Revisiting the Lehman Brothers Collapse

The Business of Banking and its Inherent Crises

In order to understand our economy it is necessary to take a critical, nononsense look at banking. It is a disruptive force that tends to induce and amplify instability even as it is an essential factor if investment and economic growth are to be financed.”

— Hyman Minsky in Stabilizing an Unstable Economy (1986)


How does a 158-year old U.S. investment bank with billions of dollars in assets suddenly collapse? The lines above from Hyman Minsky contain a clue.

The failure of Lehman Brothers at the end of the summer of 2008 and the financial turmoil that ensued is often styled a Minsky-moment. It revealed the inherent instabilities of the financial system while the subsequent bailouts of numerous other banks and financial institutions demonstrated the U.S. government’s regard for the banking system as an essential factor for financing investment and economic growth, among other things. Following Minsky’s advice, we must take a nononsense look at banking, which will lead us to an examination of the business model of banking and its inherent crises.

First, a Word on Bailouts

If you’re under the impression that banks create money, as I’ve demonstrated (here), then you might feel a bit confused. I made the claim that banks create money each and every time they extend a loan by simultaneously creating a loan asset and a deposit liability on their balance sheets. (If you want a quick recap of the idea, you can watch former IMF economist Michael Kumhof below).

Source: YouTube

So here’s the problem: if banks create money, why would they ever need a bailout? The answer of course can be found in something else I’ve previously discussed — the hierarchy of money.

The money that banks create comes, as I just mentioned, in the form of deposits, which are liabilities of the bank that issues them. Deposits are just IOUs, mind you IOUs that can be used to make payments and settle debts, which is why they are considered money. As IOUs, they are promises to pay. Promises to pay what? You might be wondering. They are promises to pay cash, which is money issued by the government. Cash is king.

That private bank-issued money is lower on the monetary hierarchy than government-issued money implies that banks face two separate, although related, types of potential crises — that of liquidity and that of solvency. We’re going to take a look at each, and in the process I will treat a question that I left open at the end of my last post, which is, what is it that actually does constrain the creation of money by banks?

Liquidity and Solvency

Both liquidity and solvency are related to one’s ability to pay one’s debts. Distinguishing between the two, however, is not so straightforward, as LSE Professor Charles Goodhart writes, “Liquidity and solvency are the heavenly twins of banking, frequently indistinguishable. An illiquid bank can rapidly become insolvent, and an insolvent bank illiquid.” Let’s see what he means.

Liquidity is all about being able to access cash when it’s needed. If you can settle your current obligations with ease, you’ve got liquidity. If you’ve got debts coming due and you don’t have the cash to settle them, then you’ve got a liquidity crisis. To be sure, a deposit account is about as good as cash in this context because it can be redeemed for cash on demand, or can be used to settle a wide variety of payments without the need to first convert to cash. For households, cash and deposits are about as liquid as assets come.

Assuming you’re running low on either one of those liquid assets, you could try selling some of your valuables, such as your car, maybe your fancy watch, or maybe even your old baseball card collection. The problem is that these items are not easy to sell in a pinch (you might be able to sell them quick but you may have to offer a ridiculously low price to do it), which is precisely what makes these valuables illiquid compared to cash.

However, the fact that you have valuables that you could sell suggests that you might still be solvent, which means that you own more than you owe. In other words, the total value of your assets is equal to or greater than the total value of your liabilities. The liquidity problem that you have is that the assets that you own are not in the form of cash, and may take time before they can be converted into cash. But your debts are due now.

To understand what this means for banks, it helps to think about the very nature of the banking business model; that is, what it is that banks do in order to make profits. Remember, banks don’t create money just to be nice by helping people and businesses fund their various endeavours; they do it for the same reason any business provides a product or service — for profits.

The basic business model of banking is to issue short-term liabilities in order to acquire long-term assets. The primary liabilities banks issue are deposits, which are usually redeemable on demand and normally pay out a small rate of interest to deposit-holders. The assets acquired by banks are loans, such as car loans, credit card loans, or mortgages, which are paid back over time, but which earn an interest rate greater than that paid on deposits. This difference between the rates that banks charge on the loans they make and the rates they pay on deposits is the primary source of bank profits.

Note, however, the time difference between which deposits may be redeemed and that which loans are paid back. Deposits are redeemable in the short-term while loans are paid back over a much longer term. This is what the industry calls maturity transformation, but it is also what makes the business of banking inherently risky. To see why, let’s return to the concept of solvency.

A bank is solvent if the total value of its assets is equal to or greater than the total value of its liabilities. We can represent this solvency condition using the most basic accounting identity equation: Assets = Liabilities + Equity. The total value of what the bank owns, its assets, is exactly equal to what the bank owes, which are its liabilities and equity. Liabilities can be thought of as what the bank owes to deposit-holders and other lenders, while equity can be thought of as what the bank owes to its shareholders.

From the above equation there are three possible scenarios, two of which produce situations of solvency and one of which produces a situation of insolvency: 1) if Assets > Liabilities, then Equity > 0 and the bank is solvent; 2) if Assets = Liabilities, then Equity = 0 and the bank is solvent; 3) if Assets < Liabilities, then Equity < 0 and the bank is insolvent. The three scenarios are illustrated below.

Despite the three scenarios, the fact that we can reduce a bank’s condition down to solvency or insolvency makes it seem as though the whole notion of solvency is a binary condition. Reality, however, is much more complex, which is why Matthew C. Klein at the Financial Times writes, “Banks, by virtue of their unusual business model, exist in a netherworld between solvency and insolvency.”

The Netherworld Between Solvency and Insolvency

In considering how it is that a bank might go from a situation of solvency to one of insolvency it’s worth taking a moment to consider different types of bank assets, and their relative liquidity. A bank’s most liquid assets are cash and reserves. Sometimes the word reserves is used in a way that it is inclusive of cash, but the more specific usage of the word is to describe the deposit accounts that individual banks hold at the central bank (the central bank is the bank for private banks). Slightly less liquid assets, although not by much, would be government securities (i.e. government debt). Then there are loans to the private sector, both businesses and households alike. Finally, there would be a bank’s physical assets, such as the property and buildings that it owns.

Now, a hypothetical situation: suppose that for some reason an abnormal number of depositors redeem their deposits for cash or transfer their deposits to another bank. In the first instance, the bank would be paying out cash, and in the second instance, the bank would be transferring reserves to another bank via the central bank’s balance sheet (illustrated below). In either case, this bank is now running low on liquid assets.

In Situation 1, we have Households redeeming their deposits at Bank A for cash. This is illustrated above as a reduced deposit liability for Bank A matched by a reduction of Bank A’s cash holdings; for the Households, we have a swap of assets (i.e. Households are exchanging their deposits for cash).

In Situation 2, Households are once again swapping assets, but this time in the form of an exchange of deposits at Bank A for deposits at Bank B. Thus, Bank A’s deposit liability is extinguished while Bank B’s deposit liability increases. But, that’s not the end of the story. Remember, while households may treat banks’ liabilities as money, banks treat each other’s liabilities as credit. That means that Bank A is now indebted to Bank B and must settle this debt using a form of money higher up on the monetary hierarchy (notice how the figure, showing the central bank on top, suggests an institutional hierarchy that mirrors the hierarchy of money). Thus, the debt is settled with a transfer of reserves via the central bank’s balance sheet. Bank A’s account at the central bank is debited and Bank B’s account is credited.

Either Situation 1 or Situation 2, or a combination of both, result in a reduction of liquidity for Bank A. If deposit redemptions occur on a large enough scale, this could lead to a severe shortage of liquidity, and possibly a liquidity crisis.

There are several options a bank suffering from a shortage of liquidity could consider. One option is to sell other relatively liquid assets, such as government bonds. Normally, however, the more liquid the asset the less interest income it earns for the bank, so banks tend to try and keep these liquid assets to a minimum. If depositor redemptions are in excess of these relatively liquid assets, then the bank may have to try selling less liquid assets. But, the problem with trying to sell less liquid assets is precisely that they are less liquid, meaning the bank may have to sell them at below what would be considered a fair price in order to convert them into cash on such short notice.

Therein lies the problem for the bank. If it sells its illiquid assets at ‘fire-sale’ prices that are below the current value at which they are held on the bank’s balance sheet, then the bank’s balance sheet is going to take a hit. Enough hits like that and the value of the bank’s assets could fall below the value of the bank’s liabilities, knocking out shareholder equity and pushing the bank into a situation of insolvency.

Of course, there are other options the bank could consider. One of those is, rather than selling its illiquid assets, the bank could try using them as collateral to borrow in money markets, say from another bank. This option allows the bank to secure the necessary liquidity without having to take on losses from selling assets at less than desirable prices. The only catch here is that potential money market lenders have to be willing to accept the assets offered as collateral. If they have reason to question the quality of the assets, say because the assets might be backed by a borrower that is unlikely to repay, then they may demand higher compensation or refuse lending altogether. On the other hand, if the bank’s solvency is in question, then they may not be able to obtain a loan even against higher quality collateral.

It’s worth backing up for a moment. I began this example by supposing that an abnormal number of depositors were redeeming their deposits with the bank without an explanation of why this abnormality might be happening. Here’s one possible reason: market participants question the quality of the bank’s assets and begin to fear that the bank may be insolvent; the fear that the bank is insolvent then causes depositors to rush to get their money out. That’s what you call a classic run on the bank.

Professor Goodhart’s claim about illiquid banks quickly becoming insolvent and insolvent banks quickly becoming illiquid is probably starting to make a lot more sense now. But, there is still another option — borrow from the central bank.

In my last post, I said that banks make their lending decisions and then look for reserves later, and that the central bank supplies reserves on demand to the banking system. Well, for banks with short-term liquidity problems, this is precisely one of the functions that central banks were designed to perform. However, there is a catch. While central banks don’t mind supplying reserves as part of their monetary policy objectives to manage the interest rate or to help out banks struggling from a temporary liquidity crisis, they generally don’t like supplying them to banks that are insolvent.

It is that reality that Lehman Brothers was hit with nearly ten years ago, even though in the bank’s own opinion it had “a surplus of assets over liabilities of $28.4 billion,” according to The New York Times. Others disagreed. Bank of America’s CEO at the time asserted that the investment bank was in the hole by $66 billion. Lehman had found itself in the netherworld between solvency and insolvency, and the Federal Reserve, as the bank’s last hope, declared a verdict of insolvency. Like that, a U.S. bank with a 158-year history of creating money to lubricate the wheels of commerce by lending to its clients, collapsed.

The Inherent Limits to Bank-Created Money

The lesson to be learned from Lehman Brothers is that while banks create money, one shouldn’t take that to mean that they create it willy-nilly. Banks do face certain constraints. The one that I have been highlighting is imposed by the very nature of the banking business model itself.

As I mentioned, banks are in the business of making profits. They have costs, such as paying interest on deposit accounts, and they have revenue, which are the returns they earn on the loans they make. Their costs are more or less certain, but their revenues depend on whether the loans are paid back or not. Because of the uncertainty surrounding revenues, banks have to be careful about who they make loans to, since loans that aren’t repaid end up becoming worthless assets. Worthless assets don’t bring in revenue, and if you acquire enough of them, you could be paving the road to insolvency.

That is the inherent limit to creating money that all banks face, and yet the forces of competition push banks continually to that limit. Nobody wants to make bad loans, but being overly cautious means falling behind one’s competitors. Following the worst of the financial crisis, Lloyd Blankfein, CEO of Goldman Sachs, wrote about his profession in 2009, “…our self-interest in preserving and expanding our market share, as competitors, sometimes blinds us — especially when exuberance is at its peak.” In addition to the reference to self-interest and competition, the point about exuberance is also important. When the economy is booming, people generally find it easier to pay their debts. If people are easily paying their debts, then what’s the harm in giving them more?

While Lehman Brothers found out the hard way what the harm is, it’s hard to know if the U.S. banking sector as a whole learned the same lesson. The fear that gripped financial markets after Lehman collapsed would have undoubtedly caused other banks to suffer a similar fate, and the Fed knew it. With the whole banking sector at risk, solvent or not, the U.S. central bank wasn’t going to let the entire financial system crumble on its watch.

So, what’s the harm in making more loans when the government puts such a high premium on the role you play in the economy, making you arguably, “too big to fail”? If you’re too big to fail, then the apparent inherent limit to the creation of money appears to be a soft one. Because of the implicit guarantee that banks have from the government in times of crisis, profit-making private banks will be tempted to test the limits of that guarantee. This is just one instance of what economists call moral hazard, and it is for this moral hazard reason, that governments have generally found it prudent to impose some external limits. But that is an explanation that will have to wait for another post.