Every student of money and banking is told that a key factor leading to the Great Depression was the breakdown of the US banking system. However, a closer look at the numbers shows a surprising resilience of the banking system, which continued to make profits even at the bottom of the Depression, whereas most other sectors made losses.
There is a general consensus today that we must do everything possible to avoid large-scale bank failures, and that this was not done during the 1930s. However, the impact of the widespread failures that took place then was much more limited than is generally assumed.
The number of bank failures rose from an annual average of about 600 during the 1920s to 1,350 in 1930 before peaking in 1933 when 4,000 banks were suspended. Over the entire period 1930-1933, one-third of all US banks failed. However, deposit losses remained limited even during this turbulent period at a cumulative 4% (with an annual peak of 2.15% in 1933) of total deposits (of all commercial banks). How can one reconcile these relatively modest losses with the large number of bank failures?
Table 1. Commercial bank suspensions during the Great Depression
||Number of Suspensions
||Deposits (millions of USD)
||Losses Borne by Depositors (millions of USD)
||Losses as % of deposits
||Loss rate in %
One reason why losses to depositors were ultimately quite limited is that, even in failed banks, depositors still received about 80 cents on the dollar, on average.1
Another key reason was that the degree of concentration in the banking industry was much lower then than it is today. In the mid-1930s, the top three banks held about 11% of the total assets of the industry; in 2008 they held about 40%. Although about one-third of all banks failed between 1930 and 1933, they accounted for only about 20% of all deposits.
The downside of this fragmentation of the banking sector was that few institutions were regarded as systemic. The failure of any one of the numerous “mini” banks did thus not arouse particular concern. Hence little was done to prevent the numerous bank failures that did occur. However, the continuing latent, if actuarially relatively small, risk of banks undermined confidence. Given the absence of a federal deposit insurance system, this undermined the functioning of the banking system as banks had to be extremely cautious given the potential for runs (mass withdrawals). The negative feedback loop that operates today – weakness of demand leading to more firms failing and hence bank losses – was thus amplified by the lack of an effective deposit insurance system and a generally higher willingness to allow banks to fail.
The need for a federal deposit insurance system in the US was one key lesson learnt, and the creation of the FDIC as part of the New Deal certainly contributed to the recovery. The lesson that confidence of depositors in the banking system is crucial has been amply applied in Europe during the present crisis, when EU governments expanded the existing deposit insurance systems to cover the holes that the Northern Rock episode exposed.
The US authorities had to relearn this lesson after the Lehman debacle. Given the large impact of the modest deposit losses during the 1930s, it should not have come as surprise that a failure of this size should have extreme consequences. As an investment bank, Lehman did not have any deposits from the general public, but its assets of about $660 billion were equivalent to about 5% of the entire US banking system.
Moreover, Lehman alone constituted a major part of the bank bond market, as it had issued about $600 billion of short- and long-term bonds, for a total of about $1200 billion emitted by all commercial banks together.
Can banks survive a depression?
The Great Depression certainly led to a collapse in corporate profits. The corporate sector went from profits amounting to over $10 billion (about 10% of GDP) in 1929 to a collective loss of about $1.5 billion (about 2.5% of GDP) in 1932 (see table 2).
Table 2. Corporate Profits before tax, million USD
|Domestic industries (including financial industry)
|Finance, insurance, and real estate
Oddly enough, in the banking sector, which is thought to have been the most affected by the crisis, profits stayed positive throughout this period. Bank’s profits declined by “only” a little more than 40%, more or less in line with nominal GDP. Figure 1 shows that bank’s profits were indeed relatively stable as a proportion of GDP (just below 1% of GDP).
Figure 1. Profits in US during the 1930s (before tax as % of GDP)
However, the financial sector outside banking did suffer losses. Banks thus appear to have been an island of relative stability. The run-up to and aftermath of the current crisis show similar features, with corporate profits buoyant before the crisis but collapsing with the onset of the bust. This boom-and-bust pattern is also pronounced in the financial sector outside banking.
The data for the last 25 years displayed in Figure 2 show that in general banking profits display little correlation with GDP growth. This property is shared neither by the rest of the financial sector nor the corporate sector as whole.
Figure 2. Profits in US after 1980 (before tax)
As most observers tend to lump banks and the larger financial sector together, it is widely assumed that banks’ profits also increased during the last boom. However, this was not the case. The profits of (commercial) banks as measured by national accounts did not noticeably increase up to 2007. The data from the Great Depression suggest that banks might be more resilient than the wider financial sector.
Another reason for the widespread impression of large profits in banking might be the confusion between the national income accounting concept and the numbers reported in financial statements. The national income accounts do not “mark to market”. Financial transactions and capital gains and losses related to transactions of financial securities are thus not included in the national accounts. During the boom phase, profits in the national income accounts tend thus to be lower than those reported to financial markets (and vice versa during the bust). Profits as measured by the national accounts should thus give a better picture of underlying, operating profitability of the sector.
The recent batch of relatively reassuring profits reported by many US banks can also be understood in the light of switch away from mark-to-market accounting. That most banks can report positive profits resembles the experience of the Great Depression.
How can one explain the relative stability of banks’ profits in general and in particular during the Great Depression? Simply put, it seems that banks are on average able to charge enough of a risk premium to cover the increase in non-performing loans during downturns. The present crisis confirms this tendency. Much has been made of the IMF’s recent headline estimate of over $4trillion in aggregate losses to the financial sector expected from the present crisis. However, this figure is an estimate of the total over four years (2007-10), and banks account for only about 60% of the overall amount. Moreover, about one-half of the losses expected by banks ($2.4 trillion, see table 1.3 of the Global Financial Stability Report, April 2009) derive from the markdown of securities (which would not be included in national income accounts).
The global average overall loss rate on loans for banks is estimated to be around 5.1%. Given that this is assumed to be the cumulated loss rate over four years, banks should be able to absorb it with a commensurate increase in their spreads. If the losses accrue mostly towards the end of this period, an increase in the average spread applied by banks of less than 2%should be sufficient to keep banks (on average at least) from making large losses. Banks do appear to be charging such spreads. For example, the ECB reports that the rates charged by euro area banks to corporate customers are now around 4.8%; about 2.5% – 3%higher than marginal funding costs as measured by Euribor rates or the rates paid on savings deposits. In the US the ‘prime rate’ (the rate charged by banks to their best customers) is at 3.25%; about 3% above marginal funding costs embodied in the federal funds or the commercial paper rates. This should be sufficient to deal with the losses that can be expected even under the current economic conditions.
Despite its much greater severity, the Great Depression did not actually lead to much higher loss rates. As shown in figure 1.30 of the IMF’s Global Financial Stability Report, commercial bank loan charge-offs peaked for only one year at a bit above 5%, but the average for the early 1930s remained between 2 and 3%.
The resilience of “normal” banking operations to a recession or even a depression strengthens the case for a separation of commercial and investment banking activities. The classic banking operations of deposit-taking and lending tend to remain profitable even under stressed conditions. But this classic function of banking would not be such a cause of concern today if the investment banking arms of banks had not gotten into trouble by investing in “toxic” assets. At present, the authorities in both the US and Europe have little choice but to make up for the losses on “legacy” assets and wait for banks to earn back their capital. But to prevent future crises of this type, policymakers should make sure that losses from investment banking arms cannot impair commercial banking operations.
Federal Deposit Insurance Corporation (FDIC) (1998), A brief history of deposit insurance in the US
IMF (2009) Global Financial Stability Report (GFSR), Responding to the Financial Crisis and Measuring Systemic Risks, April
1 Between 1908 and 1917, eight states established deposits insurance funds but by end of the 1920s they had all failed (FDIC, 1998)