exposure to corp A because our payments system is deeply flawed? The fact that the checking system is set up to bounce checks before they are paid means that I am surprised the system would allow the kind of flaw you appear to be claiming exists in electronic payments.

    1. 1) Not true. Investment banks have transaction accounts with commercial banks. The balances on those transaction accounts sit on commercial banks' balance sheets as liabilities. Unless you are suggesting that investment banks should have no access to payments systems, then commercial banks will always have exposure to investment banking activity. That would apply with a G-S separation in place. It would also apply if commercial banks were prevented from borrowing wholesale funds, as some have suggested.

      2) All balances on customer deposit and transaction accounts are bank liabilities. When a customer deposits funds in a commercial bank account, that customer is lending funds to the bank. That applies whatever sort of customer is making the deposit - retail, corporate, another bank. So investment banks' transaction account balances form part of commercial bank funding.

      What caused the financial system nearly to fail in 2008 was a massive run on shadow banking, when institutions pulled funds on a simply enormous scale. A bank run is a simply huge drain on commercial bank transaction accounts. Because the balances on the transaction accounts that are drained in a bank run form part of commercial banks' funding, a shadow bank run places the liquidity of commercial banks at risk. That is why the Fed provided liquidity to investment and shadow banks. It was to stop the bank run that was threatening the entire commercial bank/payments network.

  15. 1) I would call IB deposits, the exposure of IBs to banks. Only if banks have assets due from IBs are the banks exposed to the IBs on their balance sheets.

    2) I think I understand your point now. As long as the IBs have deposits, they can be part of a bank run that is destabilizing to the banking sector. (I would not call this a payments system problem.)

    The point of G-S is that as long as the asset side of the bank balance sheet includes no IB liabilities, then the central bank can intervene to stop the bank run without distorting asset prices in the financial system. This is why we have central banks that act as lenders of last resort. As long as banks have conservative lending practices (yes, that's a big if but one everybody understands) then the system can survive the bank run with central banks support. Of course, if the system is rotten to the core and banks take on inappropriate risks, no legislation can save it.

    G-S is not designed to stop bank runs or to eliminate the need for a central bank, but to stop bank finance of investment banking activity.

    1. 1) IB deposits are the exposure of banks to IBs. They are bank liabilities and they can be pulled at a moment's notice like any other sort of demand deposit.

      2) It is absolutely a payments system problem. It is disruption of funds flows, not stocks, that really puts the financial system at risk.

      Commercial banks having loans TO IBs really isn't the issue, apart from the public anger about "banks gambling with our deposits", which Warren played to. In fact that belief is mistaken. The figures show that commercial banks' exposure to IBs was on the liability side, not the asset side - in other words, that IBs were funding commercial banks, not the other way round. And the current regulatory pressure on banks to reduce reliance on wholesale funding supports this. Commercial banks being exposed to IBs on the liability side, through wholesale borrowing and demand deposit accounts, is far more dangerous to the banking system than commercial banks lending to IBs. It is deposits that run, not loans.

      In the financial crisis central bank funding was used to support liquidity in the SHADOW banking system. It is that central bank support of shadow banking that supporters of G-S wish to end. They wish to limit central bank liquidity support to FDIC-insured institutions only, and they think that if they put a strong enough separation in place, a shadow bank run such as happened after Lehman can be allowed to run its course, bringing down investment banks, because it will not affect commercial banks. That is what is fundamentally wrong. A run on shadow banks is by definition also a run on commercial banks. To achieve a separation of that kind would mean that investment banks would have to have their own settlement systems. That's the point I'm making.

      G-S does not do what its supporters want it to do. It does nothing to address the problem of deposit runs, which was the real cause of the near-failure of the financial system in 2008, and it aims to fix something that actually wasn't the problem at all.

      I might add that what commercial banks DID have on the asset side of their balance sheets was not IB liabilities, but supposedly safe securities backed by residential mortgages. These ended up worthless because of collateral value collapse in the sub-prime crisis. Are you suggesting that commercial banks should not be allowed to buy securities that are marketed as safe investments, in case they turn out not to be? G-S would not achieve that.

    2. In 2007-08, there were two principal runs: on repo and and on bank commercial paper (wholesale funding). Only the IBs funded themselves in a significant measure via repo, commercial banks did not (though presumably some of the major post G-S commercial participated in this IB form of finance). Thus, the IBs ran on the IBs in the repo market -- this run had nothing to do with commercial banking, though it had a lot to do with the Fed's rescue efforts and the expansion of central bank support to the IBs.

      The IBs were not significant wholesale funding investors, so the IBs didn't run on the banks. Other parties ran on the banks. I haven't look up the data, but I would imagine that IB deposits with the banking system actually increased in 2008 as the IBs pulled every line of credit they could get their hands on.

      I don't get what IB bank deposits are supposed to have had to do with the bank runs of 2007-08.

    3. Repo is collateralised cash lending. One side is cash, the other is securities. When there is a run on repo, it is cash that is being pulled. And cash moves through commercial bank accounts - because ALL cash moves through commercial bank accounts. If cash is involved, commercial bank accounts are involved.

      I did say it was a shadow bank run. Shadow banks - not just IBs - are customers of commercial banks and have deposits with them. When investors pull funds from shadow banks, those funds are pulled from accounts in commercial banks, ultimately (maybe through several layers).

      I pointed out in the post that a bank run actually only moves money from one commercial bank to another. It is the flows THEMSELVES that cause the damage because of the funding problems they cause for individual commercial banks - as with any bank run. The fact that the run happened at arms length makes no difference if the knock-on effect is that commercial banks have massive funding holes.

    4. "When there is a run on repo, it is cash that is being pulled."

      So you're saying that because Bear Stearns held it's cash with a bank, the March 2008 run on Bear Stearns was actually a run on the bank at which Bear Stearns had its account.

      But, so what. That's why we have a lender of last resort. The central bank supports the bank where Bear has its accounts, and assuming the bank has appropriate security for those account (I believe clearing banks like JP are usually pretty careful about this), the bank ends up with Bear Stearns' assets -- and when the process is complete the bank doesn't need central bank funding anymore.

      That's just how a banking system is supposed to function. (Except for the fact that Bear Stearns wasn't allowed to fail.) You won't convince me that J.P. Morgan Chase was going to collapse if the Fed hadn't stepped in.

    5. JPM wasn't going to collapse. Citigroup was, then likely BofA.

    6. The G-S supporters do not understand how runs on shadow banks affect commercial banks. They think that by imposing separation between the business of shadow banks and the business of commercial banks, central bank funding can be limited to FDIC-insured organisations, and therefore that runs on shadow banks will not require central bank intervention. My point is that as shadow bank cash assets sit in commercial banks in the form of demand deposits that will run when there is a run on shadow banks, refusing liquidity support to shadow banks is actually impossible.

      I totally agree that the reason for the LOLR is to support banks so that runs are not fatal - or better, to guarantee liquidity so that runs don't happen in the first place. But limiting liquidity guarantees to FDIC-insured institutions is pointless. In practice you end up guaranteeing everything, as the Fed was forced to. And in answer to your question - yes, JPM could have collapsed if the Fed had not intervened. After Lehman, the entire financial system imploded and banks were unable to fund themselves.

      There can be no meaningful separation of commercial and investment banking while investment and shadow banks are customers of commercial banks. I really question what this proposed separation is supposed to achieve, since it wouldn't prevent investment banks from lending to commercial banks, and it would do absolutely nothing to mitigate the risk to commercial banks from runs and/or major failures in shadow banking.

    7. My view of what the banking system should be is very different from yours I guess. But I think we agree on one thing: shadow banking (i.e. repos and bank CP) needs to be very carefully regulated (or perhaps eliminated entirely) in order for the financial system to be stabilized. I would argue that we want to implement these measures in addition to G-S. And I don't really understand why detractors of G-S think that in order to be valuable the law must solve all of our financial instability problems at once.

    8. Also, I doubt the JPM would have failed. More likely it would have been the last man standing with ownership of many of the erstwhile financial assets of Citi, Bank of America, et al. That's because of how we've structured claims on financial assets -- a problem which G-S also won't resolve.

    9. This post is simply saying that G-S is not fit for purpose. I have not disclosed here what I think the banking system should look like in the future. But I have elsewhere. Here are some links to give you an indication of where I think we should be going with bank reform. I focus on the UK, but the important issues are similar in the US, really.

      Does that give you a flavour? I admit my thinking has moved on - the Pieria posts are a lot more radical than the older Coppola Comment posts. I have one more post to write for Pieria which will be about digital financing, low-return fund management and the enhanced role of insurance - these three I think will be the key features of the new financial system of the future.

      Regarding the US: I certainly don't think shadow banking should be eliminated. I think it should be brought out of the shadows into the light, so we can see it properly. Then we will have a better understanding of how it works and what the risks really are. And then we will be able to regulate and support it properly. The two go together, by the way - if you support something you need to regulate it, and vice versa.

      It's worth bearing in mind, too, that too much regulation has perverse effects. The very existence of the shadow banking system is at least partly due to people's desire to avoid regulation.

    10. Thank you for an interesting exchange!

  16. The real issue is not glass steagal, this is the fake beast.
    If it passes, then it opens the door to changing the real beast:
    Commodity Futures Modernization Act

    This is the one big banks do not want anybody discuss about.

  17. The purpose of GS is not to refuse liquidiy support to IBs and shadow banks themselves. It is to deny liquidity support to deposits funded from "risky" loans. If an IB puts deposits with an retail bank that makes a "safe" loan then that is OK.
    It would not be possible for an investment bank to use funds from a deposit to fund a loan and then simmultaneously use those same funds again to fund a deposit into a retail bank.

    1. Erm, no, that shows you don't understand how funding works. ALL funding goes through commercial banks. That applies regardless of the nature of the lending activity. It is not possible in the funding process to distinguish between money created by "risky loans" or "safe loans". All money looks the same once it is in a commercial bank insured deposit account.

      Oh, and there is no such thing as a safe loan.

    2. But those deposits are not being used to fund "risky" loans

    3. Indirectly, yes they are. It makes no difference whether the loan in question is what you call "safe" or "risky" (although I don't agree with that distinction anyway - all loans are risky). The money that is used to fund them always goes through commercial banks. There isn't a separate distinguishable pot of money that is used to fund "risky" loans. It's all the same money.

    4. Really the whole concept of deposits "funding" loans is wrong anyway. Deposits are created when loans are made, and those deposits are placed in insured accounts with commercial banks even though the originating loan is what you call "risky".

    5. I don't call loans risky or safe that why I used quotation marks.

      Once the deposit created by the loan is deposited in a second bank is then becomes funded by THAT bank's loans.

    6. Regarding your description of how bank runs are partly due to the financial position of their depositors. At the moment banks risk weight their borrowers in terms of default risk. To get a fuller picture of the bank's position the banks also need to risk weight their depositors in terms of withdrawal risk.

  18. This comment has been removed by the author.

  19. While it's obvious that Glass-Steagall wouldn't have prevented the crisis, to write off the usefulness of such an effort is as silly as claiming that it is irrelevant.

    The intent of the law has already been well explained by Warren, and she has answered all criticism of these points....over a year ago:

    "In my conversation with Ms. Warren she told me that one of the reasons she’s been pushing reinstating Glass-Steagall — even if it wouldn’t have prevented the financial crisis — is that it is an easy issue for the public to understand and “you can build public attention behind.”

    She added that she considers Glass-Steagall more of a symbol of what needs to happen to regulations than the specifics related to the act itself."

    In other words: this is old news, a new G-S would NOT irrelevant, and there's a good case for this being a symbolic effort as part of a larger movement towards more effective reform in banking.

    And everyone knows deposits don't fund loans (for the 10000th time), but deposits are a form of funding for banking operations, and thus there is an indirect effect of "funding risky behavior" when commingling deposits with IB activities.

    It's also patently false that we would have to bail out IB institutions upon separation anyway. It would in fact be much easier to let them merely fizzle out and let the market sort out the remains (Sorry Goldman, don't have liquidity? go suck an egg, let KKR pick your bones for what's worth a damn). Combined with more effective risk management regulatory efforts, G-S could end up being an important part of a new FDR-like foundation for 21st century banking.

    But the bigger point here is that nobody, and I mean nobody, involved with the revival of this bill is claiming that it would have fixed the crisis or that it's an end-all solution. The only people who are saying that are those serving as mouthpieces for the banking sector who want to present the issue as a false dilemma like this is an all-or-nothing deal.

  20. This comment has been removed by the author.

  21. It's me again. I think I've found the flaw in your payments system argument.

    When there was a repo run on Bear, JPM and the other banks had to meet Bears demand for cash. They didn't even need central bank support to do this.

    The demand for cash reduces their reserves. They may need to borrow reserves on the Federal Funds Market to replace those reserves. If the cash leaves the US banking system entirely, that will tend to drive up the Fed Funds Rate -- but also under the current monetary policy regime lead the Fed to increase the reserves to maintain the policy rate. If the money stays in the U.S. banking system the only cost to the banks is that they have to pay the federal funds rate to borrow -- of course they actively manage their interest rate risk so the asset side of their balance pays more than the federal funds rate, and providing cash to meet Bear's demands may cost them some increment of income, but has no destabilizing effect on the banking system whatsoever.

    The reason that Bear's failure put Citi and BoA at risk was precisely because they were post G-S repeal banks that were engaged in IB activity and they were thus exposed to the asset price movements created by Bear's failure (i.e. they were in the repo market and unlike JPM had no idea what they were doing). In short, you are convincing me that it was because G-S was repealed that the shadow banking system was able to destabilize the commercial banking sector.

    1. Umm no. JPM etc. did need central bank support to do that. ALL payments need central bank liquidity. When deposits are pulled, payments are made. Central banks provide liquidity to support payments - it's a fundamental function of central banks. JPM etc. may not have needed EXCEPTIONAL Fed liquidity, if between them they already had enough reserves to support the outflows and they were wiling to lend to each other - although that wasn't the case after Lehman. But to say that they didn't need central bank liquidity is simply wrong.

      I think the flaw in your thinking comes from your notion that cash can "leave" the banking system - i.e. that in some way cash balances held by investment banks are not within the federal reserve banking system. This is wrong. Electronic cash never leaves the federal reserve system. Cash can only leave the federal reserve system as physical notes and coins, and then only until it is re-deposited. All electronic cash movements, whatever their source, occur WITHIN the federal reserve system. The only way in which reserves can be reduced is by Fed open market operations, which as you say drive up the Fed Funds rate. Movements of reserves around the banking system do not reduce the overall level of reserves, only their distribution.

      Liquidity failures (bank runs or market freezes) cause asset prices to collapse, because when banks can't fund themselves by borrowing they are forced to sell assets at fire sale prices to plug their funding gaps. Once again in your focus on the asset side you are not seeing the significance of deposits. It is deposits that run, not loans. Bear hedge funds closure caused asset price collapse because of the run on ABCP, not the other way round: admittedly the value of MBS was already dropping because of rising mortgage defaults (the subprime crisis), but the catastrophic fall in MBS prices that occurred after that closure was due to the run itself.

    2. You write: The banking system did not need extraordinary liquidity "if between them they already had enough reserves to support the outflows and they were wiling to lend to each other - although that wasn't the case after Lehman."

      Basics of modern monetary policy: If the banks don't "have enough reserves between them" the Fed will supply those reserves. You may call this central bank support, but it certainly is not extraordinary central bank support.

      The Libor market broke down after Lehman. The Federal Funds market did not, although there were a few unusually priced transactions on it during this time period.

      "The only way in which reserves can be reduced is by Fed open market operations, which as you say drive up the Fed Funds rate. "

      First, I consider in my discussion both the case of reserves being reduced and the case of them not being reduced.

      Second, this is in accurate. If I take cash out of my bank and put it under my mattress, I have reduced the reserves in the banking system. When I wrote this I was considering the possibility that the transfer of cash from the U.S. to say Germany would have a similar effect, i.e. a transfer of reserves out of the U.S. system and into the German one, but perhaps I'm failing to account for the foreign exchange market correctly.

      "Liquidity failures (bank runs or market freezes) cause asset prices to collapse, because when banks can't fund themselves by borrowing they are forced to sell assets at fire sale prices to plug their funding gaps."

      Well, yes. But in 2008, the asset price failures were not cause by runs on commercial banks, but by runs on investment banks (arguably including some of the post G-S repeal commercial/investment banks).

      I think there is a causality problem in your analysis. The fact that there is an asset price collapse does not imply that there was a run on commercial banks.

      You are using the word "deposits" in a way that unnecessarily confuses the issue. Wholesale funding is arguably comparable to deposits, and for that reason should probably be restricted or eliminated, but we survived the ABCP collapse of 2007 in relatively good shape (yes, with the help of the Fed).

      It was the repo runs of 2008 that almost did us in. And these were runs on investment banks and on investment bank funding, not on commercial bank funding. Using vocabulary that treats repos as "deposits" is misleading and has the effect of camaflouging the key role investment banking, not commercial banking or the payments system, played in the collapse of 2008.

    3. I specifically said that money could be removed from the federal reserve system in the form of physical cash.

      You are still trying to create an artificial distinction between cash deposits created by repo and any other sort of deposit. That distinction does not exist in practice. ALL cash except notes and coins in circulation sits on commercial bank balance sheets. Therefore, if there is a run on cash funding of investment banks, BY DEFINITION there is also a run on commercial banks. As with any run, if it is not too big and the commercial banks themselves will lend to each other it can be accommodated without exceptional Fed liquidity support. But if the run is too big for commercial banks to accommodate, or they stop lending to each other, it is utterly insane then to say "no you can't have any more liquidity because it's caused by investment banks pulling their funds and we don't support those".

      Commercial banks and payments systems are always at the heart of any bank run, whatever the nature of the institutions that are apparently being run upon. Of course, if you closed down payments systems when a bank run started you would stop it in its tracks, but then you would stop everything else too. There is simply no way of continuing to provide normal OR exceptional liquidity support to retail and commercial payment services while denying it to investment banks.

    4. "if the run is too big for commercial banks to accommodate, or they stop lending to each other"

      My point is that the nature of our reserve system is that the run cannot be too big for the commercial banks to accommodate. As you noted, the reserves generally do not leave the system, thus the only problem that can arise is that the commercial banks "stop lending to each other."

      But in the U.S. in 2008, this did not happen. The commercial banks continued to lend to each other and the crisis was caused by the investment banks (yes, including the post G-S repeal commercial/investment banks) ceasing to lend to each other.

      You are attributing the 2008 crisis to events in the commercial banking system that never happened. "Commercial banks and payments systems are always at the heart of any bank run, whatever the nature of the institutions that are apparently being run upon." is an assertion that is not based on the events that actually took place.

      We have a system where runs can take place on investment banks. That is a fact demonstrated by the history of the 2008 crisis. What caused the run is a different question from whether or not at the time it was appropriate policy to protect the investment banking system from collapse. I actually think the policy was right, but the instability arose from investment banking, not commercial banking, and this fact implies that G-S is sensible policy.

    5. It simply is not possible for a cash run to take place on investment banks without affecting commercial banks. As I said, all cash sits on commercial bank balance sheets. No deposits can "run" without involving payments systems, and the gateway to those is via commercial banks. Even if cash is physically withdrawn from the system, that happens via commercial banks. You seem to think that somehow cash deposits managed to run without any impact on commercial banks. That is simply untrue - it was untrue in 2008, too. Commercial banks DID stop lending to each other after Lehman, and the Fed provided extraordinary liquidity support, initially to the commercial banks and to the investment banks/shadow banks that were the source of the run. Runs absolutely can take place on investment banks, but the reason that they endanger the system is because all money transactions go through the same commercial banks and payments systems. I don't see why you find this so difficult to understand.

      However, if you find my explanation insufficient, I suggest you read this speech by Sandra Krieger.

      Your conclusion - that investment banking instablility justifies G-S - is illogical. Instability in the institutions being run upon is no justification for G-S, since G-S does absolutely nothing to prevent the Fed having to provide additional liquidity in the event of such a run. Nor does it do anything to regulate and stabilise the unstable institutions. It throws them to the wolves while failing to address the real issues.

    6. From Sandra Kreiger:

      "In March 2008,. . . it became clear that neither clearing banks, nor overnight cash investors, were well prepared to manage a dealer default. . . . the liquidation of such large amounts of collateral under the extreme market pressures would have created fire sale conditions, large liquidity dislocations and undermined confidence in the whole market. To avoid these adverse systemic consequences, the Federal Reserve stepped in and created a special lender of last resort-like facility to lend to dealers against their tri-party repo collateral. The facility effectively backstopped the market in the immediate circumstances surrounding Bear Stearns’s failure."

      This supports my statement that the run on repo was a run on investment banking "dealers" and that the Fed's liquidity had the goal of saving investment banking, not commercial banking.

      "Runs absolutely can take place on investment banks, but the reason that they endanger the system is because all money transactions go through the same commercial banks and payments systems."

      The reason I find this so difficult to understand is that, as Krieger agrees, the dangers directly caused by the investment banking run are "fire sales." She does not state the repo run destabilized the system via the payments system, as you claim.

      I don't know whether G-S will be enough to prevent such Fed intervention to save investment banks not commercial banks from ever happening again, but I don't see why your theories, which are in fact inconsistent with the facts, should convince us not to try it.

    7. Just want to say: you're a really smart person. Your post on "Financial Dislocation" was brilliant.

      I really appreciate that you have continued this dialog -- I find it extremely useful to understanding, not just your comments, but those that others make about Glass-Steagall. It has also helped clarify my own views.

      Thank you for kindly providing a forum where this discussion can take place. Sites like yours are what give the web value.

    8. My "theories" are actually an accurate description of how the banking system works. I posted Krieger's speech because she makes the deep connections between investment and commercial banks (she calls them clearing banks) very clear. Sadly you have homed in on the bits of her speech that appear to support your mistaken idea of how banking works, and ignored the rest.

      The reason that "fire sales" were so dangerous was not just that they endangered investment banks. They endangered ALL banks, because the assets being sold to meet depositor claims were so widely held as collateral. Such a major fall in asset prices was disastrous for bank balance sheets of all kinds. This is important, but it is not directly the concern of this post.

      However, you've more than slightly missed my point. I did not claim that the repo run destabilised the system "via the payments system". I said that because a repo run takes place through the same set of commercial (clearing) banks and payment systems as retail and commercial transacations, it is not possible to deny liquidity support to investment banks being "run" upon without also denying liquidity for retail and commercial transactions.

      Where you go wrong is in your repeated assertion that 2008 was a run on investment banks "instead of" commercial (clearing) banks. Actually 2008 was a run on investment banks that impacted commercial (clearing) banks in ways too large for them to handle. Krieger used the term "liquidity disclocation", which is perhaps somewhat obscure but in effect means that commercial banks experienced severe funding difficulties as a consequence of the run on shadow banks. The Fed initially provided emergency liquidity support to commercial banks impacted by the run on shadow banking. It only extended support to shadow banks when the collapse of asset prices due to the continuing run threatened the entire banking system. And the aim was to stop the run.

      You will note that Krieger does not recommend G-S or anything like it. She recommends increasing the resilience of investment banks by improving their liquidity and capital buffers. And she suggests that ALL banks, commercial and investment, require central bank liquidity support both in normal times and in crises.

    9. Further to my last....yes, Krieger does say that the Fed supported a primary dealer directly. But it had no choice. Primary dealers are not simply "investment banks", as you suggest. They are, in many ways, agents of the State. They are the principal means by which the Fed transmits monetary policy: they are required to participate in the Fed's open market operations and they are intimately involved in monetary policy decision-making. And they are the main market makers for government debt, participating in primary auctions and reselling government securities to the markets and the general public. No way is the Fed going to allow a primary dealer to fail. The consequences not only for the financial system but also for the economy as a whole are way too great.

      Thanks, by the way. I appreciate the debate.

    10. "a run on investment banks that impacted commercial (clearing) banks in ways too large for them to handle"

      The question for me is whether G-S will reduce the likelihood that this happens again. Certainly alone it will not: we need reforms of both repo and wholesale funding markets too. G-S may give us more choices in the next crisis -- and I'm not sure that more can be asked of financial regulation.

      While I have great respect for the folks at the Fed, they played an important role in getting us here in the first place. I'm not convinced that they will be able to leave their faulty frameworks for understanding the economy and financial system behind and successfully design regulation for the future.

    11. Designing regulation for the future involves understanding how the system actually works and where the risks really are. For me, the case for G-S simply is not made. There are other regulations that in my view would be far more effective in our current system, and we should be focusing on these rather than wasting time and energy resurrecting the ideas of the past.

  22. Its of no consequence to GS if a few IB deposits are protected by providing temprary liquidity support to retail banks. What GS is concirned about is solvency issues of IBs destablising the system permanently, which under GS retail banks would not be permanently affected by.

    1. 2007 and 2008 showed that banks fail primarily due to sudden outflows of liquidity. There may also be solvency issues, but the proximate cause of disorderly collapse is almost always due to lack of liquidity such as that caused by a bank run. Serious liquidity problems impact not only the institution itself but every institution with which it is connected - including the commercial banks which hold its cash assets. G-S does nothing whatsoever to improve the liquidity of investment banks. Therefore as far as the REAL danger to the banking system is concerned, it is useless.

      I would remind you also that because liquidity crises cause asset prices to fall dramatically, a bank can become insolvent as a consequence of a major run even if it was previously healthy.

  23. That is an interesting point about the knock on effect of assets

    but the solvency of commercial banks is not effected by IB deposit withdrwals . its the commercial bank assets that are important . A large number of IB deposits would be a temporary liquidity problem and have central bank support.

    1. Typo - A large number of withdrawls from IB deposit accounts held at commercial banks

    2. Dinero, there is no difference between providing liquidity to a commercial bank that is experiencing funding difficulties due to a run on investment bank deposits, and providing liquidity to the investment banks whose deposits are running. The liquidity provided to the investment banks goes to the commercial banks. It is exactly the same thing in practice.

      Commercial bank assets are important only to the extent that their values are affected by fire sales of the same asset classes by investment banks experiencing runs. What on earth makes you think that a G-S separation would end this sort of contamination? G-S would not stop commercial banks holding securities as part of their capital structure - and it wouldn't stop investment banks doing the same.

      Please don't wheel out the old chestnut about commercial banks having loans to investment banks on their balance sheets. It's not true.

    3. I'm not defending G-S itself. I'm just defining the precise dynamics of finance. The goal of G-S is to let Investment banks fail without letting Retail banks fail. If, as you point out, Retail banks require some liquidity support in the process, than the goal of G-S is not thwarted, so your refutation of G-S based on that particular dynnamic is wrong.
      If you are saying that loans create deposits and so therefore the soundness of the nations currency as a whole can not be free from loan default risk, regardless of how deposits are treated, than I agree with that. I made that point myself earlier.

    4. I'm not saying either of those.

      I am saying that when there is a massive bank run as there was in 2008, the ENTIRE financial system is involved. Trying to distinguish between different types of bank in the provision of liquidity support in a crisis is a fool's game. Even if sufficient liquidity support is available to commercial banks as you suggest, lack of liquidity in other institutions causes asset prices to fall, which threatens the solvency not only of investment banks but of commercial banks. Massive bank runs on the scale of 2008 have to be stopped, and if that means providing liquidity to anything and everything, then that is what the Fed will do. G-S will make absolutely no difference to that.

      More fundamentally, because investment banks are customers of commercial banks, when commercial banks are provided with liquidity support, that is in effect liquidity support of investment banks - just as it is of corporates and retail customers. G-S cannot prevent investment banks placing deposits with commercial banks. Nor should it.

    5. OK i get it, you think that big liquidity squeezes on reatil banks are a big enough problem in themselves. I agree.
      How about this for a solution. Make Basel require banks to risk weight their depositors in terms of withdrawal risk.
      What do you think.

    6. Not sure how it would be achievable in practice. But certainly getting banks to pay more attention to the structure of the liabilities side of their balance sheet would be a good idea.

  24. Might I suggest Modern Monetary Theory could provide an answer for a robot driven world.

  25. Again, g-s won't solve any problems
    This is a collateral based closed loop system.
    In a closed loop, if something clogs in one spot, it affects the whole loop.

    The core problem is with the Commodity Futures Modernization Act that allows shadow banking to create naked leverage infinitely.
    That's why there are runs in the shadow banking system which inevitably affect the commercial system due to plumbing as Coppola beautifully explains.

    There is only one solution:
    - Regulate derivatives as they were before the CFMA was adopted.
    - Define maximum leverage of derivatives

  26. Maggie's ghost: what is haunting Europe

  27. but the solvency of commercial banks is not effected by IB deposit withdrwals . its the commercial bank assets that are important . A large number of IB deposits would be a temporary liquidity problem and have central bank support...!!!

  28. The temporary problem of liquidity but central bank support, there are several other assets for loans are important.


Post a Comment

Popular posts from this blog

WASPI Campaign's legal action is morally wrong


What really happened to Signature Bank NY?