The Failure of the Bank of the Commonwealth: An Early Example of Interest Rate Risk

The biggest risk in Commonwealth’s portfolio was the duration of its municipal securities portfolio. The interest rates on most municipal securities and loans were fixed, so the value of these obligations would drop if rates increased. Figure 3 reports the average duration of Commonwealth’s municipal security portfolio. It nearly doubled from 12 years and 11 months in 1965 to 23 years and 4 months in 1968. Commonwealth’s strategy was to bet that interest rates would decline, and if this happened, Commonwealth would earn a large capital gain on the securities.

A second factor behind Commonwealth’s strategy was the treatment of capital losses and capital gains by the federal tax code. While the federal tax code taxed corporate capital gains at a lower rate than corporate income, roughly 25 percent versus 50 percent in the 1960s, capital losses were treated differently for commercial banks. Net realized capital losses could be expensed against income, which was taxed at the high corporate income tax rate of roughly 50 percent.⁷ And, while income from municipal obligations was tax free, if interest rates went up, a bank could still sell the securities, incur the capital losses to reduce current taxes, and reinvest the proceeds in similar tax-exempt securities at their new lower price to receive roughly the same tax-free income as before. As long as the bank had positive income, the capital losses would be expensed and reduce taxable income. Commonwealth took advantage of this asymmetry by purchasing securities in order to take advantage of cyclical fluctuations in interest rates (Rose, 1968). Of course, this strategy does not work if the bank has losses, and this situation became a problem for Commonwealth.⁸

According to Irvine Sprague, who was on the FDIC board of directors in 1972 when the FDIC bailed out Commonwealth,

Commonwealth’s pattern, which was repeated at other banks, was to sell off the safe, steady, and staggered federal securities in the bank’s portfolio and load up on low-grade, long-term municipal securities that bore higher interest rates. After almost a decade at low relatively stable levels, interest rates had been drifting upward in the late 1960s, and COMAC was trying to lock them in. COMAC believed the rise was cyclical and that rates were ready to fall. If that happened, all those high-yielding municipals in the 5 percent tax-free range would surge in value and the client banks could sell them at a princely profit. . . .

As early as January 1968, representatives of the Chicago Fed met with Commonwealth’s board and expressed concern to the directors about the course bank management was following. Parsons, as Commonwealth chairman, strongly defended their policy, saying it would produce vast capital gains for the bank. He said the projections of Dr. Gies showed that a downturn in interest rates would occur between July 1, 1968, and July 1, 1969. (Sprague, 1986, pg. 59)

Unfortunately for Commonwealth, Gies’ forecast was wrong. Between July 1, 1968, and July 1, 1969, the yield on 20-year Treasury securities at a constant maturity actually increased from 5.35 percent to 6.29 percent.⁹ Similarly, short-term rates increased over this period, as well. What had happened was that because inflation continued to increase in this period, the Federal Open Market Committee tightened monetary policy in an attempt to reduce inflation (see Figure 4). Figure 5 reports unrealized losses on Commonwealth’s securities portfolio as a percentage of capital.¹⁰ These calculations report losses for the entire portfolio, but most of the losses were from the municipal securities portfolio.¹¹ Commonwealth’s losses by this measure grow as long-term rates stay high, and they nearly reach 100 percent of capital by 1970.

At the time, all securities were accounted for at book value, so the losses would only get recognized through earnings if a bank sold its underwater securities. Unfortunately, Commonwealth was faced with selling the securities because of loan losses during the 1970 recession and liquidity problems. Part of Parsons’ strategy for Commonwealth and the other banks he controlled was to fund their growth with a mix of high-interest certificates of deposit and wholesale borrowing from other banks in the fed funds market. Figure 6 shows the liability mix of Commonwealth and the growth in borrowing in the fed funds market. What the figure does not show, however, is that much of the time deposits were being funded by interest rate sensitive depositors who did not live in Detroit (Rose, 1968). Furthermore, Regulation Q interest rate caps were not raised in 1969 (Cook, 1978; Gilbert, 1986), a situation which made it harder for Commonwealth to raise funds with price-sensitive time deposits.¹² These reasons prompted Commonwealth’s difficulty in raising funds. As a result of this challenge, Commonwealth applied to open a branch in the Bahamas to raise funds in the Eurodollar market, but the Federal Reserve Board denied the application on March 31, 1970. Furthermore, in its denial, the Board noted that “the general character of management and the bank’s financial history and condition, including the liquidity and capital positions, mitigate against approval” (Sprague, 1986, pg. 62). As Sprague (1986) reports, it was very unusual for the Federal Reserve Board to issue such a scathing statement about the character of an applicant. After the release of this statement, Commonwealth’s demand deposits declined by around 18 percent from January 1, 1970, to April 30, 1970 (Gies, 1975). As Figure 6 illustrates, Commonwealth initially replaced these lost deposits with borrowing in the fed funds market, but that source of funding declined in the second half of 1970 and then was replaced by discount window loans. Commonwealth’s ongoing problems meant that it could no longer continue.

Resolving Commonwealth

As Commonwealth’s decline continued, in July of 1970 the Federal Reserve used the threat of removing access to the Fed’s discount window in order to force Parsons and several of his partners to resign as directors and officers of the bank and to agree to a cease and desist order in which, among other things, Commonwealth had to cut ties with COMAC, stop paying dividends, and reduce its size (Sprague, 1986).

With Commonwealth being kept alive via Fed lending, the next step was to determine how to deal with the bank. For regulators, the preferred way of handling a failing bank is to find a strong bank to acquire it. Unfortunately, there was no such bank available in Commonwealth’s case. Michigan’s banking laws did not let out-of-state banks operate in Michigan, so only a Michigan bank could acquire it. Furthermore, Michigan’s banking laws also prevented banks from opening a branch more than 25 miles from its headquarters, so only a Detroit-based bank could legally acquire Commonwealth. According to Sprague (1986), the three largest banks in Detroit had a 77 percent share of deposits, and letting one of them acquire Commonwealth’s 10 percent share would have increased the concentration within the Detroit banking market to an unacceptable level (Sprague, 1986).

In the absence of an acquisition, this meant that Commonwealth would fail. When a bank fails, it is not put into bankruptcy, but, instead, into FDIC resolution. The two primary ways the FDIC resolves a failed bank is to provide assistance to an acquiring bank (a purchase and acquisition) or to liquidate it. The criteria set by law for how to resolve a failing bank has changed over time, but, historically, a purchase and acquisition is the most commonly used method. Liquidation typically has been used for only the smallest banks (Horvitz, 1986).

Given the concentration in the Detroit market referred to above, the FDIC ruled out a purchase and acquisition, but it was not willing to liquidate the bank. At the time, regulators believed that a $1.2 billion bank—Commonwealth’s size in 1972—was too big to fail (Nurisso and Prescott, 2017, 2020). Instead, the FDIC invoked the rarely used essentiality clause of the Federal Deposit Insurance Act of 1950 that let the FDIC assist a bank in order to keep it operating if the FDIC deemed it “essential” to the local community. The FDIC’s assistance to Commonwealth was only the second use of this power.¹³ The FDIC forced Commonwealth to sell much of its municipal securities portfolio and recognize the losses and lent the bank $60 million to replenish its capital. Commonwealth then limped along until it was acquired by Comerica in 1984.

Comparison with SV Band First Republic

Silicon Valley Bank got into trouble for the same reason Bank of the Commonwealth did. It bought large quantities of long-duration, fixed-rate securities in a period of low interest rates and low inflation, and then both rates and inflation increased.

A second similarity was that both banks had unstable funding bases. For SVB, the unstable funding came from its uninsured deposits, which were 94 percent of its domestic deposits at the end of 2022.¹⁴ A third similarity was that SVB’s uninsured depositors ended up being protected by financial regulators, just like Commonwealth’s, albeit by different means. Since the Great Depression, uninsured depositors of a failing large bank have rarely lost their funds (Horvitz, 1986; Stern and Feldman, 2009), and despite the many changes to banking law since Commonwealth’s failure, the outcome for uninsured depositors at these two banks was the same.

One difference between the two banks was the degree and speed of the withdrawals and how that affected the failing bank’s resolution. Historically, most banks have enough insured and stable deposits that a combination of discount window lending and other sources of lending can keep a bank operating until a solution is found. This was the case with Commonwealth and for the other too-big-to-fail bailouts of the 1970s and 1980s (Nurisso and Prescott, 2017, 2020). As discussed above, Commonwealth did have several sources of unstable funding, but it also had some stable funding. As indicated in Figure 6, the bank retained a sizeable amount of its demand, time, and savings deposits after 1970. While the Call Report provides no information on the extent to which its deposits were FDIC insured, presumably a sizable fraction of them were insured. As a result, Commonwealth’s run played out over time and gave regulators more time to resolve it.

In contrast, most of SVB’s deposits were uninsured and, as is well-documented, many were held by a small number of depositors who were highly connected to each other and could quickly initiate withdrawals (Board of Governors, 2023). As a result, the speed and size of the run on SVB was so fast and so large that there was not time to borrow from the discount window, let alone to find a buyer. Instead, regulators shut it down and used the systemic risk exception contained in the Federal Deposit Insurance Corporation Improvement Act of 1991 to protect uninsured depositors.¹⁵

Instead, the resolution of Commonwealth looks more like that of First Republic, at least in the sense that there was more time to find an acquiring bank. First Republic was a bank based in San Francisco, California, that specialized in catering to a wealthy clientele. Like Commonwealth, First Republic invested in long-duration, fixed-rate assets, though its investments were primarily in residential mortgages. Forty-six percent of its assets were in first-lien, 1–4-family residential mortgages, and most of these mortgages had a long duration. About 47 percent of these mortgages had a rate that was fixed for five to 15 years, and another 31 percent had a rate that was fixed for at least 15 years. First Republic looked much like a savings and loan in the 1970s, but unlike those institutions, it had a large fraction of uninsured depositors: roughly two-thirds of its deposits were uninsured.¹⁶

When First Republic experienced a run starting in March 2023, several large banks lent to it to keep it operating, and this lending gave the FDIC time to find an acquirer. On May 1, 2023, the FDIC resolved First Republic by arranging a purchase and acquisition in which JP Morgan bought all of First Republic’s deposits and most of its assets, and the FDIC contributed funds to the purchase.¹⁷ The assisted purchase by JP Morgan raises another similarity with Commonwealth. As discussed earlier, concentration concerns along with Michigan banking law limited the pool of acquirers for Commonwealth. In the case of First Republic, the pool of acquirers was slightly limited because the Riegle–Neal Interstate Banking and Branching Efficiency Act of 1994 prevents a bank from acquiring another bank if the acquisition would give the acquiring bank more than a 10 percent market share of deposits nationwide. Indeed, the eventual acquirer, JP Morgan, already exceeded this threshold before the acquisition. There is an exception to the law, however, if the acquired bank is failing, and the exception was used in this case (Eisen and Ackerman, 2023).

A difference between Commonwealth and both SVB and First Republic is the source of interest rate risk. Unlike Commonwealth’s, both SVB’s and First Republic’s interest rate risk exposure came primarily from residential mortgages.¹⁸ In SVB’s case, the risk was from holding mortgage-backed securities, while for First Republic it was from directly holding residential mortgages. In the late 1960s, however, the commercial banking sector did not hold many residential mortgages. Instead, most were held by the thrift industry, that is, savings and loan associations, mutual savings banks, and savings banks, rather than by commercial banks. Unlike today, the thrift industry in the 1960s was sizable, holding half of the assets of the commercial banking industry. Thrifts were required by law and regulation to mainly hold residential mortgages, many of which were fixed rate and of a long duration at the time. Indeed, the thrift industry held around 57 percent of residential mortgages during the late 1960s, while commercial banks only held about 15 percent.¹⁹ Unrealized losses on the thrifts’ residential mortgages, along with regulatory forbearance and then deregulation in the early 1980s, led to the Savings and Loan Crisis, which played out over two decades. For histories of that important event, see Kane (1989) and White (1991).

Conclusion

As in the recent episode of inflation, the inflation of the late 1960s was preceded by a long period of low interest rates and low inflation. In both periods, some commercial banks extended the duration of their assets, betting that rates would decrease or at least not increase. Instead, rates increased, and several of these banks failed. The details of how Commonwealth, SVB, and First Republic were resolved differ as a result of changes in banking law, bank structure, and payments technology. Nevertheless, in all three cases the response of regulators was similar, and uninsured depositors were protected.