The Great Depression as a saving glut

by Victor Degorce (EHESS & European Business School) & Eric Monnet (EHESS, Paris School of economics & CEPR).

This blog is part of our EHS 2020 Annual Conference Blog Series.


 

GreatDepression
Crowd at New York’s American Union Bank during a bank run early in the Great Depression. Available at Wikimedia Commons.

Ben Bernanke, former Chair of the Federal Reserve, the central bank of the United States, once said ‘Understanding the Great Depression is the Holy Grail of macroeconomics’. Although much has been written on this topic, giving rise to much of modern macroeconomics and monetary theory, there remain several areas of unresolved controversy. In particular, the mechanisms by which banking distress led to a fall in economic activity are still disputed.

Our work provides a new explanation based on a comparison of the financial systems of 20 countries in the 1930s: banking panics led to a transfer of bank deposits to non-bank institutions that collected savings but did not lend (or lent less) to the economy. As a result, intermediation between savings and investment was disrupted, and the economy suffered from an excess of unproductive savings, despite a negative wealth effect caused by creditor losses and falling real wages.

This conclusion speaks directly to the current debate on excess savings after the Great Recession (from 2008 to today), the rise in the price of certain assets (housing, public debt) and the lack of investment.

An essential – but often overlooked – feature of the banking systems before the Second World War was the competition between unregulated commercial banks and savings institutions. The latter took very different forms in different countries, but in most cases they were backed by governments and subject to regulation that limited the composition of their assets.

Although the United States is the country where banking panics were most studied, it was an exception. US banks had been regulated since the nineteenth century and alternative forms of savings (postal savings in this case) were limited in scope.

By contrast, in Japan and most European countries, a large proportion of total savings was deposited in regulated specialised institutions. Outside the United States, central banks also accepted private deposits and competed with commercial banks in this area. There were therefore many alternatives for depositors.

Banks were generally preferred because they could offer additional payment services and loans. But in times of crisis, regulated savings institutions were a safe haven. The downside of this security was that they were obliged – often by law – to take little risk, investing in cash or government securities. As a result, they could replace banks as deposit-taking institutions, but not as lending institutions.

We prove our claim thanks to a new dataset on deposits in commercial banks, different types of savings institutions and central banks in 20 countries. We also study how the macroeconomic effect of excess savings depended on the safety of the government (since savings institutions mainly bought government securities) and on the exchange rate regime (since gold standard countries were much less likely to mobilise excess savings to finance countercyclical policies).

Our argument is not inconsistent with earlier mechanisms, such as the monetary and non-monetary effects of bank failures documented, respectively, by Milton Friedman and Anna Schwartz and by Ben Bernanke, or the paradox of thrift explained by John Maynard Keynes.

But our argument is based on a separate mechanism that can only be taken into account when the dual nature of the financial system (unregulated deposit-taking institutions versus regulated institutions) is recognised. It raises important concerns for today about the danger of competition between a highly regulated banking system and a growing shadow banking system.

Business bankruptcies: learning from historical failures

by Philip Fliers (Queen’s University Belfast), Chris Colvin (Queen’s University Belfast), and Abe de Jong (Monash University).

This blog is part of our EHS 2020 Annual Conference Blog Series.


 

 

BankruptBlog
The door of a bankrupt business locked with a chain and padlock. Available at Flickr.

 

Business bankruptcies are rare events. But when they occur, they can prove catastrophic. Employees lose their jobs, shareholders lose their savings and loyal customers lose their trusted suppliers.

Essentially, bankruptcies are ‘black swan’ events in that they come as a surprise, have a major impact and are often inappropriately rationalised after the fact with the benefit of hindsight. While they may be extreme outliers, they are also extremely costly for those affected.

Because bankruptcies are so rare, they are very hard to study. This makes it difficult to understand the causes of bankruptcies, and to develop useful early warning systems.

What are the risk factors for which shareholders should watch out when evaluating their investments, or when pension regulators audit the future sustainability of workplace pension schemes?

Our solution is to exploit the historical record. We collect a dataset of all bankruptcies of publicly listed corporations that occurred in the Netherlands over the past 100 years. And we look to see what we can learn from taking this long-run perspective.

In particular, we are interested in seeing whether these bankruptcies had common features. Are firms that are about to go out of business systematically different in terms of their financial performance, corporate financing or governance structures than those that are healthy and successful?

Our surprising result is that the features of bankrupt corporations vary considerably across the twentieth century.

During the 1920s and 1930s, small and risky firms were more likely to go bankrupt. In the wake of the Second World War, firms that did not pay dividends to their shareholders were more likely to fail. And since the 1980s, failure probabilities have been highest for over-leveraged firms.

Why does all this matter? What can we learn from our historical approach?

On first glance, it looks like we can’t learn anything; the drivers of corporate bankruptcies appear to change quite significantly across our economic past.

But we argue that this finding is itself a lesson from history.

The development of early warning failure systems needs to take account of context and allow for a healthy degree of flexibility.

What does this mean in practice?

Well, regulators and other policy-makers should not solely rely on ad hoc statistical models using recent data. Rather, they should combine these statistical approaches with common sense narrative analytics that incorporate the possibility of compensating mechanisms.

There are clearly different ways in which businesses can go bankrupt. Taking a very recent perspective ignores many alternative routes to business failure. Broadening our scope has permitted us to identify factors that can lead to business instability, but also how these factors can be mitigated.

Taxation, fiscal capacity and credible commitment in eighteenth-century China: the effects of the formalization and centralization of informal surtaxes

by Max Hao (Peking University) and Kevin Liu (The Hong Kong University of Science and Technology).

 

In premodern Europe, famine relief was inadequately provided until the late 19th century.  In contrast, in late imperial China, preventing starvation helped legitimate the state and played a key role in reducing internal conflicts. The Qing state operated a network of granaries and developed sophisticated procedures to report famines and supply relief. However, as shown in Figure 1, the frequency of famine relief recorded in the Qingshilu (veritable records of Qing) was much lower than the frequency of disasters under the reigns of Shunzhi (1644-1661) and Kangxi (1661-1722). In contrast, in the reign of Yongzheng (1723-1735) and in the early years of Qianlong (1736-1760), famine relief became more responsive to disasters. Why did the Qing state take on the responsibility for “nourishing the people” only after 1723?

Picture 0
Figure 1: Province-level frequency of disasters, famine reliefs and tax exemptions, 1645-1722 (Four-year moving average)

To solve this puzzle, we explore the effect of a reform that formalized and centralized part of the fiscal system in premodern China: the ‘huohao turned to public reform instituted in Emperor Yongzheng’s reign (1723-1735).   ‘Huohao’ denotes all the informal surtaxes collected by county governments. Because the bulk of formal tax revenue was remitted directly to the central government, provincial and county governments retained only a limited share of this income with limited discretion over its expenditure.  Consequently,  they imposed huohao or surtaxes to finance their own expenses. However, because these informal revenues were unsanctioned and unmonitored by the central government, they were largely pocketed by officials.

The ‘huohao turned to public’ reform was an endeavor to formalize huohao in order to achieve two interconnected policy goals: to reduce corruption and enhance provincial fiscal capacity by centralizing control. First, huohao was collected at rates designated by the provincial governor and delivered to the provincial treasury. Second, 60% of the remitted funds were allocated to county magistrates and provincial governors as ‘anticorruption salaries’ to finance their regular expenses, reducing their incentive to collect informal surtaxes and seize public revenues. More importantly, 40% of the formalized houhao was assigned as public funds to finance irregular expenses at the governors’ discretion. The total annual revenue from the formalized huohao amounted to 4.5 million taels of silver, of which 1.8 million were reserved as public funds. In sum, by formalizing and centralizing informal surtaxes, the reform enhanced provincial fiscal capacity by giving provincial governors more resources and a greater incentive to spend them on public goods. The design of the reform is illustrated in Figure 2.

 

Picture 1

 

Because corruption is extremely difficult to explore empirically, we mainly focus on whether the second goal was achieved. The timing of when the reform was initiated and completed differed between provinces, but the reasons behind these variations  were largely exogenous to provincial characteristics, enabling us to use them to test the effects of reform. We test whether the huohao reform raised the frequency of famine relief in periods of disastrous weather by exploiting the different timing of the reform process across provinces. We restrict our dataset to 1710-1760, when the bulk of famine relief was financed by the provinces.

Using a prefecture-level panel dataset, we find that in times of extreme drought and floods, the frequency of famine relief increased after the reform by 1.05 times per prefecture, which was more than 100% of the standard deviation of the dependent variable relative to that under non-disastrous weather. By exploring the dynamics of the reform’s impact, we find no pre-trend, which supports the exogeneity of the reform’s timing. Our results are robust to controlling for other initiatives by the central government, such as tax exemptions, the allocation of tribute grain and central fiscal revenues, the enforcement of bureaucratic monitoring, and other concurrent fiscal reforms. We also find that famine relief effectively reduced grain prices when disasters occurred, indicating that public funds were spent on famine relief which had a beneficial impact on the population. Further, we find that the reform’s impact was greater when areas faced exceptional flooding compared to exceptional droughts, and greater in prefectures which had difficulties collecting taxes, suggesting that the reform facilitated the intertemporal and spatial redistribution of financial resources.

For this tax reform to have a sustained effect on provincial government capacity,  central government would need to resist the expropriation of these new revenues for its own use.  However, in premodern China, there were no institutional constraints on this dispossession.  After emperor Qianlong (1736-1796) succeeded to the throne, the central government began to make regular checks on the expenditure of provincial public funds, forced the inter-provincial transfers of funds, and expended them on projects previously financed by central revenues. These actions forced provincial governments to reduce their expenditure on famine relief and withhold anticorruption salaries meant for county administrators. This finding highlights that it was the lack of credible commitment that accounted for the short-lived success of this fiscal reform. Viewed from this perspective, the reform provides a valuable lesson about the role of political institutions in the Great Divergence.

 

To contact the authors:

Max Hao (maxhao1003@pku.edu.cn)

Kevin Liu (kevin.liu@connect.ust.hk)

A Silver Transformation: Chinese Monetary Integration in Times of Political Disintegration during 1898–1933

by Debin Ma (London School of Economics and Hitotsubashi University)  and Liuyan Zhao (Peking University)

The full paper is due to be published in The Economic History Review and is currently available on Early View.

 

chinese coins
Two 19th Century Chinese Cash Coins. Available at <https://www.coincommunity.com/forum/topic.asp?TOPIC_ID=70505>

Despite the political turmoil, the early 20th century witnessed  fundamental economic and industrial transformations in China.  Our research documents the most important but neglected aspect of this development:  China remained on the silver standard until 1936 while many countries remained on gold.  Nonetheless, the Chinese silver regime defies easy classification because  its silver basis was traditionally not in coinage, but in the form of privately minted ingots called sycee, denoted by a unit of account called tael.  During our study period, sycee circulated alongside standardized silver coins such as Mexican and later Chinese silver dollars. We know relatively little about the operation of the silver exchange and monetary regime within China, in contrast to the large literature on the gold standard during the same era.

We present an in-depth analysis of China’s unique silver regime by offering a systematic econometric assessment of Chinese silver market integration between 1898 and 1933.  As a result of this integration, the dollar-tael exchange rate, the  yangli, became the most important indicator of the Chinese currency market. We compile a large data set culled from contemporary publications on the yangli across nineteen cities in Northern and Central China, and offer a threshold time series methodology for measuring silver integration comparable to that of gold points.

Picture 1
Figure 1. Silver point estimates between Shanghai and Tianjin in 10-year moving windows, Jan. 1898–March 1933. Source: Ma and Zhao (per article in the Economic History Review, 2019)

We find that the silver points between Shanghai and Tianjin, the two most important financial centers in Central and Northern China, declined  steadily from the 1910s for the rest of the period (Figure 1).  Our estimates of silver points from the daily rates of nineteen cities during the 1920s and 1930s also reveal that there was no substantial difference in the level of monetary integration between the Warlord Era of the 1920s and the Nanjing decade of the 1930s. Figure 2 provides a simple linear plot of  the distance between Shanghai and the estimated silver points of those cities paired with Shanghai during the 1920s and 1930s. This Figure shows a positive relationship between silver points and the distance from Shanghai, indicating the rise of a monetary system centered on Shanghai.

Our silver point estimates are closely aligned with the actual costs of the silver trade derived from contemporary accounts. Moreover, the silver points help predict corresponding transaction volumes: the majority of large silver exports from Shanghai occurred when the  yangli spread was above the silver export points;  only limited flows occurred when it fell within the bounds of the silver points. The econometric results reveal that monetary integration between Shanghai and Tianjin improved in the 1910s—precisely during the Warlord Era of national disintegration and civil strife—and these improvements spread to other cities in Central and Northern China in the 1920s and 1930s.

Picture 2
Figure 2. Silver points and distance. Source: Ma and Zhao (per article in the Economic History Review, 2019

Our research provides a historical analysis of the causes of monetary integration, attributing a central role to China’s infrastructure and financial improvements during this period. One plausible driving force was the rise of new transport and information infrastructure, for example, the completion of the Tianjin-Nanjing Railway, and the Shanghai-Nanjing and Shanghai-Hangzhou Railways constructed between 1908 and 1916, which linked the Northern and Southern China. Compared with road or water transport, railroads offered much faster, cheaper and safer delivery, an advantage far more significant for high-value silver shipments than low-value high-bulk commodities.

Another, more important factor was monetary and financial transformation indicated by the rise of a modern banking system from the end of the 19th century. Although it was the government that issued national dollars, banking communities played a key role in defending its reputation and purity. Overtime, the ‘countable’ dollar outperformed the ‘weighable’ sycee as a medium of exchange, gaining an increasing share in China’s monetary system. This eventually paved the way for the currency reform of 1933, which abolished the sycee and the tael, establishing the dollar as the sole standard. A notable monetary transformation was the increasing popularity of banknotes. The system of Chinese bank note issuance was largely run on a model of free banking with multiple public and private banks, Chinese or foreign, issuing silver-convertible banknotes based on reputation mechanism. Thus, the increasing note issue from the 1910s provided a much more elastic currency to smooth seasonality in the money markets and enhance financial integration.

 

To contact the authors:

Debin Ma (D.Ma1@lse.ac.uk)

Liuyan Zhao (zhly@pku.edu.cn)

Turkey’s Experience with Economic Development since 1820

by Sevket Pamuk, University of Bogazici (Bosphorus) 

This research is part of a broader article published in the Economic History Review.

A podcast of Sevket’s Tawney lecture can be found here.

 

Pamuk 1

New Map of Turkey in Europe, Divided into its Provinces, 1801. Available at Wikimedia Commons.

The Tawney lecture, based on my recent book – Uneven centuries: economic development of Turkey since 1820, Princeton University Press, 2018 – examined the economic development of Turkey from a comparative global perspective. Using GDP per capita and other data, the book showed that Turkey’s record in economic growth and human development since 1820 has been close to the world average and a little above the average for developing countries. The early focus of the lecture was on the proximate causes — average rates of investment, below average rates of schooling, low rates of total productivity growth, and low technology content of production —which provide important insights into why improvements in GDP per capita were not higher. For more fundamental explanations I emphasized the role of institutions and institutional change. Since the nineteenth century Turkey’s formal economic institutions were influenced by international rules which did not always support economic development. Turkey’s elites also made extensive changes in formal political and economic institutions. However, these institutions provide only part of the story:  the direction of institutional change also depended on the political order and the degree of understanding between different groups and their elites. When political institutions could not manage the recurring tensions and cleavages between the different elites, economic outcomes suffered.

There are a number of ways in which my study reflects some of the key trends in the historiography in recent decades.  For example, until fairly recently, economic historians focused almost exclusively on the developed economies of western Europe, North America, and Japan. Lately, however, economic historians have been changing their focus to developing economies. Moreover, as part of this reorientation, considerable effort has been expended on constructing long-run economic series, especially GDP and GDP per capita, as well as series on health and education.  In this context, I have constructed long-run series for the area within the present-day borders of Turkey. These series rely mostly on official estimates for the period after 1923 and make use of a variety of evidence for the Ottoman era, including wages, tax revenues and foreign trade series. In common with the series for other developing countries, many of my calculations involving Turkey  are subject to larger margins of error than similar series for developed countries. Nonetheless, they provide insights into the developmental experience of Turkey and other developing countries that would not have been possible two or three decades ago. Finally, in recent years, economists and economic historians have made an important distinction between the proximate causes and the deeper determinants of economic development. While literature on the proximate causes of development focuses on investment, accumulation of inputs, technology, and productivity, discussions of the deeper causes consider the broader social, political, and institutional environment. Both sets of arguments are utilized in my book.

I argue that an interest-based explanation can address both the causes of long-run economic growth and its limits. Turkey’s formal economic institutions and economic policies underwent extensive change during the last two centuries. In each of the four historical periods I define, Turkey’s economic institutions and policies were influenced by international or global rules which were enforced either by the leading global powers or, more recently, by international agencies. Additionally, since the nineteenth century, elites in Turkey made extensive changes to formal political institutions.  In response to European military and economic advances, the Ottoman elites adopted a programme of institutional changes that mirrored European developments; this programme  continued during the twentieth century. Such fundamental  changes helped foster significant increases in per capita income as well as  major improvements in health and education.

But it is also necessary to examine how these new formal institutions interacted with the process of economic change – for example, changing social structure and variations in the distribution of power and expectations — to understand the scale and characteristics of growth that the new institutional configurations generated.

These interactions were complex. It is not easy to ascribe the outcomes created in Turkey during these two centuries to a single cause. Nonetheless, it is safe to state that in each of the four periods, the successful development of  new institutions depended on the state making use of the different powers and capacities of the various elites. More generally, economic outcomes depended closely on the nature of the political order and the degree of understanding between different groups in society and the elites that led them. However, one of the more important characteristics of Turkey’s social structure has been the recurrence of tensions and cleavages between its elites. While they often appeared to be based on culture, these tensions overlapped with competing economic interests which were, in turn, shaped by the economic institutions and policies generated by the global economic system. When political institutions could not manage these tensions well, Turkey’s economic outcomes remained close to the world average.

Loans of the Revolution: How Mexico Borrowed as the State Collapsed in 1912–13

by Leonardo Weller (São Paulo School of Economics – FGV)

Read the full article on The Economic History Review – published in Augus 2018, available here

 

Mexico borrowed £6 million abroad in 1913, amidst a civil war that destroyed the state and killed over two million people. Civil wars tend to make creditors wary because of their inevitable consequences: if the borrowing government wins, it will need to spend on reconstruction, leaving it short of cash to pay their debt back; in case of defeat, the new incoming government is bound to repudiate its enemy’s debt. The Mexican loan of 1913 is unusual because the bankers in charge knew that the government was likely to lose or, at best, fight a long and bloody war. Paribas, the head of the syndicate that underwrote the loan, received first-hand reports from an agent in Mexico City, according to whom:

 ‘The political situation is (. . .) obscure because the country is still infested by rebellious bands while General Huerta is only president of the republic in a provisory character (…), and no one knows how that will end’.

Victoriano Huerta assassinated Francisco Madero, leader of the revolution who deposed the long-serving autocrat Porfirio Díaz who was in office at various times between the 1870s and 1911. The report from the Mexican agent was accurate: Huerta aimed to re-established Díaz’s stable regime, but his counter-revolution fostered an unlikely but powerful alliance between popular insurgents such as Emiliano Zapata and moderate politicians linked to Madero. Pressured by the financial toll of the war, the Huerta administration defaulted on its entire debt – including the loan take out 1913 – in 1914.   The insurgents took Mexico City and deposed the dictator, but the war continued and Mexico became a failed state. Peace only came in 1917, but the government in charge of reconstructing the country did not pay the sovereign debt.

LUCHA
Figure 1. Lucha Revolucionaria. Source: Diego Rivera, Palacio Nacional, Mexico City.

Paribas and its fellow syndicate members underwrote the 1913 loan at rather poor borrowing conditions: 6 per cent interest, 90 per cent discount, and 10 years maturity, which result in a 4 per cent risk premium (a measure of credit cost), twice higher than the premium applied to the Mexican debt already floating on the London Stock Exchange at the time. In plain language, the loan was remarkably expensive vis-à-vis market conditions. This discrepancy appears in the graph below: The solid line is the premium at which the secondary market traded old Mexican bonds, and the dot is the premium at which the banks issued the new loan in 1913.

 

Table 01. Mexican risk and reports in The Times in Mexico.

2
Sources: Calculated from Investor’s Monthly Manual, Bulletin de la Côte, The Times, 1911–14.

 

The bankers themselves considered the loan as ‘too severe a burden on the Government’, but agreed that the operation had to be arranged ‘in our favour’ because of the ‘terrible political circumstances’ in the country. In line with this dire assessment of Mexican political conditions, Paribas sold its entire share of the 1913 loan. In fact, Paribas acted solely as an underwriter to float the bonds on the international market, as opposed to underwriter and final creditor (holding a share of the bonds). The bank also liquidated all its other Mexican assets it held, including a significant share of the Banco Nacional de México. A conflict of interest explains why Paribas underwrote the 1913 loan: The new credit created confidence among the public and sustained the price of all Mexican securities, which enabled Paribas to eliminate its exposure to Mexico without realising losses. The liquidation was profitable overall: in particular, the bank sold the 1913 bonds at a 6 per cent margin.

Undoubtedly, Paribas’ gains were at the expense of others. Why, then, did bondholders agree to purchase the debt? Figure 1 suggests (and econometric tests confirm) that positive press reports influenced the market. The Times published negative news on Mexico when the revolutionaries deposed Díaz and the counter-revolutionaries assassinated Madero, in 1911 and 1913, respectively, but it subsequently altered its editorial stance by publishing good news and generally remaining going quiet on the country. Meanwhile, bondholders continued buying Mexican bonds in spite of the civil war and, as a result, Mexican risk stayed below 2 per cent, a relatively low rate.

The public read over-optimistic news, while the bankers had access to pessimistic but accurate reports from their agents in Mexico. Thus, Paribas benefited from asymmetric information, which explains why it could profit at the expenses of the final creditors.

This case study is at odds with the most recent historical literature on sovereign debt, which stresses the role of debt underwriters as gatekeepers, responsible for guiding the market. The literature asserts that bankers produced signals that separated the trustworthy borrowers from the rest. In contrast, Paribas exploited the market´s disinformation to profit from the liquidation of its Mexican businesses.

 

To contact the author:

leonardo.weller@fgv.br

@leoweller

 

Did efficient options pricing lead or follow the development of the Black Scholes Merton model? Evidence from the interwar London Metals Exchange

by David Chambers and Rasheed Saleuddin (Judge Business School, University of Cambridge)

This research is due to be published in the Economic History Review and is currently available on Early View

 

In early 1998 a nervous options trader was asked to fill an order from one of the world’s largest global investors. The fund’s manager, believing that Canadian short term rates would fall in the very near future, wanted to buy the option – but not the obligation – to buy two year bonds for the next two weeks at a ‘strike’ price of 103 per cent of par when they were actually trading at 102. If rates fell enough upon the option’s expiry in two weeks, the bond would trade above 103, and the hedge fund would pocket the difference between the actual price and the strike of 102.

The average volume in options on these bonds for the week was probably a few hundred million dollars of par value, but this client was looking for options on $2 billion. It would be impossible for the trader to find the exact matching trade in the market from another client, and he would have to ‘manufacture’ the options himself. How, then, to calculate the price?

The framework for this analysis was largely developed by Fisher Black, Myron Scholes and Robert Merton, as published in 1972 (Black and Scholes 1972, Merton 1973). But one crucial input into the so-called Black-Scholes-Merton (BSM) model was difficult to estimate: expected future volatility (as measured by standard deviation of returns) of the underlying bond over the next two weeks. The trader looked first to the recent past: What had the two-week volatility been over the past few months? The trader also knew that unemployment numbers were due out in three days and that some uncertainty always surrounds such a release. In the end, the trader used the BSM model and a volatility input of slightly higher than the past two weeks’ observations to account for the uncertainty in unemployment and the large size of the trade. Upon execution, the trader then used the BSM model to calculate the amount of the underlying bond to sell short to hedge some of the risks to the original trade. That trader was one of the co-authors of an article — ‘Commodities option pricing efficiency before Black, Scholes and Merton’ — recently published in the Review. In this study, the authors David Chambers and Rasheed Saleuddin examine a commodity futures options market for the interwar period to determine how traders might have made markets in options before the advent of modern models.

It is often thought that market prices conform to newly-implemented models rather than obeying some natural laws of markets before such laws are revealed to observers. It has been suggested that equity options, specifically, were ‘performative’ in that they converged to BSM-efficient levels shortly after the dissemination of this model in the early 1970s (MacKenzie and Millo 2003). On the other hand, some claim that it was the advent of liquid exchange trading around the same time that led to BSM efficiency (Kairys and Valerio 1997).

Evidence of efficient pricing before the 1970s is sparse and mixed. There are very few data sets with which to test efficiency, and the few that have been used are far from ideal. Two papers (Kairys and Valerio 1997, Mixon 2009) use one-sided indicative advertised levels targeted to retail investors, without any indication that these were prices upon which investors traded. Another paper uses primarily warrant data, yet the prices of warrants, even in modern times, are often far from BSM efficient, for well-understood reasons (Veld 2003). In any event, these studies find that, on average, prices were far from BSM efficient levels. There is little attempt in this early literature to determine if prices were dependent on the most important BSM model parameter –observed volatility.

This study uses a new data set: prices at which the economist John Maynard Keynes traded options on tin and copper futures traded on the interwar London Metals Exchange. In turns out that Keynes traded at levels that were – on average – as efficient as modern markets. Additionally, the traded prices appear to have varied systematically with the key input to the model, observed volatility (Figure 1), with 99% significance and very high R2.

Untitled
Figure 1. Scatter Diagram of Implied Volatility vs. Historic Volatility of 3-month options on Tin and Copper futures, 1921-31. Source: Chambers and Saleuddin (2019)

How was it possible that Keynes’ traders and brokers were able to match BSM efficient prices so closely? There is some suggestion that options traders in the 19th and early 20th centuries well understood options theory. Indeed, Anne Murphy (2009) had identified a perhaps surprising degree of sophistication and activity among the options traders in 17th century London. Certainly, by the turn of the previous century, options traders had a strong grasp of many of the fundamentals of options trading and pricing (Higgins 1907). Yet current understanding of the influence of volatility of the underlying asset was still in its infancy and several contemporary breakthroughs in theory were not disseminated widely. Finance scholarship hints at one possible explanation: For options such as those traded by Keynes, the relationship between the key BSM valuation parameter, volatility, and option price are quite straightforward to estimate (Brenner and Subrahmanyam 1988). It may have been the case that market participants were intuitively taking into account BSM without an understanding of the model itself. This conclusion is, of course, pure speculation – but perhaps therein lies its fascination?

To contact the authors:

David Chambers:
d.chambers@jbs.cam.uc.uk

Rasheed Saleuddin,
Rks66@me.com
@r_sale.

It is only cheating if you get caught – Creative accounting at the Bank of England in the 1960s

by Alain Naef (Postdoctoral fellow at the University of California, Berkeley)

This research was presented at the EHS conference in Keele in 2018 and is available as a working paper here. It is also available as an updated 2019 version here.

 

Naef 3
The Bank of England. Available at Wikimedia Commons.

The 1960s were a period of crisis for the pound. Britain was on a fixed exchange rate system and needed to defend its currency with intervention on the foreign exchange market. To avoid a crisis, the Bank of England resorted to ‘window dressing’ the published reserve figures.

In the 1960s, the Bank came under pressure from two sides: first, publication of the Radcliffe report (https://en.wikipedia.org/wiki/Radcliffe_report) forced publication of more transparent accounts. Second, with removal of capital controls in 1958, the Bank came under attack from international speculators (Schenk 2010). These contradictory pressures put the Bank in an awkward position. It needed to publish its reserve position (holdings of dollars and gold ) but it recognised that doing so could trigger a run on sterling, thereby creating a self-fulfilling currency crisis (see Krugman: http://www.nber.org/chapters/c11032.pdf).

For a long time, the Bank had a reputation for the obscurity of its accounts and its lack of transparency. Andy Haldane (Chief Economist at the Bank) recognised, for ‘most of [it’s] history, opacity has been deeply ingrained in central banks’ psyche’.

(https://www.bankofengland.co.uk/speech/2017/a-little-more-conversation-a-little-less-action). One Federal Reserve (Fed) memo noted that the Bank of England took ‘a certain pride in pointing out that hardly anything can be inferred by outsiders from their balance sheet’, another that ‘it seems clear that the Bank of England is being pushed – by much public criticism – into giving out more information.’ However, the Bank did eventually publish reserve figures at a quarterly, and then monthly, frequency (Figure 1).

Transparency about the reserves created a risk for a currency crisis so in late 1966 the Bank developed a strategy for reporting levels that would not cause a crisis (Capie 2010). Figure 1 illustrates how ‘window dressing’ worked. The solid line reports the convertible reserves as published in the Quarterly Bulletin of the Bank of England. This information was available to market participants. The stacked columns show the actual daily dollar reserves. Spikes appear at monthly intervals, indicating the short-term borrowing that was used to ensure the reserves level was high enough on reporting days.

 

Figure 1. Published EEA convertible currency reserves vs. actual dollar reserves held at the EEA, 1962-1971.

Naef 1

 

The Bank borrowed dollars shortly before the reserve reporting day by drawing on swap lines (similar to the Fed in 2007 https://voxeu.org/article/central-bank-swap-lines). Swap drawings could be used overnight. Table 1 illustrates how window dressing worked using data from the EEA ledgers available at the archives of the Bank. As an example, on Friday, 31 May 1968, the Bank borrowed over £450 million – an increase in reserves of 171%. The swap operation was reversed the next working day, and on Tuesday the reserves level was back to where it was before reporting. The details of these operations emphasise how swap networks were short-term instruments to manipulate published figures.

 

Table 1. Daily entry in the EEA ledger showing how window dressing worked

Naef 2

 

The Bank of England’s window dressing was done in collaboration with the Fed. Both discussed reserve figures before the Bank published them. During most of the 1960s, the Bank and the Fed were in contact daily about exchange rate matters. Records of these phone conversations are parsimonious at the Bank but the Fed kept daily records (Archives of the Fed in New York, references 617031 and 617015).

During the 1960s, collaboration between the two central banks intensified. The Bank consulted the Fed on the exact wording of the reserve publication (Naef, 2019) and the Fed communicated on the swap position with the Bank, to ensure consonance between the public statements. Indeed, the Fed sent excerpts of minutes to the Bank to allow excision of anything mentioning window dressing (Archives of the Fed in New York, reference 107320). Thus, in December 1971, before publishing the minutes of the Federal Open Market Committee (FOMC) for 1966, Charles Coombs (a leading figure at the Fed) consulted Richard Hallet (Chief Cashier at the Bank):

‘You will recall that when you visited us in December 1969, we invited you to look over selected excerpts from the 1966 FOMC minutes involving certain delicate points that we thought you might wish to have deleted from the published version. We have subsequently deleted all of the passages which you found troublesome. Recently, we have made a final review of the minutes and have turned up one other passage that I am not certain you had an opportunity to go over. I am enclosing a copy of the excerpt, with possible deletions bracketed in red ink.’

Source: Letter from Coombs to Hallet, New York Federal Reserve Bank archives, 1 December 1971, Box 107320.)

 

Coombs suggested deleting passages where some FOMC members criticised window dressing, while other members suggested the Bank would get better results ‘if they reported their reserve position accurately than if they attempted to conceal their true reserve position’ (https://fraser.stlouisfed.org/scribd/?item_id=22913&filepath=/docs/historical/FOMC/meetingdocuments/19660628Minutesv.pdf). However, MacLaury (FOMC), stressed that there was a risk of ‘setting off a cycle of speculation against sterling’ if the Bank published a loss of $200 million, which was ‘large for a single month’ in comparison with what was published the previous month.

The history of the Bank’s window dressing is a reminder of the difficulties central banks face in managing reserves, a situation similar to how investors today closely monitor the reserves of the People’s Bank of China.

 

 

To contact the author: alain.naef@berkeley.edu

 

References:

Capie, Forrest. 2010. The Bank of England: 1950s to 1979. Cambridge: Cambridge University Press.

Naef, Alain. 2019. “Dirty Float or Clean Intervention?  The Bank of England in the Foreign Exchange Market.” Lund Papers in Economic History. General Issues, no. 2019:199. http://lup.lub.lu.se/record/dfe46e60-6dfb-4380-8354-e7b699ed8ef9.

Schenk, Catherine. 2010. The Decline of Sterling: Managing the Retreat of an International Currency, 1945–1992. Cambridge University Press.

All quiet before the take-off? Pre-industrial regional inequality in Sweden (1571-1850)

by Anna Missiaia and Kersten Enflo (Lund University)

This research is due to be published in the Economic History Review and is currently available on Early View.

 

Missiaia Main.jpg
Södra Bancohuset (The Southern National Bank Building), Stockholm. Available here at Wikimedia Commons.

For a long time, scholars have thought about regional inequality merely as a by-product of modern economic growth: following a Kuznets-style interpretation, the front-running regions increase their income levels and regional inequality during industrialization; and it is only when the other regions catch-up that overall regional inequality decreases and completes the inverted-U shaped pattern. But early empirical research on this theme was largely focused on the  the 20th century, ignoring industrial take-off of many countries (Williamson, 1965).  More recent empirical studies have pushed the temporal boundary back to the mid-19th century, finding that inequality in regional GDP was already high at the outset of modern industrialization (see for instance Rosés et al., 2010 on Spain and Felice, 2018 on Italy).

The main constraint for taking the estimations well into the pre-industrial period is the availability of suitable regional sources. The exceptional quality of Swedish sources allowed us for the first time to estimate a dataset of regional GDP for a European economy going back to the 16th century (Enflo and Missiaia, 2018). The estimates used here for 1571 are largely based on a one-off tax proportional to the yearly production: the Swedish Crown imposed this tax on all Swedish citizens in order to pay a ransom for the strategic Älvsborg castle that had just been conquered by Denmark. For the period 1750-1850, the estimates rely on standard population censuses. By connecting the new series to the existing ones from 1860 onwards by Enflo et al. (2014), we obtain the longest regional GDP series for any given country.

We find that inequality increased dramatically between 1571 and 1750 and remained high until the mid-19th century. Thereafter, it declined during the modern industrialization of the country (Figure 1). Our results discard the traditional  view that regional divergence can only originate during an industrial take-off.

 

Figure 1. Coefficient of variation of GDP per capita across Swedish counties, 1571-2010.

Missiaia 1
Sources: 1571-1850: Enflo and. Missiaia, ‘Regional GDP estimates for Sweden, 1571-1850’; 1860-2010: Enflo et al, ‘Swedish regional GDP 1855-2000 and Rosés and Wolf, ‘The Economic Development of Europe’s Regions’.

 

Figure 2 shows the relative disparities in four benchmark years. If the country appeared relatively equal in 1571, between 1750 and 1850 both the mining districts in central and northern Sweden and the port cities of Stockholm and Gothenburg emerged.

 

Figure 2. The relative evolution of GDP per capita, 1571-1850 (Sweden=100).

Missiaia 2
Sources: 1571-1850: Enflo and. Missiaia, ‘Regional GDP estimates for Sweden, 1571-1850’; 2010: Rosés and Wolf, ‘The Economic Development of Europe’s Regions’.

The second part of the paper is devoted to the study of the drivers of pre-industrial regional inequality. Decomposing the Theil index for GDP per worker, we show that regional inequality was driven by structural change, meaning that regions diverged because they specialized in different sectors. A handful of regions specialized in either early manufacturing or in mining, both with a much higher productivity per worker compared to agriculture.

To explain this different trajectory, we use a theoretical framework introduced by Strulik and Weisdorf (2008) in the context of the British Industrial Revolution: in regions with a higher share of GDP in agriculture, technological advancements lead to productivity improvements but also to a proportional increase in population, impeding the growth in GDP per capita as in a classic Malthusian framework. Regions with a higher share of GDP in industry, on the other hand, experienced limited population growth due to the increasing relative price of children, leading to a higher level of GDP per capita. Regional inequality in this framework arises from a different role of the Malthusian mechanism in the two sectors.

Our work speaks to a growing literature on the origin of regional divergence and represents the first effort to perform this type of analysis before the 19th century.

 

To contact the authors:

anna.missiaia@ekh.lu.se

kerstin.enflo@ekh.lu.se

 

References

Enflo, K. and Missiaia, A., ‘Regional GDP estimates for Sweden, 1571-1850’, Historical Methods, 51(2018), 115-137.

Enflo, K., Henning, M. and Schön, L., ‘Swedish regional GDP 1855-2000 Estimations and general trends in the Swedish regional system’, Research in Economic History, 30(2014), pp. 47-89.

Felice, E., ‘The roots of a dual equilibrium: GDP, productivity, and structural change in the Italian regions in the long run (1871-2011)’, European Review of Economic History, (2018), forthcoming.

Rosés, J., Martínez-Galarraga, J. and Tirado, D., ‘The upswing of regional income inequality in Spain (1860–1930)’,  Explorations in Economic History, 47(2010), pp. 244-257.

Strulik, H., and J. Weisdorf. ‘Population, food, and knowledge: a simple unified growth theory.’ Journal of Economic Growth 13.3 (2008): 195.

Williamson, J., ‘Regional Inequality and the Process of National Development: A Description of the Patterns’, Economic Development and Cultural Change 13(1965), pp. 1-84.