Bank capital is the buffer on a bank’s balance sheet that allows it to absorb losses, particularly credit losses. Although there is a great deal of excitement about bank liquidity — bank runs, just like in “It’s a Wonderful Life”! — but the main danger is the capital buffer being wiped out (insolvency). A bank run might feature at the end of the bank’s lifetime (quite often, regulators just step in), but the trigger is the insolvency. This article discusses bank capital at a high level, from a macroeconomic viewpoint.
Note: This is yet another preliminary draft from my banking manuscript. I may add a description of things like the Basel III bank capital framework, but in order to explain things like that properly, I would have to assume more familiarity with accounting than what I want to target in my book. Meanwhile, the more technical I get, the more likely I make an error, or the text becomes out of date as regulatory frameworks change. Meanwhile, there is the issue that regulations are country specific.
Constraints On Lending Growth (Again)
I will now repeat what I have previously discussed: funding flows (“liquidity”) are not truly a constraint on growth for the financial system, since funding flows are circular. Rather, the constraints are willingness to borrow/lend, and the allied notion of the ability to absorb credit losses. Bank capital rules are a regulatory approach to limit the credit risk to banks.
A bank that is growing rapidly is going to hit its capital constraints fairly quickly. However, the aggregate banking system by definition grows at the average rate for all banks. And what we will see in practice is that realised growth rates of aggregate bank loan books are generally going to be less than the theoretical maximum that capital rules would imply.
The reason why capital rules generally do not act like a hard constraint in practice on economy-wide borrowing is that the banking system can work around the existing constraint faster than creditworthy borrowers tend to show up at their doors. Capital can be increased, and securitisations can get assets off the banks’ balance sheets. Furthermore, non-bank lenders — who are not covered by bank capital rules — will just increase their market share if lending is more “expensive” for banks. (The expected return on equity is invariably much higher than other sources of funding for banks.)
I will return to the question of how this is done after running through the basics of what bank capital is.
Balance Sheets and Equity
In this discussion, we are looking at one set of corporate accounting statements: the balance sheet. This is the value of assets and liabilities frozen at one point of time. (The other main accounting statements are the income statement and the cash flow statement, which tells us about cash/income flows during an accounting period.)
- Assets are all the things a corporation owns. Accountants follow accepted rules to give all of these things a value in the local currency (I will use dollars for simplicity).
- Liabilities are all the things that a corporation owes to other entities (mainly short-term and long-term debt). Since most debts have a dollar value associated with them, they are generally easy to assign a dollar value to. (It is possible that the corporation might owe money, such as the possibility of losing a lawsuit. Those are contingent liabilities, and just appear in notes to the financial statements until the uncertainty about the payments disappears.)
The general hope is that the value of assets is greater than liabilities. The gap between them is theoretically what the owners of the firm would get if they sold all the assets and paid off all the debts. This means that balance sheets follow the relationship:
Assets = Liabilities + Equity.
For a non-financial firm, if the assets were liquidated all at once, the proceeds will typically be far below the carrying value on the balance sheet. Although this might offend some purists who think we should use market values for everything, this is done to make financial statements more useful. For example, factory equipment is valued at their purchase price less a depreciation allowance that reflects the finite expected lifetime of equipment. As long as the factory remains in operation — a going concern — we want to relate the profit stream from the factory to the cost of the equipment (e.g., a return on assets). If we arbitrarily cut the price of factory equipment in half because of a hypothetical fire sale prices (when no fire sale is planned), we just doubled the return on assets. As such, we only care about the liquidation value when the firm is about to be liquidated — typically in a bankruptcy.
Financial assets are claims on cash flows, and thus do not physically depreciate. This means that financial firms whose assets mainly consist of financial assets, the liquidation value is supposed to be relatively close to the balance sheet carrying value. (You could have exceptions such as in the case of fraud.) This means that the equity value is a meaningful approximation of the liquidation value of the firm, and this allows financial firms to operate with higher percentage of their balance sheet as debt.
Bank Capital versus Equity
We can now turn to the question of bank capital. Bank capital is a somewhat expanded notion of “equity.” For a bank, we say that
Assets = Liabilities + Capital,
instead of the more usual Assets = Liabilities + Equity that I was taught in the various introduction to accounting courses that I took.
Capital includes the following parts of the bank’s “capital structure”:
- Common equity, which consists of shares owned by the ultimate owners of the bank. (It might be possible that there are multiple classes of shares, but that might be precluded in some jurisdictions). Assuming things go well, these shareholders are paid dividends out of profits — and there is no upper limit to those payment beyond the profitability of the bank.
- Preferred shares. These are classes of shares that offer a fixed dividend that follows some type of contractual schedule. A typical structure is a perpetual fixed coupon. (The contractual limitation of payments is what distinguishes them from common equity.) The bank has the right to suspend payments on the preferred shares without triggering a default; the preferred shares will have some kind of contractual protection (e.g., dividends cannot be paid to common shares without making the preferred shareholders whole). Preferred shares are considered to be “equity” for any firm, they just are more prominent for banks since they tend to be the major issuers of preferred shares.
- Some subordinated debts managed to be considered a type of bank capital (presumably through the miracle of bank lobbying). For me, the most entertaining type of capital in this group are contingent convertible bonds (“co cos”) — mainly because I think the nickname is a hoot. The idea behind contingent convertibles is that regulators will convert the bonds into common equity in a financial crisis. And since the regulators are taking the action, this means that bond holders cannot try to drag the bank into court contesting the action — they need to go after the central government.
The bank capital rules are complex, and “capital” is divided into tiers. Tier 1 capital is just common equity, whereas the other instruments show up in the other tiers. (E.g., see this explanation from the Bank of International Settlements: https://www.bis.org/fsi/fsisummaries/defcap_b3.htm.) There are differing limits on banks’ balance sheets using these different definitions.
Capital adequacy regulations look at the amount of assets on the bank’s balance sheet versus the various definitions of capital. (These are different than liquidity regulations, and somewhat complementary with leverage rules.) However, not all assets are treated equally. One set of fundamental measures is the ratio of risk-weighted assets to capital.
The process of risk weighting is straightforward: each class of assets is given a risk weighting as a percentage, and the amount of assets held is multiplied by that percentage.
For example, imagine a really small bank had $100 in assets, with
- .$50 in assets with a 100% risk weighting,
- $30 of assets with a 50% risk weighting, and
- $20 of assets with a 0% risk weighting.
The bank’s risk-weighted assets would equal $50×1 + $30×0.5 + $20×0 = $65 of risk-weighted assets.
From the point of view of fixed income investing, risk weighting shows up a lot in discussions of sovereign bonds. It is typical that local sovereigns have a risk-weighting of 0% — which means they do not affect the amount of risk-weighted assets. This property is one good explanation why bond markets shrug off whatever credit rating agencies pontificate about the local sovereign — the 0% weight means that the bonds are exceedingly attractive for banks.
Having assets with a 0% risk weighting eliminates a problem that would appear if all assets were weighted equally. If a bank experienced a rush of inflows, the asset size of its balance sheet will grow in line with the increase in deposits. Since a deposit inflow does not generate a profit, capital would be unchanged. If risk weighting were not used, the bank would have to refuse to take inflow as the ratio of non-weighted assets to capital might breach a limit. However, with risk weighting, it can just buy 0% risk-weighted assets (including “buying deposits at the central bank,” which is a funny way of saying accumulated reserves) and keep its risk-weighted assets constant.
The objective of the capital rules is to force banks to hold a greater percentage of their balance as capital (of varying classes) as the mix of assets on their balance sheets gets riskier — with “riskier” being defined by the bank regulations. One of the problems we ran into in 2008 is that the riskiness of some types of assets ended up being much higher than the risk-weighting suggested.
Although bank capital regulations create an upper limit for an individual bank’s ability to expand its loan book, that upper limit will grow over time. It is the job of management to ensure that capital grows along with the bank.
The first approach is to issue securities.
- So long as capital markets are not disturbed, it is relatively easy to issue new preferred or subordinated bank capital securities. (Note that capital markets can freeze up, typically when all the senior managers in Europe go on vacation and the juniors are manning the trading desks, but it can happen whenever equity investors get unusually loopy. Although certain corners of the internet can get excited about these market closures, the usual outcome is that things calm down within a couple of weeks.) Note that these instruments will help capital ratios that rely on wider definition of capital, but not the Tier 1 definition.
- Banks could issue some additional equity, although that tends to be unusual. Issuing more shares via stock-based compensation or convertible securities is more likely. It is going to be much more difficult to issue equity if a bank is in trouble, which explains why there is no easy way to get away from solvency concerns.
- In Europe, there is a willingness of companies to issue warrants — the right to buy shares at a fixed future price (equivalent to a call option).
The second approach — and the preferred method to grow common equity — is to retain earnings. Profits that are not paid out as dividends end up increasing common equity.
For economies that are featuring steady nominal GDP growth, it would be fairly reasonable for a bank to expect to have a 10% return on equity. Since investors in bank equities are typically interested in dividends (unlike tech investors, who value companies based on whatever stoned Silicon Valley residents think is cool), banks need to pay out a healthy dividend to keep the C-suite’s stock options in the money. So, the bank might have a 50% dividend pay out ratio. The net result for the bank is that it grows its common equity by 5% in a year (50% of the 10% return on equity). This is enough grow its loan book in line with overall GDP (and the banking industry in aggregate cannot grow its loan book faster than GDP indefinitely).
To allow for a faster growth rate, you just retain more earnings. Dropping the dividend pay out ratio to 30% means that the bank would grow its equity by 7%.
However, such a growth rate is not going to be enough for an aggressive CEO. If things are going well, a small bank will have a higher than usual return on equity, which would allow a correspondingly higher organic growth rate. Although many small banks have successfully grown into large ones, fast growth also entails running greater risks. When we look at big exciting bank blow ups, they tend to be the fast moving ones (although some staid old banks can go kablooey, like Barings Bank).
And as a final note, an alternative way to deal with capital regulations is to get the risky assets off the bank balance sheets via securitisations. The bank earns fees from originating and servicing the transactions, and gets the assets off its balance sheet onto the balance sheets of institutional investors.
Email subscription: Go to https://bondeconomics.substack.com/
(c) Brian Romanchuk 2022