Tuesday, July 21, 2015

A Capital Fed Ruling

The Fed just released it's latest missive to the big banks, and the answer is capital, lots more capital.

Three cheers for the Fed.

They are increasingly understanding that no matter how much they try to micromanage asset decisions, it's impossible to regulate away risk from the top. And "liquidity" will vanish the minute it's needed. Joke version -- liquidity standards are like requiring everyone on an airplane to carry a thousand bucks, so they can buy a parachute if the engines blow up. Just who will be buying "liquid" assets in the next crash?

So,  just raise capital, lots more capital, and slowly let the rest fade away.

A minor complaint: The Fed did it right but said it  wrong.
..under the rule, a firm that is identified as a global systemically important bank holding company, or GSIB, will have to hold additional capital...
No, capital is not "held." Capital is issued. Capital is a source of funds, not a use of funds. Capital is not reserves.  Please all, stop using the word "hold" for capital.
"A key purpose of the capital surcharge is to require the firms themselves to bear the costs that their failure would impose on others," Chair Janet L. Yellen said. "In practice, this final rule will confront these firms with a choice: they must either hold substantially more capital, reducing the likelihood that they will fail, or else they must shrink their systemic footprint, reducing the harm that their failure would do to our financial system. Either outcome would enhance financial stability."
Issuing (not holding!) more capital does not make firms "bear costs." Firms never bear costs. They pass costs on to customers, workers, shareholders, or (especially for banks!) the government.  The slight argument for higher "costs" is that equity gets to leverage with less subsidized too-big-to-fail debt; that's not a cost, that's a reduction in subsidy. If (if) the cost of equity capital is high by some MM failure, then equity receives higher returns and borrowers pay higher costs. This is a surprising quote. Ms. Yellen is usually accurate in such matters.

But that's a minor complaint. I'd rather they raise capital and explain it wrong rather than the other way around. And of course, I'd rather they keep going. I'm also a skeptic that big banks are "systemic" and little banks are not, and thus should be allowed to continue with sky high leverage. But we'll get there.


A reader asks why I'm so persnickety about language. In this case, it's important. I think everyone recognizes that more capital leads to more financial stability. When an equity-financed bank loses money, share prices decline, but there are no failures or freezes. However, if you think capital is "held," and it "costly," then you think that banks shifting to issuing equity or retaining dividends to obtain funds has a cost to the economy, and regulators should require as little capital as possible. If you recognize that capital is issued, does not tie up funds, does not reduce the amount available for lending, then your mind is open to obtaining financial stability with lots and lots more capital.


  1. John,

    Can you explain why using more precise language in this case matters? Granted, standard usage is technically incorrect and annoying. But ... is that all?

    1. The language is very important here. If regulators simply said that banks need to raise more equity (as opposed to debt) to finance themselves, then things would be clear. Instead, bank apologists, lobbyists, etc. deliberately use the term "bank capital," which sounds like cash on the balance sheet (it isn't). Anywhere else, bank capital is known as equity.

    2. Anonymous,

      I think the reason for the imprecise language is because a bank can obtain capital by retaining it's earnings as well as from issuing equity.

    3. Yes, saving cash increases equity. Equity = Assets - Liabilities. Cash is an asset. Saving your profits as cash increases your assets and your equity. Capital and equity are the same thing.

    4. Anonymous,

      Yes, I know. But consider John's statement:

      "No, capital is not held. Capital is issued."

      In the case of cash savings, capital is not issued though those cash savings constitute a portion of the equity of a company.

    5. In the case of cash savings deposited in the bank you have a liability from the banks point of view - not an asset. This is the opposite of capital. It is a callable loan to the bank. In the event the bank's investments fail and jeopardize the solvency of the bank, these deposits will still have to be paid back. What Professor Cochrane is saying is the bank should not just rely on deposits. A significant portion (i.e. 20%) of the bank's funding should be raised by selling stock in the bank. These shares of bank stock are issued - not held. If bank investments fail, you have a 20% cushion before deposit insurance becomes necessary.

    6. Ben,

      Not cash savings in the sense of depositors, cash savings in the sense of profits retained by the bank.

      Profits obtained by a bank that are not returned to share holders in the form of dividends and are not used to buy back shares are equity held by the bank.

  2. Here’s a neat little truism: if all forms of state backing and subsidy for private banks are removed (which they should be) then private banks are then funded just by equity. I.e. banks’ capital ratio becomes 100%.

    Anyone who thinks they have a deposit at such a bank is deluded: since, they stand to lose their deposit, there deposit is actually a share in the bank. To be more accurate, it’s a sort of preference share.

    1. Just because debt and equity share a risk of loss does not mean that they are the same. Nor does the ability of a clever draftsman to manufacture preferred shares having most of the characteristics of debt. Regulators do not count such instruments as equity.

      Most importantly equity has no right to repayment before the completion of the liquidation and winding up of the issuer. Indeed, the repayment of the share purchase price to an equity holder before that time would be a breach of fiduciary duty by the directors unless they could prove that the original purchase price is the fair market value of the shares.

      Further, equity may not be repurchased by the issuer if doing so would impair the issuer's capital. Repayment of a debt cannot impair capital.

      Debt has many other rights over equity, such as the right to recover fraudulent transfers, and the absolute preference in liquidation.

    2. Fat Man,

      One other distinction between debt and equity concerns redemption versus market value. With equity, they are the same thing. With debt, they are not.

      Any company that tries to redeem equity must purchase it back at the going market price. If company sells 1,000 shares of equity for $10 a share and the market value of those share rises to $20 per share, that company must pay $20,000 to buy back it's equity.

      A company that retires it's debt need only pay the redemption (not market value) of the debt. For instance if the same company issues $20,000 worth of 5 year bonds with a 5% interest rate and the market value of those bonds rises to $30,000, the company need only pay a maximum of $20,000 plus 5 years of 5% interest ($20,000 * 1.25 = $20,500 using simple interest) to redeem the debt.

      With callable debt, a pre-payment arrangement is reached where the issuer does not pay the full interest.

    3. Should be $20,000 * 1.25 = $25,000

  3. Hi

    You are scrutinizing wheher the appropriate word is to hold or to issue for describing capital.

    How are you defining capital here.

  4. I think Dr. Cochrane is pointing out that if we "hold" money (such as reserves), that is a balance sheet asset. If we "issue" capital that is a liability (equity capital is on the liability side of the balance sheet). He is saying that the Fed obviously knows financial accouting and it therefore should be more precise in its language.

    1. This is really simple (although made deliberately confusing by banks):

      "bank capital" = equity (i.e. issuing common/preferred stock and retained earnings).

      That's it. I think the term "bank capital stock" comes from early accounting days. Nevertheless, it simply means equity. Firms finance themselves with a mix of liabilities and equity (currently, that's about 90% / 10% of assets for banks). Regulators and others (Anat Admati) want that to be more like 70% / 30% (or 0% / 100% for John Cochrane).

    2. Anonymous,

      Firms finance themselves with a mix of liabilities (debt), equity, and retained earnings.

      There are many small companies (mom and pop) that finance themselves entirely from their earnings. They took on debt to get started, paid off the debt, and now run as a simple cash flow enterprise. Often they are augmented by regular paying jobs.

      And so is the Fed telling banks they must retain a greater portion of their earnings as a rainy day fund? Fewer stock buybacks, fewer / smaller dividend payments, etc.?

      Reserves are a separate issue. Money held by a bank as reserves does not come from it's profits. It comes from deposits.

    3. "And so is the Fed telling banks they must retain a greater portion of their earnings as a rainy day fund? Fewer stock buybacks, fewer / smaller dividend payments, etc.?"

      Yes. Instead of financing assets (cash, business loans, mortgages, etc.) with debt, the Fed wants banks to use equity -- either by issuing stock or cutting dividends and thereby increasing retained earnings. More equity finance, less debt finance.

  5. It's fine that MORE capital to be issued/held by Big Banks is the right, "market regulation".

    If the bank issues shares to increase its capital, it receives cash -- and the cash "capital injection" is then kept, held by the bank, and NOT lent out.

    What are the new levels? Probably not even close to the 50% that I think all recipients of TARP funds should be required to hold (after issue). The capital reserve requirement is the issue here, and such capital is held (already issued) in reserve to buffer shocks to the system.

  6. From the link:
    >>Under the final rule and using the most recent available data, estimated surcharges for the eight GSIBs range from 1.0 to 4.5 percent of each firm's total risk-weighted assets
    Better direction, but not very far in terms of distance.

  7. Increasing capital element of a balance sheet does "tie up" funds.
    If capital requirement is doubled then a given amount of capital that was supporting two balance sheets of size 2x is now "tied up" on one balance sheet of size 1x.

    1. In what sense is money supplied by shareholders “tied up” any more than money supplied by depositors or bondholders “tied up”. When a bank (or indeed any other type of corporation) obtains funds from shareholders / bondholders etc, the corporation does something with that money: in the case of a bank, it lends to it mortgagors for example.

    2. Because the deposits are new money that did not exist before , the funds from Shareholders or Bondholders is previously existing money that did exist before. If Bank A has a ratio of 10 and raises 10 dollars in capital is can create 100 dollars in deposits. If Bank B has a ratio of 5 and it raises 10 dollars it can create 50 dollars.

    3. Ralph,

      Think fractional reserve lending. When a bank lends out deposits (up to a maximum determined by it's required reserves), it is effectively creating another set of deposits. For instance Joe has $1000 deposited at Bank A. Bank A lends $900 (10% held in reserve) to Sue. Sue uses that $900 to buy a new bike from XYZ Biking Company. XYZ Biking Company deposits that $900 at Bank B. Now there are $1,900 in deposits instead of the original $1,000.

      When a bank sells bonds or equity, there is not a double set of deposits and hence no "new money" created.

  8. "I think everyone recognizes that more capital leads to more financial stability."

    That's clearly true if everything else stays the same. But it's an open question whether the financial system will be more stable after it fully adapts to the regulations.

    In other words, it's easier to make one part of the system safer than to make the system as a whole safer.

    [On terminology, I remember reading a Fed paper that talked about reserves "funding" banks...of course the only bank funded by reserves is the Fed]

  9. I agree, the language is wishy washy. If I knew nothing about this and just read the headline title, I would assume the Fed was just requiring banks to raise the amount of reserves they had to hold in the system rather than issuing equity.

  10. I continue to pump for my proposal to make the 50 biggest BHCs pay hold additional capital, and to require that they issue an equal amount of tradeable subordinated debt. The market in that debt and is associated default swaps could act as the canary in the coal mine for the issuer.

  11. KISS in government and related langauge.
    How about, "If a bank desires FDIC deposit insurance, it must issue equity no less than 50 percent of loans outstanding."

    Thatt said, I am not sure if a Fed bailout is the end of the world. If the Fed prints up a lot of money at the bottom of the recession (better if before), it likely will be a healthy thing to do.

    1. Ben,

      "...it must issue equity no less than 50 percent of loans outstanding..."

      Fine idea except the markets may decide that a bank's equity is only worth 20% of it's loans outstanding.

      If we rely on markets to price a bank's equity and we allow markets to determine the value of a bank's outstanding loans, then there is no guarantee that markets will decide that the total value of a bank's outstanding equity is worth 50 percent of the loans that it holds.

      Sure, a bank can have $1 million worth of loans and sell $500,000 worth of equity. But the value of those loans and the value of that equity will change over time. And they will not necessarily move in lockstep.

    2. Frank-- I was wondering the same thing. Publicly traded equity will fluctuate in value. Hmmm.

    3. Frank,

      Why would it matter if markets decide that a bank’s loans are only worth half their book value? In the simple case where a bank is funded just by equity, the value of that equity would also halve.

      The beauty of that is the bank does not go insolvent. In contrast, the slightest drop in the value of loans in a highly levered bank means it is technically insolvent, if not actually insolvent (think Northern Rock).

    4. Ralph,

      "In the simple case where a bank is funded just by equity, the value of that equity would also halve."

      Not necessarily. In determining the value of a bank's equity, you need to look at more than just the value of it's loans under control.

      There are things like name recognition, reputation, and prestige (JP Morgan versus Bob's Savings and Loan) and intellectual property held by the company (patents, trademarks, etc.) that help to determine the value of a company's equity.

      And since banks can't borrow to fund the construction of a new office building / make any capital equipment improvements, then logically the liquidation value of those physical assets becomes equity owner property.

    5. Ralph,

      When interest rates are rising and all of the existing loans held by a bank are performing, the market value of those loans held by a bank will fall, while the market value of the bank's equity may remain unchanged.

    6. Ralph,

      "Why would it matter if markets decide that a bank’s loans are only worth half their book value?"

      What you should realize is that real time 100% equity financing for bank lending is literally impossible.

  12. Increases to cash savings are a use of funds. Any increase on the balance sheet is a use of funds. This is basic financial accounting.

    Assume a bank makes a profit. It can pay a dividend or retain the profit as cash. If the bank decides not to issue the dividend and instead keeps more cash, you say it issued equity?

  13. All is credit except gold! R'mbr money is somebody's promise, somewhere.

  14. At a higher capital ratio the profit from a loan is spread over a greater amount of shareholder funds. If at a higher capital ratio, a prospective loan will garner less profit per shareholder contribution than is required to attract shareholder contributions then the loan won't be made and so a higher capital ratio does decrease lending and or increase the cost of it.

    1. You've evidently not heard of Merton Miller and Franco Modigliani. They got economics Nobel Prizes for amongst other things debunking the above sort of idea. Criticisms have been made of MM, but the criticisms are pathetic, far as I can see.


    2. In Miller and Modligliani, for a firm , the shareholder contributon goes up and extra assets are aquired in line with the extra funding , and the profit also goes up. So the return per contribution is the same.
      But in raising Bank Capital ratios the shareholder contrbutiong goes up and the same assets are aquired as before , the extra funding, in line with the hgher ratio, does not result in extra assets being aquired, and so the profit remains the same. So the return per contributon goes down. So Miller and Modligiliani is not applicable for the Bank Capital ratio analysis.

    3. Ralph,

      Miller Modigliani states that the value of a firm is independent of the financing means selected (debt vs. equity). The key term here is "value of a firm". Dinero is commenting on the profitability of a firm. The value of a firm is not exactly the same thing as it's profitability.

      Also, see:

      Redemption rules invalidate Modigliani-Miller.

      Suppose the shareholders in a company say they have had enough and demand that the company CFO buy back all outstanding shares in the company from them using profits from future quarters. The company must pay the market value of those shares to redeem them.

      Suppose the bondholders of a company say they have had enough and demand that the company CFO retire all existing debt using profits from future quarters. Here, the bondholders are not entitled to market price for the bonds. They are only entitled to the previously agreed upon interest and principle payments.

      Miller Modigliani assumes that the only buyer for a company's debt / equity is someone / some entity other than the company itself.

    4. Frank,

      I don't think you have to go into contracting details to recognize that MM is breaks down in reality. The most proximate issue is always the value of tax shields associated with debt interest tax deductions. I would expect the value of these to dwarf the value of security legal details.

      Also, I'm not sure why the company has to pay the market price for shares (or at least not the market price as represented by the exchange price); they can always execute a private transaction negotiated with a large shareholder.

    5. Anonymous,

      "Also, I'm not sure why the company has to pay the market price for shares (or at least not the market price as represented by the exchange price); they can always execute a private transaction negotiated with a large shareholder."

      It's more a price ceiling that the company can expect to pay to retire it's own security issuances. A company cannot be expected to pay more than the going market price for it's outstanding equity shares. Likewise the price ceiling for a company to buy back it's outstanding debt is not the market value of that debt, it is the previously agreed upon terms of the debt (principle and interest).

  15. As an economist turned banker turned regulator, I think we're talking past each other. True, capital is issued, not held. But unlike most industries, capital is not a primary source of funding for loans and investments. Instead, it finances risk taking by providing a buffer to absorb unanticipated losses. Considered this way, 50% capital levels make sense only in the riskiest firms, e.g., venture capital.

    1. You seem to be making the popular claim that the risk of banks failing should be reduced to reasonably low levels, which means the chance of taxpayers having to bail them out is reasonably low. My answer to that is that taxpayers (quite rightly) make absolutely no promise whatever to bail out butchers, bakers or candlestick makers, so why should there be the remotest chance of their having to bail out money lenders?

      The promise by an entity that lends to repay depositors’ money is fraudulent because those entities lend on depositors’ money and there is no such thing as a totally safe set of loans or investments. Ergo money lenders should be funded just by equity. Or as I think John Cochrane has put it, “runnable liabilities” should be completely removed from banks’ balance sheets.

      As to sums which depositors want to be totally safe, those should be lodged in a totally safe manner, e.g. just lodged at the central bank and/or invested in short term government debt.

    2. Ralph,

      My name is J. P. Morgan. I just purchased the U. S. Steel Company from Andrew Carnegie for over $200 million. Is the U. S. Steel Company a bank? Can it borrow money under your proposal?

  16. But I don't care how much capital a bank "issues". I care how much capital they carry on their books as the difference between assets and liabilities. That capital cushion might represent common or preferred shares that have been issued, sub-ordinated debt or retained earnings. It might even include unrealized capital gains on real estate holdings.

    1. Absalon,

      It would have made more sense it the Fed would have come out and said that banks must lower their debt to equity ratios. Everyone knows what debt is and everyone knows what equity (including retained earnings, shares outstanding, and real estate holdings) is.

      The problem is determining a bank's capital adequacy. How much in outstanding liabilities (debt and equity) should a bank carry for a given loan portfolio. A truly run free bank makes loans solely from it's retained earnings.

      Introducing either debt or equity sold by a bank places a market claim on the value of that loan portfolio. This becomes problematic when the value of a bank's combined debt and equity fluctuate independently of the portfolio. And so a bank can become undercapitalized when the market value of it's liabilities (debt and equity) take a hit relative to the market value of it's loan portfolio.

      I think the Fed realizes this. Both debt / equity ratios and capital adequacy are important issues in banking regulation. The Fed chose to address the latter (capital adequacy) without mentioning the former (debt / equity ratio).

  17. Hello John,

    Just to double check.

    What if a bank has usd100 in loans and usd30 in equity (simple example with 100% risk weighted loans) and wants to create a new loan/deposit of +usd10. It will need to raise at least some equity -- eg usd5 -- to maintain the 30% equity/assets.

    So in the margin, to expand lending the bank needs more equity.

    But in a world where the central bank doesn't come to the rescue --no subsidized too-big-to-fail-- people just have to accept that investing in a bank is risky.

    Is this a good enough compromise or you see 100% equity bank as the optimum -- eg retail deposits at the central bank or, out of bounds for banks?

    I assume that if banks can´t get more equity to expand lending then it´s a sign that the extra lending is too risky-- hence it´s ok if it doesn't happen eg bad projects. So no loss for society. Is this correct?

    Fantastic blog!

  18. I think Yellen's point is that she wants banks to shift from debt financing to equity financing because debt imposes a macroeconomic externality in the form of higher probability of crisis.

    Banks do not properly bear these risks, because they do not take into account the macro externalities, and because they expect bailouts. Yellen is concerned about banks "bearing" the costs in the sense that they incorporate the full costs of the marginal loan into their decisions. The language makes sense because presumably losses to owners = equity holders enter into bank decision-making, while losses to customers, tax-payers, bond-holders, and the general public (through macro externalities) do not.

    It makes sense to think of banks as "holding more capital", as the converse of "issuing less debt", if you think of capital as a residual A - D (assets minus debt). Also, I thought it was generally accepted that issuing equity is costly because of informational problems (which is why we have debt contracts at all, rather than fully contingent claims).

    1. Jonathan,

      "It makes sense to think of banks as holding more capital, as the converse of issuing less debt, if you think of capital as a residual A - D (assets minus debt)."

      That is an unusual definition of capital. When looking at a company's balance sheet it has assets and liabilities. Debt and shares outstanding are treated as liabilities of the company. Buildings, intellectual property, retained earnings, and in the case of banking loan portfolios are treated as assets.

      Capital can be broken into three forms - physical capital (buildings, machinery, etc.) that are owned by a company, cash obtained from the sale of liabilities (debt / equity issuance), and cash from retained earnings. The shares / bonds that a company sells are not capital, the cash that is obtained from the sale of shares / bonds is capital.

      So when Janet Yellen insists that banks "hold more capital" does that mean that banks should reduce their reliance on debt AND equity issuance and instead operate from retained earnings only?


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.