"How many times do we have to go down the duration mismatch road with fractional reserve lending and nearly $9 trillion of Fed QE to prove the current banking doesn't work?"
Then Mish points out that the Fed would not do this because the Fed / government prefers inflation and the control it gets from "fixing" the crisis it creates.
I will continue to argue that to be effective in addressing these issues we need to first disentangle the reasons people put money in banks.
People put money in banks because of the facilities banks provide for accepting and disbursing funds. It is very convenient for me to have a middle man who will hold the money my employer gives me as salary, and disburse it in a safe and secure fashion to the people I designate. Even better that now this can all be done electronically. I, and most people, would be willing to pay for this convenience, and we do pay in an opaque fashion in terms of fees charged to sellers by payment processors and bank accounts that get very low to no interest.
People, sometimes the same people, also put money in banks because banks pay for the privilege of accepting funds they can aggregate into larger, hopefully more profitable, investments. In this case I may even accept significant limitations on my ability to withdraw those funds.
For the payment processing function it makes a lot of sense to backstop all or nearly all the funds because they are intended to be liquid, even to the point of accepting a slight negative interest rate on the balances since the funds will be turning over rapidly.
For the investment function however, backstopping all or nearly all of my deposits very definitely creates a moral hazard that I will simply seek out the highest yield offered with little or no regard to how risky the firm is. This behavior then redoubles as it encourages the management of banks that are in trouble to offer higher returns, and thus make even riskier investments to support them.
John Cochrane proposes to separate these functions. Payment processing would be in banks that are required to back deposits with riskless assets. The investment function would be in banks that do not have deposit insurance.
My cynical reaction is that I think government wants to influence the investment function, which probably means that it prefers things the way they are.
It was reported ROKU had roughly $500 million in SVB. I don't know what to make of that. It doesn't seem sensible to handle consumer deposits under the same operation and regulatory structure as a corporate / investment accounts.
It does seem logical that business accounts to handle money flows (payroll, accounts receivable and accounts payable, expenses) should be handled differently than accounts designed to hold money for the purpose of holding money. My understanding is this used to be the case - a checking account and a savings account had different rules for client access. Now a days, it seems savings accounts don't even exist. At least who would bother with a savings account given the interest on the savings accounts is zero - the same as it is with the checking account?
'To make uninsured deposits safe for the depositor would require a complete redesign of banking.'
@ Arnold: That is entirely the point though. 90 years of deposit insurance has meant that depositors don't track their banks' risk, so the risk gets pushed further and further.
Yellen says not all deposits are insured. Only those deposits at banks she and Joe Biden decide are important.
"A bank only gets that treatment," she told U.S. Republican Senator James Lankford, if supermajorities of the boards of the Federal Reserve, the Federal Deposit Insurance Corp and "I, in consultation with the president, determine that the failure to protect uninsured depositors would create systemic risk and significant economic and financial consequences."
(2) A call is made on the bank and it lacks the cash or the ability to borrow to deliver.
Regulation and oversight should prevent cases of #1 from being excessively large. Should, but 2008 was exacerbated by gross mismanagement of financial firms, or at least the faltering economy revealed a lot of firms were not managed to handle a recession.
The issue of a bank not being able to pay current liabilities seems to be a fundamental risk of banking. On the surface this seems a simple problem to address. Simply require banks to always be able to satisfy a call on the bank! Either always hold cash and have lines of credit to satisfy a demand on deposits or have policies that slow the rate deposits can be withdrawn.
It seems to me that banks are expected to handle demands on deposits they actually are not able to literally support. I appreciate the models that say bank clients won't all pull their money at the same time. Yet now we see that social contagion is real. Panic is all too real. So perhaps banks ought to implement throttles that literally prevent large deposits from being wired out on a whim.
But I'm thinking like an engineer and not like a financier.
Your way #2 is precisely the risk Arnold has addressed. Anything that reduces the probability of depositors making a call on their deposits or delays the ability to call will provice incentives for a bank to take more risks. Your suggestion of restrictions on depositors to make calls doesn't reduce the risk of the loans the bank makes. It does reduce the risk of loss to the bannk in the same way that deposit insurance would. Thus moral hazzard occurs either way. As my father woud say "Anything that keeps Uncle Vinny from saying 'Time to pay up.' gives me a better chance of winning big on the ponies."
Which tells me the only way to run a bank that handles revolving accounts is to require all deposits labeled for such purpose to be invested in short duration bills.
You are on the right track. A more general rule might be to have the asset maturities match the liablilities and equity maturities. We might both agree on that. However, others might argue that regulators could never enforce that well and that the current system even with occasional disruptions might cost less. They might also argue that the cost of switching to that system, even if it is regulation effective, would exceed any benefits from reducing disruptions.
Politicians in particular might prefer a less rigorus maturity match. Hard to argue for easier terms for home ownership when the banks are restricted from lending more in that market. Also a desire to lengthen the terms of government debt could be hindered by these restrictions. Even with more bank failures, the politicians can argue, as Joe Biden just did, "We'll get to the bottom of who's at faul and punish them."
Still don’t understand why depositors can spend their days browsing stupid phones and social networks but can’t monitor if their deposits are safe. It is a matter of priorities in life: get a picture on Instagram or be able to get your savings back? Lastly, one of the issues with rating agencies is the alignment of interest....currently more with issuers than investors....because the real client is the one who pays the bill, as always. Until this changes there will always be an inherent conflict.
Arnold, I have spent nearly all of my career lending money to and consulting with independent mortgage bankers and banks who make mortgage loans, and I know you spent time at Freddie Mac. I like the Cochrane proposal to split the governmental support based on functions of banks, but I wonder if there's a more fundamental problem that is affecting the entire system: the federally-insured, 30-year mortgage with no prepayment penalty.
Banks have clear incentives through the capital requirement rules to put their deposit investments into Treasury and mortgage-backed securities (you reference why with your comments about Niall Ferguson's book). SVB followed this approach, which would have looked entirely prudent to the regulators and others, by investing in Treasuries and MBS. However, now many critics are slamming SVB's management for not hedging these investments to match the duration of the MBS, which is where most of the banking industry's unrealized and realized losses occurred, ultimately causing the run on the bank.
But how would banks, some of the largest investors in MBS currently, hedge their MBS? There are no financial hedges large enough to cover banks' investment in MBS and the cost of the ones that are available is so high that they eliminate any benefit of owning MBS. So I wonder if we're not back to the same problem underlying the 1980s S&L crisis: fundamental and unsolvable mismatch on a long-term asset funded by a short-term liability. Banks got burned on MBS then, and in the Financial Crisis, and now, so I fear they'll never buy many MBS again. This would devastate the housing market, which is not what politicians, regulators, businesses, and American society want as evidenced by our multi-year economic incentives and economic distortions for homeownership.
The losses that banks are taking on low-coupon MBS (which is nearly the entire market due to the Fed's ridiculously low rates in response to the pandemic) are directly attributable to the prepayment risk MBS investors face. When MBS investors buy a current coupon MBS, they expect the MBS to have a duration of 7-10 years as the underlying mortgages pay off, primarily due to refinances but also sales of the home or payment in full. This duration bounces around depending on the rates the Fed causes in the market, but the variation is usually manageable by most banks.
However, when the Fed ratcheted up rates so dramatically over the last year or so, finally responding to inflation, what consumer is ever going to refinance his or her low-rate mortgage or buy a home with a mortgage rate that is perceived to be high? The result is that anybody trying to sell MBS to meet a liquidity run (SVB) or any bank holding MBS needing to mark them to market (all other banks and institutional investors) has a much longer duration, low-coupon investment that is upside down relative to market rates, cost of funds, or cost of capital. Hence, they need to recognize enormous unrealized or realized MTM on their MBS in their financial reporting.
And doesn't it all come down to the 30-year fixed rate mortgage, which U.S. regulators are almost alone in the world in promoting? More importantly, loan guidelines from government insurers and GSE purchasers prohibit prepayment penalties in mortgages. Consumers get a free put on their 30-year mortgage so that very few mortgages remain outstanding that long-- unless you have a clueless Fed that for more than two decades encourages low rates, ignores "transitory" inflation, and then is forced to overcorrect with a rapid rate increase that drives up mortgage rates by 3 to 4 percentage points.
I fear that the way the current mortgage guidelines are structured with free puts for mortgagors that cause prepayment and duration risk for banks and other MBS investors will kill off or drive up much higher the cost of mortgages and housing. It's not that we can't redesign the mortgage market to offer shorter-duration mortgages or have explicit or implicit prepayment penalties (the rest of the world does this and U.S. subprime and many U.S. non-conforming mortgages include them now), but the economic disruption is going to be very painful, and this crisis of confidence is banks and the financial system is going to last longer and make for a nasty recession.
Sorry for the long comment, but I'm interested in your and your readers' comments.
"If you take the cosmic justice framing, you end up saying idiotic things, like “the shareholders of SVB lost everything, and executives got fired, so there is no moral hazard issue.” At best, what happened to the shareholders and executives satisfies cosmic justice. But it does nothing to eliminate moral hazard."
That "does nothing to eliminate moral hazard" is not fair - if the shareholders were bailed out and executives stayed with their nice salaries - this would be a bigger moral hazard. So it does *something* to eliminate moral hazard - your arguments only show that it does not *enough*.
If the moral hazard of deposit insurance lies with bank directors and officers, could that hazard be substantially reduced by exposing them to personal liability to the FDIC on a no-fault basis via personal guarantees as a condition of insurance? Currently, I believe bank directors and officers only have liability to the FDIC for fraud and other breaches of fiduciary duty. Allowing the FDIC to pursue the personal assets of officers and directors on a no-fault basis via personal guarantees might be a powerful deterrent to excessive risk-taking.
Riffing on your remark about SVB's depositors deserving to lose a few percent: suppose the FDIC policy was to not insure 100% of deposits but only 100-X percent. Is there a value of X that is simultaneously large enough to give depositors reasonable incentive for prudence and defeat the strategy of offering above market yields backed by government guarantee, and small enough that it still makes people feel that the system is safe, and the FDIC can credibly (from a political economy point of view) commit to actually make people take that X percent loss when it comes to it?
probably would not work. Think about yourself. Do you want to spend time going over your bank's derivative book to see if they are doing a good job of hedging their risks?
Assuming that Next Time The Fed/FDIC Will Get It Right seems much less realistic than just opting to push depositors into full-reserve/narrow banking (which is much simpler to regulate)
I am deeply skeptical that loaning deposits is really the only possible liquidity base for modern finance.
There is the systemic question on bailing out uninsured deposit. Then there are the specifics of any bail-out.
In this case the bond holders at the parent will be the beneficiaries of the bailout as the parent has a multi-billion claim on the bank. So more responsible banks pay a levy. Funds trading the parent bonds get a windfall. Similarly there was a senator the other day questioning as to why the banks in his state were going to pay a levy to make whole the deposits of large tech firms.
The FDIC appears to have retained some powers to modify which deposits benefit from this bail-out, we shall see if they use them.
Mish Shedlock calls for Full Reserve banking
https://mishtalk.com/economics/the-perfect-solution-to-the-banking-crisis-is-to-make-a-truly-safe-bank
"How many times do we have to go down the duration mismatch road with fractional reserve lending and nearly $9 trillion of Fed QE to prove the current banking doesn't work?"
Then Mish points out that the Fed would not do this because the Fed / government prefers inflation and the control it gets from "fixing" the crisis it creates.
Is this the same as 'narrow banking'?
I will continue to argue that to be effective in addressing these issues we need to first disentangle the reasons people put money in banks.
People put money in banks because of the facilities banks provide for accepting and disbursing funds. It is very convenient for me to have a middle man who will hold the money my employer gives me as salary, and disburse it in a safe and secure fashion to the people I designate. Even better that now this can all be done electronically. I, and most people, would be willing to pay for this convenience, and we do pay in an opaque fashion in terms of fees charged to sellers by payment processors and bank accounts that get very low to no interest.
People, sometimes the same people, also put money in banks because banks pay for the privilege of accepting funds they can aggregate into larger, hopefully more profitable, investments. In this case I may even accept significant limitations on my ability to withdraw those funds.
For the payment processing function it makes a lot of sense to backstop all or nearly all the funds because they are intended to be liquid, even to the point of accepting a slight negative interest rate on the balances since the funds will be turning over rapidly.
For the investment function however, backstopping all or nearly all of my deposits very definitely creates a moral hazard that I will simply seek out the highest yield offered with little or no regard to how risky the firm is. This behavior then redoubles as it encourages the management of banks that are in trouble to offer higher returns, and thus make even riskier investments to support them.
John Cochrane proposes to separate these functions. Payment processing would be in banks that are required to back deposits with riskless assets. The investment function would be in banks that do not have deposit insurance.
My cynical reaction is that I think government wants to influence the investment function, which probably means that it prefers things the way they are.
It was reported ROKU had roughly $500 million in SVB. I don't know what to make of that. It doesn't seem sensible to handle consumer deposits under the same operation and regulatory structure as a corporate / investment accounts.
It does seem logical that business accounts to handle money flows (payroll, accounts receivable and accounts payable, expenses) should be handled differently than accounts designed to hold money for the purpose of holding money. My understanding is this used to be the case - a checking account and a savings account had different rules for client access. Now a days, it seems savings accounts don't even exist. At least who would bother with a savings account given the interest on the savings accounts is zero - the same as it is with the checking account?
'To make uninsured deposits safe for the depositor would require a complete redesign of banking.'
@ Arnold: That is entirely the point though. 90 years of deposit insurance has meant that depositors don't track their banks' risk, so the risk gets pushed further and further.
Yellen says not all deposits are insured. Only those deposits at banks she and Joe Biden decide are important.
"A bank only gets that treatment," she told U.S. Republican Senator James Lankford, if supermajorities of the boards of the Federal Reserve, the Federal Deposit Insurance Corp and "I, in consultation with the president, determine that the failure to protect uninsured depositors would create systemic risk and significant economic and financial consequences."
https://www.reuters.com/business/finance/yellen-tells-senators-us-banking-system-remains-sound-2023-03-16/
There are two ways a bank fails
(1) Fraud / gross mismanagement
(2) A call is made on the bank and it lacks the cash or the ability to borrow to deliver.
Regulation and oversight should prevent cases of #1 from being excessively large. Should, but 2008 was exacerbated by gross mismanagement of financial firms, or at least the faltering economy revealed a lot of firms were not managed to handle a recession.
The issue of a bank not being able to pay current liabilities seems to be a fundamental risk of banking. On the surface this seems a simple problem to address. Simply require banks to always be able to satisfy a call on the bank! Either always hold cash and have lines of credit to satisfy a demand on deposits or have policies that slow the rate deposits can be withdrawn.
It seems to me that banks are expected to handle demands on deposits they actually are not able to literally support. I appreciate the models that say bank clients won't all pull their money at the same time. Yet now we see that social contagion is real. Panic is all too real. So perhaps banks ought to implement throttles that literally prevent large deposits from being wired out on a whim.
But I'm thinking like an engineer and not like a financier.
Your way #2 is precisely the risk Arnold has addressed. Anything that reduces the probability of depositors making a call on their deposits or delays the ability to call will provice incentives for a bank to take more risks. Your suggestion of restrictions on depositors to make calls doesn't reduce the risk of the loans the bank makes. It does reduce the risk of loss to the bannk in the same way that deposit insurance would. Thus moral hazzard occurs either way. As my father woud say "Anything that keeps Uncle Vinny from saying 'Time to pay up.' gives me a better chance of winning big on the ponies."
Which tells me the only way to run a bank that handles revolving accounts is to require all deposits labeled for such purpose to be invested in short duration bills.
You are on the right track. A more general rule might be to have the asset maturities match the liablilities and equity maturities. We might both agree on that. However, others might argue that regulators could never enforce that well and that the current system even with occasional disruptions might cost less. They might also argue that the cost of switching to that system, even if it is regulation effective, would exceed any benefits from reducing disruptions.
Politicians in particular might prefer a less rigorus maturity match. Hard to argue for easier terms for home ownership when the banks are restricted from lending more in that market. Also a desire to lengthen the terms of government debt could be hindered by these restrictions. Even with more bank failures, the politicians can argue, as Joe Biden just did, "We'll get to the bottom of who's at faul and punish them."
Still don’t understand why depositors can spend their days browsing stupid phones and social networks but can’t monitor if their deposits are safe. It is a matter of priorities in life: get a picture on Instagram or be able to get your savings back? Lastly, one of the issues with rating agencies is the alignment of interest....currently more with issuers than investors....because the real client is the one who pays the bill, as always. Until this changes there will always be an inherent conflict.
Arnold, I have spent nearly all of my career lending money to and consulting with independent mortgage bankers and banks who make mortgage loans, and I know you spent time at Freddie Mac. I like the Cochrane proposal to split the governmental support based on functions of banks, but I wonder if there's a more fundamental problem that is affecting the entire system: the federally-insured, 30-year mortgage with no prepayment penalty.
Banks have clear incentives through the capital requirement rules to put their deposit investments into Treasury and mortgage-backed securities (you reference why with your comments about Niall Ferguson's book). SVB followed this approach, which would have looked entirely prudent to the regulators and others, by investing in Treasuries and MBS. However, now many critics are slamming SVB's management for not hedging these investments to match the duration of the MBS, which is where most of the banking industry's unrealized and realized losses occurred, ultimately causing the run on the bank.
But how would banks, some of the largest investors in MBS currently, hedge their MBS? There are no financial hedges large enough to cover banks' investment in MBS and the cost of the ones that are available is so high that they eliminate any benefit of owning MBS. So I wonder if we're not back to the same problem underlying the 1980s S&L crisis: fundamental and unsolvable mismatch on a long-term asset funded by a short-term liability. Banks got burned on MBS then, and in the Financial Crisis, and now, so I fear they'll never buy many MBS again. This would devastate the housing market, which is not what politicians, regulators, businesses, and American society want as evidenced by our multi-year economic incentives and economic distortions for homeownership.
The losses that banks are taking on low-coupon MBS (which is nearly the entire market due to the Fed's ridiculously low rates in response to the pandemic) are directly attributable to the prepayment risk MBS investors face. When MBS investors buy a current coupon MBS, they expect the MBS to have a duration of 7-10 years as the underlying mortgages pay off, primarily due to refinances but also sales of the home or payment in full. This duration bounces around depending on the rates the Fed causes in the market, but the variation is usually manageable by most banks.
However, when the Fed ratcheted up rates so dramatically over the last year or so, finally responding to inflation, what consumer is ever going to refinance his or her low-rate mortgage or buy a home with a mortgage rate that is perceived to be high? The result is that anybody trying to sell MBS to meet a liquidity run (SVB) or any bank holding MBS needing to mark them to market (all other banks and institutional investors) has a much longer duration, low-coupon investment that is upside down relative to market rates, cost of funds, or cost of capital. Hence, they need to recognize enormous unrealized or realized MTM on their MBS in their financial reporting.
And doesn't it all come down to the 30-year fixed rate mortgage, which U.S. regulators are almost alone in the world in promoting? More importantly, loan guidelines from government insurers and GSE purchasers prohibit prepayment penalties in mortgages. Consumers get a free put on their 30-year mortgage so that very few mortgages remain outstanding that long-- unless you have a clueless Fed that for more than two decades encourages low rates, ignores "transitory" inflation, and then is forced to overcorrect with a rapid rate increase that drives up mortgage rates by 3 to 4 percentage points.
I fear that the way the current mortgage guidelines are structured with free puts for mortgagors that cause prepayment and duration risk for banks and other MBS investors will kill off or drive up much higher the cost of mortgages and housing. It's not that we can't redesign the mortgage market to offer shorter-duration mortgages or have explicit or implicit prepayment penalties (the rest of the world does this and U.S. subprime and many U.S. non-conforming mortgages include them now), but the economic disruption is going to be very painful, and this crisis of confidence is banks and the financial system is going to last longer and make for a nasty recession.
Sorry for the long comment, but I'm interested in your and your readers' comments.
Just a steelmanning nitpick:
"If you take the cosmic justice framing, you end up saying idiotic things, like “the shareholders of SVB lost everything, and executives got fired, so there is no moral hazard issue.” At best, what happened to the shareholders and executives satisfies cosmic justice. But it does nothing to eliminate moral hazard."
That "does nothing to eliminate moral hazard" is not fair - if the shareholders were bailed out and executives stayed with their nice salaries - this would be a bigger moral hazard. So it does *something* to eliminate moral hazard - your arguments only show that it does not *enough*.
If the moral hazard of deposit insurance lies with bank directors and officers, could that hazard be substantially reduced by exposing them to personal liability to the FDIC on a no-fault basis via personal guarantees as a condition of insurance? Currently, I believe bank directors and officers only have liability to the FDIC for fraud and other breaches of fiduciary duty. Allowing the FDIC to pursue the personal assets of officers and directors on a no-fault basis via personal guarantees might be a powerful deterrent to excessive risk-taking.
Cochrane persuaded me.
I agree with what you said here about what shouldn't be done but I don't see what you've previously written about what to do as a better option.
What are you're thoughts on big banks bailing out Republic?
"Mark to market" should be made mandatory.
Riffing on your remark about SVB's depositors deserving to lose a few percent: suppose the FDIC policy was to not insure 100% of deposits but only 100-X percent. Is there a value of X that is simultaneously large enough to give depositors reasonable incentive for prudence and defeat the strategy of offering above market yields backed by government guarantee, and small enough that it still makes people feel that the system is safe, and the FDIC can credibly (from a political economy point of view) commit to actually make people take that X percent loss when it comes to it?
probably would not work. Think about yourself. Do you want to spend time going over your bank's derivative book to see if they are doing a good job of hedging their risks?
Assuming that Next Time The Fed/FDIC Will Get It Right seems much less realistic than just opting to push depositors into full-reserve/narrow banking (which is much simpler to regulate)
I am deeply skeptical that loaning deposits is really the only possible liquidity base for modern finance.
There is the systemic question on bailing out uninsured deposit. Then there are the specifics of any bail-out.
In this case the bond holders at the parent will be the beneficiaries of the bailout as the parent has a multi-billion claim on the bank. So more responsible banks pay a levy. Funds trading the parent bonds get a windfall. Similarly there was a senator the other day questioning as to why the banks in his state were going to pay a levy to make whole the deposits of large tech firms.
The FDIC appears to have retained some powers to modify which deposits benefit from this bail-out, we shall see if they use them.