In yesterday's post, I discussed the first day of the Atlanta Fed's two-day virtual workshop on monetary and financial history. In today's post, I'll discuss the workshop's second day.
Day two began with a paper presentation by Chris Cotter of Oberlin College and discussion by Hugh Rockoff of Rutgers University. Cotter's paper
("Off the Rails: The Real Effects of Railroad Bond Defaults Following the Panic of 1873") analyzes the knock-on effects of railroads' bond defaults stemming from the 1873 financial panic. About one-quarter of all U.S. railroads defaulted on their bonds then. The paper's data set combines data on bond defaults with geographic data on national banks operating in areas served by the defaulting railroads. The main result of the paper is that even though banks did not (and legally could not) hold railroad bonds, the presence of a defaulting railroad in the area served by a bank tended to contract the loans and deposits at that bank. The railroad bond defaults thus exerted systemic, negative effects on the U.S. banking system despite lack of direct exposure of banks' portfolios to the bond defaults.
In the discussion, Rockoff agreed with the paper's conclusion but proposed that the paper could be strengthened by including case studies to check for their consistency against historical narrative. Rockoff also suggested a robustness check of comparing the effects of the 1873 panic to those of an 1877 nationwide railway strike. The post-1873 economic contraction was also one of the longest in U.S. economic history, Rockoff noted, so it would be interesting to know how much the railroad bond defaults contributed to the postpanic slowdown in economic growth.
The next paper presentation was by Lee Ohanian of the University of California, Los Angeles, with a discussion by Angela Redish of the University of British Columbia. The paper ("The International Consequences of Bretton Woods Capital Controls and the Value of Geopolitical Stability," coauthored with Diana Van Patten of Princeton University), Paulina Restrepo-Echavarria of the Federal Reserve Bank of St. Louis, and Mark L.J. Wright of the Federal Reserve Bank of Minneapolis) models the world economy using a three-sector general equilibrium model (the United States, Western Europe, and the rest of the world) and uses this model to measure the impact of the Bretton Woods system of exchange controls. These controls were present from the end of the second World War until 1973, and in the model, these controls show up as taxes ("wedges") on the intersector movement of capital. The main result of the paper is that these wedges redirected very large amounts of capital away from the United States as compared to a first-best allocation, reducing growth and consumption in the United States but increasing them elsewhere. Aggregate global welfare was also reduced. Ohanian argued that despite its large domestic cost, the United States was willing to tolerate such a system for geopolitical reasons.
Redish noted in her discussion that while Bretton Woods is commonly thought of as an exchange-rate regime, in practice capital controls were necessary to afford countries some degree of monetary autonomy under fixed exchange rates. Redish also noted that the capital controls took many different forms and a closer examination of which types of capital controls were actually implemented could amplify the paper's message. She suggested that the high level of aggregation in the model obscures some potentially important cross flows of capital (for example, inflows into Germany are netted against outflows from the United Kingdom). The paper's counterfactual simulations are striking, but additional narrative could improve them. While supportive of the paper's overall conclusions, Redish noted that one unmodeled benefit of fixed exchange rates was reduced exchange rate volatility, which could have promoted capital formation. Another unmodeled benefit of Bretton Woods could have been a reduced incidence of financial crises stemming from "hot money" flows, which ideally could be weighed against the costs of inefficiently allocated capital.
The second invited lecture of the conference was presented by Catherine Schenk of the University of Oxford. Schenk's presentation described a multiyear research project that will collect and analyze data on global correspondent banking, especially as it developed in the post-WWII era ("Constructing and Deconstructing the Global Payments System 1870–2000"). The presentation focused on events during the 1960s and 1970s. The expansion of foreign exchange trading during this era led U.S. banks to found a technologically advanced, privately owned, large-value payment system (CHIPS) in 1970. Schenk explained how CHIPS enabled banks to settle the rapidly growing volumes of U.S. dollar payments from foreign exchange trading and facilitated the expansion of the global correspondent banking system. She also described how problems with CHIPS and certain other features of the correspondent banking system came to light in 1974 with the failure of a German bank, Bankhaus Herstatt. The Herstatt failure revealed the extent of the expanded correspondent system and also highlighted potential risks arising from unsettled foreign exchange trades, creating new challenges for banking regulators.
The audience discussion focused on changes in the regulatory environment coinciding with or following the Herstatt failure. William Roberds pointed out that a longer-term consequence of Herstatt was the founding of CLS in 2001 as a mechanism for coordinating foreign exchange settlements. Schenk noted that the Basel Committee on Bank Supervision was formed shortly after 1974, leading to global coordination in bank capital requirements and other supervisory standards. Robert Hetzel pointed out that 1974 witnessed another watershed bank failure, that of Franklin National, which like Herstatt was also heavily exposed in foreign exchange transactions. Responding to a question by Alain Naef (Banque de France), Schenk argued that many of the problems banks faced with foreign exchange operations in the 1970s simply resulted from a technical inability to handle an increased transaction volume. Michael Bordo noted that the technical changes in transaction technologies during this period interacted with economic forces to create profound changes in global banking.
Alain Naef of the Banque de France presented the final paper of the second day ("Blowing against the Wind? A Narrative Approach to Central Bank Foreign Exchange Intervention"). Owen Humpage of the Federal Reserve Bank of Cleveland discussed it. The paper considers the effectiveness of Bank of England foreign exchange interventions over a sample running from 1952 until 1992, using daily data the Bank has recently made available. The Bank intervened on almost 80 percent of trading days during the sample, and most interventions were not publicized. The Bank intervened to influence the exchange rates of the pound against the dollar and deutschmark. Interventions were offset (sterilized) through domestic open market operations. Naef's presentation highlighted the main result of the paper, which is that interventions were usually ineffective when they attempted to go against market trends, in which case they were estimated to succeed only 8 percent of the time.
In the discussion, Humpage placed Naef's results in the context of the extensive literature on sterilized foreign exchange intervention. He noted that many traditional theories of sterilized intervention are oriented around the idea that such interventions could serve as a signal of a central bank's private information or intent. These theories would not seem to apply to the sample Naef analyzed, however, in which interventions were rarely publicized. He also noted that endogeneity concerns are common to this type of study. He recommended an alternative empirical approach focusing on the probability of certain market movements following an intervention, which could allow for a broader range of explanatory variables (for example, whether an intervention was coordinated with other central banks). Humpage also suggested additional clarification as to whether interventions in the data set were undertaken on the initiative of the Bank of England or the UK Treasury. Lastly, he noted that despite the apparent ineffectiveness of sterilized intervention, there are long stretches in the data where pound exchange rates appear to be pegged, indicating that there may be some unmodeled interactions between Bank policy and the foreign exchange markets that should be taken into account in the analysis.
The workshop concluded with a panel discussion 
of the potential impact of central bank digital currencies (CBDCs). The first panelist, Michael Bordo, proposed that the introduction of CBDCs could be as transformative as the introduction of circulating, central-bank-issued currency in the 17th and 18th centuries. Bordo said that while there was a role for private digital currencies, CBDCs could play a stabilizing role in the digital monetary landscape. He also suggested that CBDCs could also expand the options available for monetary policy, facilitating (for example) negative policy interest rates or tiered interest rates on central bank liabilities.
The second panelist, Warren Weber, began with the observation that 80 percent of the world's central banks are now at least considering issuing central bank digital currencies, most to be eligible for use in retail (consumer) transactions. Motivating factors for this development include the declining use of physical currency in some countries and Facebook's proposal to create a private digital currency (originally named Libra and now named Diem). A related motivating factor is the concern central banks have about financial stability risks that private digital currencies pose. Weber discussed this last issue in the context of two historical episodes when privately issued paper currency was commonplace and currency issued by central banks was not (in the United States between 1786 and 1863 and in Canada between 1817 and 1890). Weber argued that many private currencies in circulation during these periods failed to meet the definition of "safe asset" since their value tended to fluctuate, and sudden losses of value could occur when an issuing bank failed or suspended payments. However, private currencies became more reliable after more stringent regulation was introduced (1863 in the United States and 1890 in Canada), and in both cases this heightened reliability was accomplished without the introduction of central bank currency. On the basis of these experiences, Weber argued that CBDCs were not necessary for reliable digital currencies to exist, although CBDCs could be desirable on other policy grounds.
The third panelist in this session was François Velde of the Federal Reserve Bank of Chicago. Velde discussed CBDCs in the context of the general history of central bank money. He argued that central bank money historically arose to fill gaps in existing monetary systems (resolving ambiguity about units of account, facilitating payments, or boosting governments' fiscal capacity), and that historically, most central banks have shied away from involvement with retail payments other than the provision of paper currency. If CBDCs become prevalent, then they will still need to be oriented around provision of a stable unit of account, Velde noted, but an unanswered question is whether a widely accessible CBDC would fundamentally alter the relationship between private and central bank money. History suggests that governments and central banks will have some degree of involvement with digital currency, as with other forms of money, but the extent and form of this involvement are yet to be determined. Velde concluded with the observation that monetary innovations often have not resulted from conscious policy decisions but from a combination of underlying societal trends and chance occurrences.
In the subsequent panel discussion, Gorton argued that the more interesting types of available digital currencies are stablecoins, which purport to maintain a constant value against a central bank currency such as the U.S. dollar. Gorton argued that to be fully credible, stablecoins will need to operate under some degree of regulation, and, as the use of stablecoins expands, lawmakers may face the issue of whether stablecoin issuance falls under the purview of bank regulation. He noted that in to promote its state-issued digital currency, China has effectively shut down private digital currency issuance. Gorton and other panelists predicted that much of the future success of digital currencies would derive from more convenient cross-border payments—for example, along international supply chains. Bordo argued that even within domestic markets, digital currencies including CBDCs could offer efficiency gains over existing payment channels. If private digital currencies become sufficiently widespread, however, Bordo argued that they could interfere with central banks' ability to conduct monetary policy.
Velde then noted that the challenges facing digital currency adoption remain daunting, with mainstream acceptance probably requiring some degree of regulation to establish sufficient scale and credibility. Gorton agreed, but said that the example of money market mutual funds showed that such regulation could be challenging to get right. A question was posed as to whether governments' fiscal demands might also promote interest in CBDC issue, to which Velde said current low rates of interest on government debt do not provide strong incentives for governments to seek seigniorage through CBDC issue. Weber and Gorton suggested that what we might see instead of CBDCs are traditional banks moving into the issue of digital currencies as these become more widely accepted.
In the audience discussion, Peter Rousseau (Vanderbilt University) proposed that early U.S. monetary history showed that government regulation was not necessary to establish functional currencies, and that problems such as those that arose with pre–Civil War state banknotes could be minimized with modern technologies such as the blockchain. Chris Meissner (University of California, Davis) questioned whether, from a political economy point of view, private digital currencies would be allowed to become widespread enough to compete with CBDCs. Gorton responded by saying that in countries such as the United States, it will not be politically feasible to outlaw private digital currencies. Hugh Rockoff then remarked that CBDCs could facilitate fiscal transfers and Bordo said that in general, financial inclusion could be bolstered through CBDCs. Maylis Avaro (University of Oxford) noted that there did exist an historical example of such "retail outreach" by the Banque de France, which offered consumer accounts. Mark Carlson (Board of Governors) questioned whether the technological feasibility of fiscal transfers through CBDCs might affect central bank independence.
Larry Wall (Federal Reserve Bank of Atlanta) observed that the increased cross-border efficiency of CBDCs could lead to increased competition between central bank currencies. Bordo stated that the historical pattern of dollarization supported Wall's hypothesis, and Gorton suggested that one major reason that China has been accelerating development of its digital currency is to promote cross-border usage and international acceptance of the yuan.
By 
Will Roberds, a research economist and senior adviser in the Atlanta Fed's Research Department