• The demands for cash are on such a scale often the whole banking system is run on that it is not possible to meet these demands because the assets of the banking system cannot be sold en masse without their prices plummeting.
• Crises are sudden, unpredictable events, although the level of fragility may be observable.
The crises described above involve different kinds of bank money. In 1857 customers of banks demanded gold and silver in exchange for their banknottl and demand deposits. In the panics of the National Banking Era, like the Panic of 190–. depositors demanded National Bank Notes in exchange for their checking accounts. In the Panic of 2007-8, depositors did not want the bonds as collateral for repo anymore. I will say more about the Panic of 2007-8 later. But at this point note that the run on repo involves the same vulnerability of bank debt that occurred with private banknotes and demand deposits. With repo, the backing collateral received by the depositor was not always U.S. Treasuries but sometimes risky bonds. The Panic of 2007-8 was structurally the same as the earlier panics.
The descriptions convey a sense of the suddenness of crises “as rapid, and apparently as unpremeditated, as the blow that brought about the French Revolution of 1848.” There is an abrupt change, like going over a cliff, or a sudden switch, a plunge. Economists like the term “regime switch” or “break-point.” Also, the descriptions emphasize the scale of the event. It is systemic. Many institutions are run on at more or less the same time, such that the demands for cash cannot be met by the system. These crises are systemic, which to repeat means that the demands for cash in exchange for bank liabilities are of such a magnitude that the demands cannot be honored. Prior to the existence of the Federal Reserve System, the banks would suspend convert-ibility rather than try to honor the demands for cash. During the Great De-pression, state governors and then President Roosevelt declared banking holidays to keep banks from having to pay out cash. These two characteristics scale and demands for cash are canonical features of systemic financial crises.
The banking system cannot possibly turn all, or significant amounts of, their liabilities into cash in a short period of time. This is because the backing assets can’t be sold fast enough without pushing their prices down. But if banks don’t have to sell assets, as in a suspension of convertibility, then it turns out the losses are minimal or at least were during the National. Banking Era. There is a very important difference between the earlier runs, when there w as no central bank and when the government did not intervene (and was not expected to intervene) and the later period, when the Federal Reserve Svstem or other central bank existed, and later still, when deposit insurance was in place, or when the government was expected to intervene. With regard to the earlier, pre-Fed period, it is clear that a financial crisis is a system-wide bank run. Mass hank runs don’t necessarily occur in the modern era, due to explicit or implicit government actions, or expectations of such actions. In the modern era, we can say that:
A financial crisis is an event where holders of short-term debt issued by financial intermediaries withdraw en masse or refuse to renew their loans or would have engaged in such a mass run had not explicit or im-plicit government intervention been in place or been expected.
While most crises involve bank runs, the definition is expanded because a clear-cut run may not occur because of expected government action. Instead of a straightforward panic, what happens depends critically on the private sector’s expectations or beliefs about the policies of the central bank or government. The lack of a bank run because of these expectations or indeed of the government’s actions creates the impression that financial crises are not related to the vulnerability of bank debt, but are about other factors, like shocks to bank capital and so on, as will be discussed below. I readily concede that the phrase “or would have engaged in such a mass run had not explicit or implicit government intervention been in place or been expected” is tricky because it poses a counterfactual. Later, I will show concretely what this means when asking what would have happened in the 1920s and 1930s in the United States had the Federal Reserve System not been in existence. The key is that the expectations and beliefs of people in the economy must be understood in a concrete, practical way so that the counterfactual can be con-structed. And later I will argue that this is an example of what economists need to do more generally. Financial crises are not uncommon, and they occur in all market econo-mies. They are sudden, unpredictable events, although the level of systemic fragility may be observable. They involve all or a significant number of banks (or other financial firms). A crisis may involve runs on banks, a sudden demand for cash on a scale that cannot be met by banks (although “banks” are not necessarily the firms that bank regulators think are banks). Bank runs may not occur if action by a government or central bank is expected, but there can still be a crisis in which the banking system is massively impaired.