Lychakov, Nikita
Working Paper
Government-made bank distress: Industrialisation
policies and the Russian financial crisis of 1899-1902
QUCEH Working Paper Series, No. 2018-11
Provided in Cooperation with:
Queen's University Centre for Economic History (QUCEH), Queen's University Belfast
Suggested Citation: Lychakov, Nikita (2018) : Government-made bank distress: Industrialisation
policies and the Russian financial crisis of 1899-1902, QUCEH Working Paper Series, No. 2018-11,
Queen's University Centre for Economic History (QUCEH), Belfast
This Version is available at:
https://hdl.handle.net/10419/186121
Standard-Nutzungsbedingungen:
Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen
Zwecken und zum Privatgebrauch gespeichert und kopiert werden.
Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle
Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich
machen, vertreiben oder anderweitig nutzen.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen
(insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten,
gelten abweichend von diesen Nutzungsbedingungen die in der dort
genannten Lizenz gewährten Nutzungsrechte.
Terms of use:
Documents in EconStor may be saved and copied for your personal
and scholarly purposes.
You are not to copy documents for public or commercial purposes, to
exhibit the documents publicly, to make them publicly available on the
internet, or to distribute or otherwise use the documents in public.
If the documents have been made available under an Open Content
Licence (especially Creative Commons Licences), you may exercise
further usage rights as specified in the indicated licence.
QUCEH WORKING PAPER SERIES
http://www.quceh.org.uk/working-papers
GOVERNMENT-MADE BANK DISTRESS:
INDUSTRIALISATION POLICIES AND THE RUSSIAN
FINANCIAL CRISIS OF 1899-1902
Nikita Lychakov (Queen’s University Belfast)
Working Paper 2018-11
QUEEN’S UNIVERSITY CENTRE FOR ECONOMIC HISTORY
Queen’s University Belfast
185 Stranmillis Road
Belfast BT9 5EE
November 2018
Government-made bank distress:
industrialisation policies and the Russian financial crisis of 1899-1902
Nikita Lychakov *
Queen’s University Belfast
How can industrial policies lead to bank distress? In the 1890s, when undergoing rapid
state-led industrialisation, the Russian Empire grew by foreign capital inflows into national
debt and by state procurement of industrial output. Concurrently, state policies incentivised,
but did not compel, commercial banks to finance industry. In 1899, the inflow of foreign
capital fell sharply, initiating a financial crisis. Using newly-collected historical data and
extensive narrative evidence, I find the banks which experienced greater distress in the crisis
had more personal connections to the government officials who were close to the epicentre
of policymaking. Moreover, these banks had more personal ties to the companies which had
been most-stimulated by state policies to expand production. Taken together, these two
findings suggest that national development policies had a destabilising impact on bank
performance.
Keywords: financial crises, bank failures, development policies, political economy,
Russia.
JEL Classification: G01, L5, O25, P16.
* E-mail: [email protected]
I thank Chris Colvin and John Turner for excellent advice and Alan de Bromhead and Lucy
Newton for valuable feedback. I am also grateful for helpful comments to participants at the
EHS residential course, conference poster session and annual conference; Queen’s
University Belfast poster session, postgraduate seminar and QUCEH seminar; UCL WEast
Workshop; Oxford University economic and social history seminar; the 8th World Congress
of Cliometrics; the 12th EHES conference; the 77th EHA conference poster session; and the
19th April International Academic Conference.
1
1. Introduction
In the mid- to late 1890s, industry in the Russian Empire showed exceptional growth. The
value of manufactured output between 1893 and 1900 rose by nearly 10 per cent per year
(Borodkin, 2011a, 2011c), making it the fifth largest manufacturing producer in the world
(League of Nations, 1945, p. 13). The state acted as a catalyst for economic growth by
procuring industrial products and stimulating industry to expand operations to match the
constantly-rising demands from the state and the private sector (Gindin, 2007c).
Simultaneously, I find evidence that state policies incentivised commercial banks to finance
this industrial expansion. In 1899, this successful setup started to come apart due to a sudden
drop in the rate of inflow of European capital to Russian government bonds and the securities
of industrial enterprises. This led to a rapid stock market decline, disastrous corporate
performance and overwhelming bank losses (Bovykin, 1967, 1984). Because banks supplied
over half of the industrial sector’s financing, banking distress threatened the successes of the
whole decade.
What caused the reversal in the robust industrial growth? Among a number of
possible causes, one reason offered by the then-Minister of Finance was the state
industrialisation policies of the 1890s, which had over-stimulated heavy industry, leading to
its overexpansion and eventual contraction (Gindin, 1996). The purpose of this paper is to
examine how far government policies contributed to bank losses in the crisis of 1899-1902, if
at all. This enquiry is important because the influence of the state on a banking system is
often significant and can be negative. When it is negative, banking distress and failure may
follow, which is likely to lead to a fall in the supply of money (Friedman & Schwartz, 1963)
and other non-monetary ill effects (Bernanke, 1983; Hall, 2010) on economic activity overall.
I identify three channels through which the Russian government affected the banks’
inclination to finance industry and, consequently, their financial outcomes. The first was
information that bank staff gathered through their personal connections with government
2
officials and corporate board members. In particular, banks obtained information about the
future direction of industrialisation policies, such as which industries the government would
protect and support, and about state procurement decisions, such as which factory would
receive a lucrative procurement contract. The second channel was the government’s
expectation that banks would finance the modernisation of industry. The third was the
development policy of the State Bank, a quasi-central bank, which provided commercial
banks with regular credit. This paper focuses on the first channel because I argue the other
two channels contributed little to the increase in the banks’ financing of industry.
In order to trace the flow of information from the state to the banks, I record personal
connections, or more formally interlocks, between members at banking boards, government
officials, and company board members.
1
Specifically, a well-informed political interlock was
recorded when a bank board member, or his sibling, was part of a group of top government
officials who were well-informed – due to their proximity to the Finance Minister Sergei
Witte – about state procurement decisions and the expected trajectory of the industrialisation
policies. Witte, as the architect of the industrialisation policies and the final decision-maker
on state procurement contracts, was at the epicentre of policymaking (Bovykin, 1967;
Shepelev, 1981).
A quite different heavy industry interlock is recorded when a bank’s board member
was also a board member in a heavy industrial firm, or when that banker’s sibling served this
function. This type of interlock presents an alternative way of tracing the effect of state
policies on bank performance, because heavy industrial firms were highly reliant on state
procurement; when the crisis caused procurement to fall, industry suffered. Moreover, the
government itself acknowledged the danger posed by the existence of corporate interlocks. In
1
For lucidity, I refer to the members of a board of directors or a management committee as simply
‘board members’, unless otherwise specified.
3
one of the few corporate laws passed in response to the crisis, authorities forbade an
executive board member of a bank to serve on a corporation’s board (Complete Collection of
Laws, 1903). I also specify three additional types of interlock: to capture bankers’
connections with a wider range of government officials and with non-financial companies,
and to trace connections with rival banks.
Next, I assess how the presence of government, industrial, and banking connections in
each bank affected bank performance during the crisis. Bank distress is measured in the form
of: (1) the net losses sustained by banks on investment portfolios over the whole crisis period;
and (2) the drop in bank share prices over the first year of the crisis, the only period during
the crisis that was largely free from direct government intervention in the stock market.
My main finding is that the banks which experienced greater distress during the crisis
had more connections to the government officials who were well-informed about the
industrialisation policies and/or state procurement contacts in the lead-up to the crisis. The
inference is that the banks which had been most influenced by these policies exposed their
finances most to heavy industrial companies and consequently lost most.
Additionally, I find that the banks which experienced greater distress during the crisis
had more personal connections to heavy industrial companies. The interpretation is that those
banks which suffered heavier losses had greatest exposure, via direct loan financing or
securities underwriting, to the companies most stimulated by state policies. Put together,
these two complementary findings indicate the negative effect of the national development
policies on bank performance during Russia’s state-led growth.
The findings of this paper add to four strands of existing literature. First, the paper
provides empirical evidence to support the long-standing belief that government can
contribute to, and even originate, banking crises. Research on this subject comes in the form
of descriptive evidence (Hammond, 1957; Cameron, 1967; Krugman, 1998; Cassis, 2002;
4
Rajan, 2011; Calomiris & Haber, 2014; Turner, 2014). For example, Calomiris & Haber
(2014) argue that the root cause of the 2007-09 financial crisis in the U.S. was the deal struck
between banks and government to loan money to unqualified mortgage borrows in exchange
for relaxed banking regulation. The unique contribution of this paper is that it educes
qualitative evidence and takes a quantitative approach to link government policies with
subsequent bank distress. My statistical analysis adds to and confirms the narrative evidence,
rather than merely claiming causality without substance.
Second, this paper augments Russian economic history. Alexander Gerschenkron
maintained that the Russian state was the ‘agens movens of industrialisation’ (Gerschenkron,
1962, p. 20), or the primary driver of economic development, largely responsible for the
country’s economic outcomes. The present research supports Gerschenkron’s arguments, and
at the same time questions Gregory’s (1991, 1994) revisionist view that the Russian state was
not a vital player in the industrialisation of the 1890s. Specifically, Gregory contends that
state expenditures on industrial products, such as military hardware, were too small to affect
industrial growth unduly. This paper shows that state procurement accounted for up to 50 per
cent of heavy industry output.
Third, this paper throws new light on the impact of bankers’ personal connections on
bank performance. Researchers have found political interlocks can have both a negative
(Duchin & Sosyura, 2012; Grossman & Imai, 2016) and positive impact (Braun & Raddatz,
2010; Acemoglu et al., 2016) on the financial outcomes of banks. This paper shows that a
negative influence is indeed a possibility. When it comes to banks’ interlocks with non-
financial firms, my research supports the established view that in times of stress this type of
connection typically has a negative effect on banks (Laeven, 2001; La Porta et al., 2003;
Colvin, 2014; Colvin et al., 2015). A distinguishing feature of this paper is that it identifies
exactly the type of information that passed through personal contacts and produced the
5
quantified effect. In contrast, nearly all academic articles on interlocks do not go beyond
hypothesising why interlocks have a negative or positive effect.
Fourth, this paper contributes to the historical literature on the role of bank-industry
connections in the economic growth of the nineteenth century. Although banks’ interlocks
with companies brought little tangible benefits during Germany’s industrialisation (Fohlin,
1999), they were essential for the economic development of New England (Lamoreaux,
1996), and improved firms’ access to credit in South Yorkshire (Newton, 1996). This paper’s
findings suggest that bank-industry ties played a significant role in the development of
Russia’s heavy industry.
The arguments are built as follows. Section 2 provides the historical setting of the
study. Section 3 outlines the proximate causes of the crisis. Section 4 quantifies government’s
role in expanding industrial production in the 1890s. Section 5 examines the information
channel through which government affected banks’ willingness to finance industry. Section 6
defines the empirical strategy and variables. Sections 7 reports empirical results and Section 8
concludes. Appendices provide summary statistics and robustness tests, explain the choice
behind dependent variables, and discuss the two other channels through which government
might have influenced banks’ investment and lending decisions.
2. Industrialisation in the 1890s
In the 1880s, Russia’s share of world manufacturing production was far behind those of other
major nations. Whereas the U.K.’s share was 26.6 per cent and France’s 8.6 per cent,
Russia’s was only 3.4 per cent (League of Nations, 1945). In 1893, the newly-appointed
Finance Minister, Sergei Witte, introduced fresh economic policies aimed at rapid
industrialisation. The goal was to catch up with the already industrialised countries in Europe
and make Russia great again in the world economic, political and military arena (Gindin,
6
1957). Witte believed an economy based solely on agriculture was at the mercy of ‘a quite
obvious, unsheltered exploitation of its productive forces’ (Gindin, 1957, p. 154) by the
already industrialised countries.
2
In fact, the humiliating fiasco in the Crimean War of 1853-
56 and the diplomatic restrictions imposed on Russia at the Congress of Berlin in 1878
vividly showed that economic backwardness and military defeats are closely related (Crisp,
1976).
Along with foreign investors, whose role is discussed in the next section, 39 joint-
stock commercial banks financed Russia’s industrial growth. Banks supplied well over half
the industrial financing, according to my estimates.
3
Ten banks were headquartered in St.
Petersburg, the capital, where the government was located; five were based in Moscow,
which was the centre of light industry; while the rest operated in provincial centres, from
present-day Poland to Siberia. Appendix Table A1 lists the names and characteristics of
banks.
Rapid industrialisation turned out to be a success. Figure 1 shows that, throughout the
1890s, 745 public companies were established, as opposed to 216 firms in the prior decade.
4
Between 1893 and 1900, the value of output manufactured by heavy and light industry rose
by 87 and 52 per cent, respectively (Borodkin, 2011a, 2011c). By 1900, Russia was
producing 5.0 per cent of global manufacturing output, which ranked close to France, the
world’s fourth largest manufacturer (League of Nations, 1945).
2
Author’s translation, as are all other instances of text translated from Russian language.
3
I estimate that banks supplied some 61 per cent of joint-stock company financing. The banks’ role in
financing heavy industrial companies was probably greater still. For comparison, Bovykin (1967)
estimates that banks supplied half of all heavy industry’s finance. Calculated from data in Bovykin
(1894); Dmitriev-Mamonov (1903); Golubev (1905); State Bank (1899); and Russian State Historical
Archive (RSHA) (collection 587, inventory 33, file 98).
4
Calculated from data in Dmitriev-Mamonov (1903).
7
3. The proximate causes of the crisis
The sound industrial growth of the 1890s was interrupted by the financial crisis of 1899-
1902, which began with a much reduced inflow of foreign capital to Russian government
bonds and the securities of industrial enterprises. Beginning from 1899, the Russian
government found it almost impossible to sell its debt on the Parisian market, which had been
Russia’s primary source of external funding (Ministry of Finance, 1900-1907b).
5
The rate of
capital inflows to corporate equity and debt securities also began to decelerate: from a 35 per
cent increase in 1899 to just 1 per cent in 1902 (Ol, 1925). This contrasts markedly with the
interest expressed by French, Belgian, and German investors in the years before the crisis
who saw in Russia a highly attractive opportunity. The net amount of foreign capital from
5
Gregory (2003) estimates that in 1898 foreigners owned as much as 57 per cent of government debt.
A bond of 159 million rubles was placed with French investors in May 1901 and another 182 million
rubles in April 1902, together equivalent to 17.5 per cent of state revenues in 1902, but that did not
make much difference to national finances (Siegel, 2015). State revenues are from Ministry of
Finance (1902d).
0
50
100
150
200
250
1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900
NUMBER OF JOINT-STOCK
COMPANIES ESTABLI
SHED
FINANCIAL CRISIS
Figure 1. Number of joint-stock companies established per year, 1890-1900
Sources: Dmitriev-Mamonov (1903).
8
abroad soared by over 66 per cent between 1893 and 1900, from 2,951 to 4,910 million
rubles (Bovykin, 1984).
6
The decline in foreign capital inflows was due to the shortage of capital in Europe,
itself attributable to various economic and political factors (Ministry of Finance, 1900a;
Witte, 1898b). The scarcity of capital was plain from the rise of official interest rates across
Europe as early as the summer of 1898. In a note to the Czar, Witte mentioned that on 29
October 1898, German interest rates stood at 5.5 per cent, unprecedented since the Baring
crisis of 1890 (Witte, 1898b). The fact that interest rates in Europe first rose six months
before the crisis began in Russia suggests that the sudden stop in capital flows was unrelated
to Russia’s industrial performance. In October 1899, the Second Boer War broke out,
depriving Great Britain of access to considerable gold supplies in South Africa, which further
tightened capital constrains on Europe’s money markets (Ministry of Finance, 1900a).
Consequently, the St. Petersburg stock exchange index began a three-year decline
from February 1899 to the end of 1901, when the index had fallen by 45.4 per cent.
7
In 1901-
02, heavy industry went into recession and its output declined by 7.7 per cent.
8
Three banks
failed during the crisis and two banks shortly after, while 35 other banks survived through
intensive help from the State Bank (RSHA, 587, 33, 101, 12-13). Despite these efforts, the
crisis lasted longer in Russia than parallel industrial crises in Germany and Belgium (Brandt,
1902). The reason for its duration, according to the contemporary view of the Association of
Industrial and Trade Enterprises, was primarily the government’s action, which the
6
The net flow of foreign capital of 1,959 million rubles between 1893 and 1900 was equivalent to
208.6 million GBP, or 10.6 per cent of the nominal British GDP in 1900. Foreign exchange is
calculated as the average rate on a 3-month bill of exchange drawn from St. Petersburg on London
between 1893 and 1900. Data on the rates are from Borodkin (2011b). GDP is from Thomas &
Williamson (2018).
7
Calculated from data in Goetzmann et al. (n. d.a).
8
Calculated from data in Borodkin (2011a, 2011b).
9
association believed, ‘exacerbate[d] the crisis in every possible way. In periods of industrial
recessions, government, one of the largest buyers, reduce[d] its procurement by an
exceptionally great extent. Such was the case with the construction of railways after 1900’
(Council of Congresses, 1914, p. 73). Other contemporary pundits agreed on these root
causes of the crisis (Ozerov, 1905; Kanel, 1906).
Other causes were thought by contemporaries to be insufficient consumer demand for
industry products, such as the use of passenger trains, that could have compensated for
declining state procurement (State Comptroller, 1902), and weak corporate and stock
exchange regulation (Witte, 1898a). Interestingly, the crisis was never attributed to
inadequate banking regulation, perhaps because Russia possessed one of the most stringent
regulatory and supervisory banking frameworks in Europe, according to a comparative study
by the Ministry of Finance (Gindin, 1960).
9
4. Government industrial policy
To determine the impact of state policies on bank performance, one must first identify how
far the government contributed both to the expansion of industrial production and the bank
financing of industry in the years leading up to the crisis. This section deals with the state’s
impact on industry.
Industrialisation policies created a protectionist environment which encouraged firms
to expand operations. Among the reasons for this stimulus were the state procurement of
industrial products which intensified after 1893; protective tariffs on industrial imports;
subsidies to key industries, including targeted loans from the state; the introduction of the
9
For example, the average capital ratio of St. Petersburg banks was over 25 per cent before the crisis,
on the basis of data in Golubev (1905). For comparison, capital ratios of UK banks in 1900 were
about 12 per cent (Sheppard, 2006). The exception was Germany, where credit banks in 1904 had a
capital to assets ratio of 44 per cent (Calomiris, 1995).
10
gold standard in 1897 to attract foreign capital to domestic enterprises; and official
encouragement to establish joint-stock companies (Gindin, 2007c). The push for rapid
industrialisation that began in 1893 is confirmed by a notably increased number of the joint-
stock companies which began to be set up, as shown in Figure 1.
In this policy of stimulus, state procurement was foremost. Constantly rising demand
from both state and private sectors throughout the 1890s motivated industrialists to expand
operations. In fact, the general press attributed the crisis primarily to industrial over-
expansion (Migulin, 1907). Perhaps no other evidence speaks stronger in support for this than
the one that comes directly from the Finance Minister. In a report at a special meeting on 10
April 1903, Witte, looking back at the crisis, acknowledged that the true cause of the
calamities to manufacturing industry had been the rapid railroad construction and industrial
development initiated by the government (Gindin, 1996). For twelve consecutive years
(1887-1899), strong demand for manufacturing products had outrun supply, encouraging
industrialists to constantly expand production capacity. In 1900, ‘the law of [continual]
growth in consumption’ (Gindin, 1996, p. 137) was broken and supply exceeded demand for
the first time. The overstretched financial position of factories, Witte added, put them at risk
of immediate failure (Gindin, 1996).
Moreover, the prospect of obtaining state orders often led firms to make over-
optimistic decisions, especially concerning the management of costs. Firms too small to
tackle state contracts sought to extend their operations solely to attract state orders. This
strategy often absorbed all the firms’ profits and drove them into debt (Afanasiev, 1900;
Gindin, 1996). A typical case was the growth and bankruptcy of a major mining enterprise,
which led to two of the three bank failures (Gertsenshtein, 1903).
10
10
Namely, the Alekseevski Mining and Joint-Stock Enterprise and the Kharkov Trade and
Ekaterynoslavsky commercial banks.
11
State procurement played a particularly important role because the state was a
dominant buyer. According to estimates by Varzar, a leading economist of the time, pig iron,
iron, steel and products manufactured from these materials across the country were in 1900
consumed in the following proportion: 48 per cent of the total by government, 22 per cent by
private enterprises, 14 per cent by the general population and 16 per cent by unidentified
parties (Ozerov, 1905). Note that the industries that extracted and employed these items in
their production accounted for over 62 per cent of the total heavy industry output.
11
Further
evidence comes from the Coalition of Southern Miners which, from their own data, indicated
that in 1899 the government’s share of procurement of such products from Southern mining
factories amounted to 47 per cent of the total; while a representative from the Coalition of
Polish Miners stated that government demand in his region amounted to a ‘mere’ 31 per cent
of the aggregate (Gindin, 1996, p. 74).
Qualitative records likewise indicate how heavily the private sector depended on state
orders. The report of the State Comptroller (1902, p. 55) to the Czar for 1901 states that
‘[c]urrently, there is little doubt that the crisis stemmed from the recent artificial and
excessive growth of manufacturing industry, which had originated in the protectionist policy,
large government orders and speculation using cheap foreign capital’.
While government procurement had a stimulating effect on industry in the 1890s, it
had the opposite effect during the crisis, because the state could no longer sustain
procurement at pre-crisis levels. In a memorandum to senior authorities, Zhukovski, the ex-
managing director of the State Bank, stressed that the crisis had resulted in a ‘lack of
government orders’ (Matveeva, 1987, pp. 52-53). The Economist, in its issue of 29 June
1901, reported that
11
Calculated from data in Borodkin (2011a, 2011b).
12
‘while production rapidly increased, the ordinary consumptive demand steadily declined and
when the Government, owing to its financial embarrassments, had to curtail its railway
building and its orders for material, the hot-house industries it had fostered collapsed and with
them the inflated market for industrial securities’ (p. 969).
5. Government information
Having established the role of government in industry’s expansion, this section now describes
the first channel through which the government affected banks’ willingness to finance
industry – namely, information about industrialisation policies that banks gathered through
personal connections, or interlocks. First, I identify which officials were most likely to
possess privileged information. Second, I examine how banks acted on privileged
information. Third, I document how banks leveraged their government connections. The
evidence suggests that banks were greatly influenced by such information.
Appendix Section C examines the two other channels through which government
affected banks’ willingness to finance industry: (1) the government’s expectation that banks
would finance industrial development; and (2) the State Bank’s policy vis-a-vis the banking
sector. The evidence suggests that these two channels played minor, if not negligible, role in
affecting banks’ decisions.
5.1. Government interlocks
The current literature suggests that banks set up interlocks with government officials in order
to benefit from officials’ industry-specific or managerial expertise; to enhance a firm’s
prestige from having a notable politician on their team; to gain access to preferential
treatment, such as protection from competitors; and to obtain preferential information on
future government actions, such as state procurement plans or regulatory changes (Faccio,
13
2006; Grossman & Imai, 2016).
12
The last two motives represent a form of rent-seeking that
can distort market competition to favour connected firms and their clients, which, in turn,
might impede overall economic growth.
Previous studies also suggest that banks establish interlocks with companies to reduce
information asymmetries and to influence clients’ corporate strategy (La Porta et al., 2003).
However, corporate connections do not come without risk. Corporate interlocks may develop
a strong influence on a bank and persuade it to finance dubious projects.
The Russian government affected banks’ decisions first through the privileged
information that they received about state procurement decisions and/or the future course of
its industrialisation policies. Banks obtained this information by establishing personal
connections with government officials and corporate board members.
Of the myriad of Russian government officials, some carried more weight and were
better informed than others. The central figure to the country’s development was Witte, the
architect of the industrialisation policies of the 1890s and the final decision-maker on
procurement contracts, corporate subsidies and other forms of economic support (Bovykin,
1967; Shepelev, 1981). One example of the centrality of Witte’s position was his drive to
consolidate power over the economic affairs of the nation in his own ‘enlightened
stewardship’ (Gindin, 1959, p. 123). Witte believed that this would ensure the successful
orchestration of industrialisation (Gindin, 1959). His power steadily grew as the Ministry of
Finance became a ‘super ministry’ (Solovyov, 2003, p. 165), whose reach expanded well
beyond its remit, for example, when it acquired absolute control over the Ministry of
Railways (Solovyov, 2003). Last, the historian Gindin (1996, 2007a, 2007b, 2007c, 2007d)
provides many examples of Witte’s omnipresent reach across Russia’s economic affairs.
12
For an extensive literature review on the role of interlocks at corporations, see Adams et al. (2010).
14
This evidence justifies the view that all major economic policy decisions emanated
from Witte and that he handled most of the state’s procurement contracts. From this, it may
be inferred that the top government officials who were in close and regular contact with Witte
were in the best position to obtain privileged information. In addition to these men, top
officials at the Ministry of Railways, Agriculture, Defence and Maritime Affairs were well-
placed to acquire privileged information, because they each had to procure goods from
private sector companies for their ministries. Such products were then used in building state-
sponsored railways, in the timber industry and in arms for the army and navy (Gindin,
2007a). But even before they placed industrial orders these officials had first to outline their
procurements plans in petitioning the Ministry of Finance for fiscal resources (Machlai,
2011).
5.2. The effect of government information
On the basis of written correspondence between bankers and government officials and
corporate board members, Bovykin (1967) documents many cases of banks obtaining and
then acting on privileged information. Banks learned from privileged information that certain
industrial companies presented low-risk, high-return opportunities. Certain industrial firms
were low-risk because the government often signed multi-year procurement contracts
(Gindin, 2007c), which convinced firms that state procurement would continue indefinitely
and at a constant level. Indeed, Figure 2 shows that, on the eve of the crisis, the bond risk
premium of heavy industrial companies (extractive and manufacturing) was as low as that of
long-established and conservative light industrial companies, represented here by the textile
industry. Furthermore, the risk premium of corporate securities was just one percent above
that of the safest government bonds.
15
In addition to being low-risk, certain industrial companies represented high-return
opportunities because government bought at above-market prices (Lyaschenko, 1949; Gindin,
2007c). Indeed, data between 1893 and 1898 reveal that heavy industries (extractive and
manufacturing) then averaged 11.7 per cent in return on equity, the ratio of net income to
capital stock.
13
Industry recipients of government orders probably made even more profit.
This suggests that the policies intended to develop heavy industry inadvertently
encouraged banks to devote more of their assets to financing heavy industry. Indeed, a
government audit of one of the troubled banks revealed that ‘in its activities, the bank broke
through the strict limits of its statute, putting significant sums of money into emerging
industrial enterprises both through acquiring shares in such enterprises and opening credit to
them’ (Bovykin & Petrov, 1994, p. 105).
13
Extractive and manufacturing industries accounted for about 90 per cent of all heavy industries’
capital stock. Calculated from data in the Ministry of Finance (1901b).
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
1898 1899 1900 1901 1902 1903 1904 1905
MINING
MANUFACTURING
TEXTILE
Figure 2. Monthly corporate bond risk premium by industry (in percent), 1898-1904
Notes: The corporate bond risk premium is the excess of current yield on corporate bonds of each of the three industries
over the current yield of the highest-rated government security, the four percent Russian government bond of 1894. Current
yield of each corporate b ond is calculated based on price and coupon data of 37 individual bonds. Gaps in the series are due
no trades in those months. Industry yield is the average yield of traded bonds in every month weighted by the outstanding
amount of each bond.
Sources: Price and coupon data is from the Ministry of Finance (1900-1907a). The quantity of bonds outstanding is from
Dmitriev-Mamonov (1903).
16
Furthermore, Witte’s reports on state revenues and expenditures over time
emphasised the need for unabated industrialisation. In one such report from 1897, Witte
states that, since the government ‘has followed the protectionist system with an unwavering
rigor and consistency’ for some time, the premature loosening of these policies would be ‘a
great political mistake and a source of major shocks to the economic organism’ (Ministry of
Finance, 1896a, p. 14). Even at the beginning of the crisis, in the official industrialisation
programme presented to the Czar, dated February 1899, Witte continued to stress the need for
uninterrupted industrialisation (Von Laue, 1954).
In this situation, it was natural for banks to expect the government to continue
procuring even in times of crisis. It might, thus, be argued that banks were more inclined to
finance riskier heavy industry projects because they expected government to support industry.
In fact, this is exactly what happened when the government organised a massive bailout of
industrial firms (Gindin, 1996). However, if banks indulged in riskier financing on the
expectation of industry bailouts, this simply makes state policies answerable for bank losses.
Witte’s continued emphasis on adhering as closely as possible to the existing course
of action may also explain why banks with well-informed connections did not envisage a
crisis in industry. Politicians and corporate board members simply may have not realised how
Witte’s promises might fail to materialise and hence did not advise their banking interlocks to
disengage immediately from financing industry.
Regarding the importance of interlocks between banks, the current literature suggests
that the presence of personal ties can result in a bank’s adopting a competitor’s corporate
strategy, which may result in the alignment of investment and lending positions (Connelly et
al., 2011). In times of crisis, banks with similar characteristics or exposure to the same
sectors can experience greater distress due to information contagion effects (Helwege &
Zhang, 2016). This implies that any corporate strategy chosen by Russian banks with political
17
and corporate interlocks could have passed to banks with no such connections via personal
ties between the banks.
5.3. Leveraging government interlocks
Banks were often compensated for their underwriting services by the securities themselves,
often content to retain the securities on their books (Bovykin, 1967). This suggests that banks
did not engage in underwriting-to-distribute. It was also common for banks to lend accepting
the industrial securities of clients as collateral (Bovykin, 1967). This kept such securities on
the books while the loan lasted. Had the banks doubted the positive prospects of their clients,
they would have refused this kind of collateral.
Moreover, individual bank board members and directors were compensated with
corporate securities (Bovykin, 1967). This suggests that some bankers had personal skin-in-
the-game and genuinely believed in the soundness of their clients. Bankers were also
rewarded by becoming members of corporate boards; or the reverse sometimes occurred
(Bovykin, 1967). However, given that bank board members held directorships and ownership
stakes with many corporations, they are best regarded as bankers on the boards of
corporations, not the other way around.
As banks became financially and interpersonally intertwined with companies, they
began to leverage their government connections to benefit their corporate clients. Some banks
directly petitioned ministries to procure goods from their clients (Bovykin, 1967). This
practice may count as rent-seeking and could have resulted in the suboptimal direction of
state orders. However, there is at least one reason to believe that banks’ petitions on their own
could not have so severely misallocated funds as to produce the deep recession of 1901-02 in
heavy industry. In the 1890s, industrial opportunities were abundant – 456 new joint-stock
companies were established between 1893 and 1898 (Dmitriev-Mamonov, 1903). Banks did
18
not have to mislead government into buying from a client-company if it was poorly-managed.
The bank could disinvest its holdings and invest in something better. In fact, Bovykin (1967)
documents how banks consistently rejected unattractive investments and lending
opportunities and sold unsuitable assets.
6. Empirical strategy and variables
Having established the role of government officials and corporate board members in banks’
decision-making in the 1890s, I next analyse the impact of interlocks on bank performance
during the crisis of 1899-1902. For that, I estimate OLS regression models, such that:
Distress
i
= β
0
+ β
1
(interlock variables) + β
2
(bank-specific variables) + ε
i
14
6.1. Dependent variables
To quantify bank distress, I use two separate dependent variables. In the first instance,
Distress
i
comes in the form of the net losses banks sustained on investment portfolios over
the whole crisis period, calculated as the ratio of net investment portfolio losses between
1899 and 1902 to the average value of the investment portfolio owned by the bank over the
same period. Note that, for the reasons discussed in Appendix Section B2, I assign a zero for
banks which made net investment profits over the crisis. I calculate the denominator only for
the years when the bank was in operation if it went bankrupt before the end of 1902.
To perform these calculations, I hand-collected investment profits and losses for
1899-1902 from banks’ individual annual reports in the Russian National Library and the
Russian State Historical Archive, both in St. Petersburg. See Appendix Table A1 for the
banks whose annual reports I located. Where annual reports were missing, I collected
14
Logarithmic and square root transformations do not substantially improve the fit of the models.
19
investment profits and losses for 1899-1900 from Statistics on short-term credit (Golubev,
1905).
For 1901-02, investment gains and losses were gathered from The Bulletin of Finance,
Industry and Trade (Ministry of Finance, 1902-1903c). Both publications were initially
compiled from individual banks’ annual reports. However, for several reasons, original
annual reports remain the most precise source for investment losses, and thus, whenever
possible, I draw data from them. Finally, I obtained end-of-year investment portfolio holdings
for 1899-1902 from Statistics on short-term credit (Golubev, 1905).
In the second instance, Distress
i
takes the form of the percentage change in the bank
share price from the maximum to the minimum level over 1899, the first year of the crisis.
During this period, the St. Petersburg Stock Exchange experienced a rapid decline of 10.6 out
of the 45.4 per cent total during the crisis (Goetzmann et al., n.d.a). I obtained bank share
prices from the dataset constructed by Goetzmann et al. (n.d.b). See Appendix Section B2 for
the reasons why these particular distress measures were chosen.
6.2. Independent variables
To capture the effect of government and corporate connections on bank performance, I add
interlock variables to my models. I register two broad categories of interlock: direct and
indirect. First, I recognise a direct interlock when the same person holds office in two entities
– a bank and a government body, or a bank and a company. This occurs only when the first,
middle and surnames of two individuals match. Second, I consider an indirect interlock when
a board member at a bank is connected to his sibling, either in the government or at a
company. This occurs when the surnames and middle names of two individuals match.
15
This
15
Another way an interlock could be established was via a father-son connection, but no such
linkages appeared. Women did not hold board positions at this time.
20
is because in Russia, the father’s first name customarily becomes the child’s middle name, a
tradition that has been followed since the 18th century (King & Wilson, 2011; Lisbach &
Meyer, 2013). What also ensures that I match true siblings is the fact that, for two individuals
to be identified as siblings, both must have made great progress in their respective careers.
After all, only a selected few became top government officials or corporate board members.
Furthermore, a good portion of bankers’ surnames, including Nobel, Bark, Gromme and
Lego, were unique and of non-Russian descent.
Next, I establish five different types of interlock: two government, two corporate and
one banking. I consider an interlock with government when a bank board member himself
(direct interlock), or his sibling (indirect interlock), is employed by a government entity – that
is, a state department, organisation, or agency – through which he may gather preferential
intelligence. To identify which government organisations to include in my dataset, I review
all the public entities that operated in Russia, as listed in the Address-Calendar (1898). I keep
nearly all entities, except the ones whose employees were absolutely unrelated to the
industrialisation process, for example, those in the Department of the Orthodox Confession. I
then gather the names of officials employed by the state in 1898 from the Address-Calendar
(1898) and the names of board members employed by banks in 1898 from Russian Banks
(Golubev, 1899). Note that I treat all government officials as equal, not differentiating higher-
from lower-ranks. Instead, I create a separate type of a government interlock, as follows.
Not all political connections are equal in value. Access to officials close to the
epicentre of policy-making should be more beneficial to bankers than ties to officials who are
remote from it. With this in mind, I record an interlock with the Finance Minister’s circle of
well-informed officials when there is a direct or indirect interlock between a bank board
member and the highest-ranked officials who were likely to be in close and regular contact
with Witte. Namely, I include top officials at (1) the Ministry of Finance, the centre of
21
policy-making as regards all national economic affairs (Raskin, 1998); (2) the Ministry of
Railways, one of the largest buyers in the country (Solovyov, 2003); (3) the Ministries of
Agriculture, Defence and Maritime Affairs which procured from companies; (4) the State
Bank, which provided loans to industry and banks; (5) the State Nobles’ Land Bank which
extended loans to the nobility, some of whom owned industrial enterprises; (6) the State
Treasury, responsible for managing national finances (Raskin, 1998); (7) the Government
Senate, the supreme supervisory body, where Witte often presented reports (Raskin, 1998);
and (8) the Council of the State, the supreme law-making body (Raskin, 1998).
Consequently, only the most influential government bodies are included. The limited number
of top officials in them should ensure that my interlock variables capture the passing of
privileged information alone.
Because banks sustained losses related to industry, I also track bank ties with the
corporate world. I consider the occurrence of an interlock with non-financial firms when a
bank board member either personally sits on the board or management committee of a non-
financial firm, or when that banker’s sibling does so. I include in my dataset a full range of
non-financial companies: heavy industrial, light industrial and non-industrial. The names of
corporate board members active at the end of 1896 are sourced from Statistics on Joint-Stock
Businesses in Russia (Pushkin, 1897). The aim was to pick a data source that was not too
close to the crisis in 1899, since it took time for ideas of newly-appointed corporate board
members to translate into bank strategy. Besides, corporate board members were elected for
five-year periods (Rudjuk, 2005), indicating that the composition of the majority of boards
was unlikely to have changed between the end of 1896 and 1899. Note that I consider
22
different types of corporate board members as equivalent: chairmen, vice-chairmen and
regular members are treated as identical.
16
Not all corporate connections are equal in value. Since the state targeted the
development of heavy industrial companies, bankers’ connections with such firms represent
another way of tracing the effect of state policies on bank performance. Hence, I record an
interlock with heavy industrial firms when a bank board member is also a board member of a
heavy industrial firm, or when the banker’s sibling is. I include in my dataset all firms
operating in heavy industries, namely: extractive industries, such as oil and minerals;
manufacturing industries that produced end-products as railroad tracks and ships; the
chemical industry; the timber industry; and state-owned railroad companies.
17
When it is
unclear whether a particular firm was heavy industrial, I consult company descriptions
provided by Dmitriev-Mamonov (1903). To identify state-owned railroad companies, I refer
to The Statistical Yearbook of the Ministry of Railways (Ministry of Railways, 1901).
To discern whether banks mimicked corporate strategy in competitor banks, I
document an interlock with banks when a bank board member is connected to a board
member at another bank, either personally or via a sibling. In all, I end up with five types of
interlock variable: bank board members connected to government officials, to the Finance
Minister’s circle, to all kinds of non-financial firms, to heavy industrial companies and to
other banks.
What do these interlocks capture? As discussed above, personal ties capture the
passing of privileged information. If they registered corruption, we would expect to see little
16
I do not rank board members because the purpose of my analysis is to identify whether corporate
connections mattered, rather than to pinpoint the board members whose rank was more consequential
than someone else’s.
17
Banks’ inability to find buyers for state-backed railroad bonds during the crisis forced them to keep
the bonds on their books, thus impairing their performance.
23
bank distress. This is because the suboptimal redistribution of state orders could not on its
own have led to the deep recession of 1901-02 in heavy industry.
It could also be asserted that bank losses had little to do with government policies and
were simply the result of banks’ speculation with industrial securities. There was indeed a
fair amount of speculative activity among banks and the general public, especially towards
the beginning of the crisis (Brandt, 1904). However, as discussed in detail in Appendix
Section B2, banks’ losses were primarily due to the massive reintermediation of industrial
securities onto banks’ balance sheets because industrial clients became unprofitable and
unable to service their bank loans. This suggests that banks’ ruinous performance was not due
to securities speculation, but rather to the industry’s economic downturn.
Finally, it could be argued that omitted factors, such as banks’ poor oversight of
industrial clients, drive the presence of both interlocks and bank losses. However, the
industrialisation policies and the information that came with them occurred first, while any
omitted factors can have resulted only from these information conditions.
My next step is to identify actual interlocks by matching banker names with the
names of government officials and in turn with the names of corporate board members. I use
a source code written specifically for this task. The matching process of 416 bankers, 3,378
corporate board members and 7,596 officials reveals that there were 49 bankers with personal
links to 63 officials and 142 bankers with personal ties to 418 corporate board members. The
matching also indicates that 74.2 per cent of all banker connections to the political and
corporate world were arranged via a sibling, while the remaining 25.8 per cent took place via
a direct interlock.
As the final step, I sum the number of bank board members with connections of either
type and use the total amounts in regression analysis. Note that if a particular banker is
connected to multiple corporate board members, I count this as a single interlock of a
24
particular type for that bank. In other words, for me to record ten bank board members with
interlocks, all ten members must have one or more corporate ties. The same method applies
to bank connections with government officials.
I also add bank-specific variables to my regressions to control for the heterogeneity of
banks. These variables are based on data from December 1898, when bankers hardly
expected the imminent crisis, yet only two months before the beginning of the stock market
downturn. My choice of these variables is as follows, with full descriptions located in Table
1: bank age, number of bank locations, leverage, liquidity, asset growth in the year before the
crisis and board size. I collect these variables from Statistics on Short-Term Credit (Golubev,
1905), with the exception of bank age, which comes from Russian Banks (Golubev, 1899).
Lastly, Appendix Tables A2 and A3 provide correlation matrices for all variables.
7. Empirical results
To preview my results, I examine the distribution of interlocks among banks. Appendix Table
A4 shows that banks located in St. Petersburg, the capital, were more connected to industry
than banks located in Moscow and the provinces. The table also shows that St. Petersburg’s
banks were the group most connected with heavy industry. Their 35 personal linkages
resulted in the number of connections per board member being four times the number of
those in Moscow or the provinces. Given that St. Petersburg banks sustained the greatest
financial losses of the three banking groups, as shown in the summary statistics in Table 2,
their high interconnectedness with heavy industry implies that personal ties may have played
a role in bank distress. St. Petersburg’s banks were also the group most connected with the
Finance Minister’s circle. Their 15 connections were three times the number of those in
25
Table 1: Definitions of variables used in regression analysis
Variables Unit Definition
Dependent variables
Net investment losses Ratio Net losses on equity and bond investments, 1899-1902 to Average value of investments, 1899-1902
Share price Ratio Percentage change in bank share price, from maximum to minimum value over 1899
Net investment profits Ratio Net profits on equity and bond investments, 1895-1898 to Average value of investments, 1895-1898
Control variables
Bank age Years Bank age
Number of locations Number Number of bank locations in 1898, including headquarters, branches, and agencies
Leverage Ratio Total debt to Total assets
Liquidity Ratio Percentage of total assets held in cash at the State Bank and other depository institutions
Asset growth Ratio Change in total assets over 1898
Board size Number Total members present on the bank board of directors and management committee
Interlock variables
Direct interlock Number A bank board member holds office at a government or corporate entity
Indirect interlock Number A bank board member is connected to his brother at a government or corporate entity
Member connected to non-financial firms Number Bank board members interlocked with non-financial firms, end of 1896
Member connected to heavy industrial firms Number Bank board members interlocked with heavy industrial firms, end of 1896
Member connected to the government Number Bank board members interlocked with government officials in 1898
Member connected to the Finance Minister’s circle Number Bank board members interlocked with government officials close to the Finance Minister in 1898
Member connected to banks Number Bank board members interlocked with competitor banks, end of 1898
Interlocks with non-financial firms Number Connections between bank board members and non-financial firms, end of 1896
Interlocks with heavy industrial firms Number Connections between bank board members and heavy industrial firms, end of 1896
Interlocks with the government Number Connections between bank board members and officials in 1898
Interlocks with the Finance Minister’s circle Number Connections between bank board members and officials close to the Minister of Finance in 1898
Interlocks with banks Number Connections between bank board members and competitor banks at the end of 1898
Share of board connected to non-financial firms Ratio Bank board members connected to non-fininancial firms in 1896 to All bank’s board members
Share of board connected to heavy industrial firms Ratio Bank board members connected to heavy industrial firms in 1896 to All bank’s board members
Share of board connected to the government Ratio Bank board members connected to government officials in 1898 to All bank’s board members
Share of board connected to the Finance Minister’s circle Ratio Bank board members connected to officials close to the Min. of Fin. in 1898 to All board members
Share of board connected to banks Ratio Bank board members connected to other banks, end of 1898 to All bank’s board members
26
Table 2: Summary statistics
N mean median SD min max
39 0.06 0.03 0.09 0.00 0.42
10 0.14 0.13 0.13 0.00 0.42
5 0.02 0.03 0.01 0.00 0.03
Net investment losses (all banks)
Net investment losses (St. Petersburg banks)
Net investment losses (Moscow banks)
Net investment losses (Provintial banks) 24 0.03 0.01 0.05 0.00 0.22
26 -0.16 -0.16 0.11 -0.41 0.00
9 -0.23 -0.22 0.09 -0.41 -0.07
4 -0.21 -0.22 0.07 -0.27 -0.14
Share price (all banks)
Share price (St. Petersburg banks)
Share price (Moscow banks)
Share price (Provintial banks) 13 -0.09 -0.07 0.08 -0.29 0.00
36 0.16 0.10 0.17 0.00 0.73Net investment profits (all banks)
Net income (all banks) 35 0.08 0.08 0.03 0.02 0.14
39 9.86 10.13 1.32 6.84 12.06
39 21.82 26.00 9.70 1.00 34.00
39 7.64 4.00 8.93 1.00 30.00
39 0.59 0.61 0.16 0.09 0.81
39 0.05 0.04 0.03 0.00 0.17
39 0.20 0.13 0.30 -0.66 1.15
Bank size
Bank age
Number of locations
Leverage
Liquidity
Asset growth
Board size 39 10.33 9.00 5.46 3.00 25.00
39 1.26 0.00 2.12 0.00 9.00
39 0.56 0.00 1.12 0.00 5.00
39 3.62 3.00 3.47 0.00 12.00
39 1.41 1.00 1.90 0.00 8.00
Member connected to the government
Member connected to the Finance Minister’s circle
Member connected to non-financial firms
Member connected to heavy industrial firms
Member connected to banks 39 1.05 0.00 1.83 0.00 7.00
39 1.62 0.00 2.71 0.00 11.00
39 0.69 0.00 1.38 0.00 6.00
39 9.67 5.00 11.30 0.00 43.00
39 3.03 1.00 4.29 0.00 17.00
Interlocks with the government
Interlocks with the Finance Minister’s circle
Interlocks with non-financial firms
Interlocks with heavy industrial firms
Interlocks with banks 39 1.08 0.00 1.90 0.00 7.00
Share of board connected to the government 39 0.09 0.00 0.12 0.00 0.38
Share of board connected to the Finance Minister’s circle 39 0.04 0.00 0.08 0.00 0.27
Share of board connected to non-financial firms 39 0.33 0.35 0.26 0.00 1.00
39 0.12 0.09 0.15 0.00 0.56Share of board connected to heavy industrial firms
Share of board connected to banks 39 0.07 0.00 0.11 0.00 0.42
27
Moscow.
18
Table 2 presents summary statistics on the interlock and non-interlock variables
for all banks.
My core evidence is presented in Table 3, which reports the results of five regression
models, the only difference between each specification being the type of the interlock
variable used. Note that these models incorporate both direct and indirect interlocks. Robust
standard errors are used in all models. Model (1) shows that, with each additional bank board
member connected to a variety of officials, both close to the Minister of Finance and not
directly related to his circle, a bank experienced 1.1 per cent of extra investment portfolio
losses, though the result is not statistically significant at standard levels. Model (2) tests
exclusively for bank board members connected to officials close to the Minister of Finance.
The model reveals that, with each additional interlock of this type, a bank lost much more, or
6.3 per cent of extra portfolio losses.
It may be concluded that the banks which had been influenced most by the
industrialisation policies exposed themselves financially to heavy industrial companies and
hence experienced greater losses. Importantly, it can be inferred that the banks which
sustained greater losses were the financiers of heavy industry, because the value of heavy
industrial equities declined by substantially more than the light industrial companies.
19
Model (3) in Table 3 provides evidence of the presence of a bank board member
connected to all kinds of non-financial firms resulting in 2.4 per cent of additional investment
18
The share of banks with connections to the Finance Minister’s circle was 28 per cent. For
comparison, Grossman & Imai (2016) find that in 1900 the share of British banks connected to
members of the Parliament was about 24 per cent.
19
The 52 heavy industrial companies traded between the peak and trough on the St. Petersburg Stock
Exchange experienced an average and median stock price decline of 56.2 and 65.5 per cent,
respectively. This contrasts to the 22.2 and 27.3 per cent declines for the 7 light industrial companies.
The closest available stock price was used for the periods when stocks were not traded. Thinly-traded
stocks and second and subsequent stock offerings were excluded. Calculated based on data from
Goetzmann et al. (n.d.b).
28
Table 3: Net investment losses 1899-1902, OLS model
1 2 3 4 5
Member connected to the government 0.011
(0.015)
Member connected to the Finance Minister’s circle 0.063
∗∗∗
(0.021)
Member connected to non-financial firms 0.024
∗∗∗
(0.007)
Member connected to heavy industrial firms 0.044
∗∗∗
(0.009)
Member connected to banks 0.013
(0.012)
Bank age -0.000 -0.001 -0.000 -0.000 -0.000
(0.002) (0.002) (0.001) (0.002) (0.002)
Number of locations 0.000 -0.002 -0.001 -0.001 -0.000
(0.001) (0.002) (0.001) (0.001) (0.001)
Leverage -0.068 0.057 -0.068 0.018 -0.055
(0.143) (0.107) (0.098) (0.073) (0.123)
Liquidity 0.604 0.018 0.505 -0.073 0.584
(0.869) (0.417) (0.674) (0.344) (0.815)
Asset growth 0.077 0.069 0.086
∗
0.083
∗
0.066
(0.059) (0.049) (0.046) (0.045) (0.052)
Board size 0.001 -0.001 -0.006
∗
-0.005
∗∗
0.002
(0.003) (0.003) (0.003) (0.002) (0.003)
Constant 0.021 0.030 0.045 0.035 0.016
(0.057) (0.077) (0.045) (0.050) (0.053)
Observations 39 39 39 39 39
Adjusted R
2
-0.038 0.345 0.357 0.572 -0.016
Robust standard errors in parentheses
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
29
losses.
Model (4) narrows the range of personal connections to bank board members
connected specifically to heavy industrial companies and reveals that the presence of this
type of interlock led to even greater distress, equal to 4.4 per cent of additional portfolio
losses. This suggests the particularly unfortunate effect of personal ties to heavy industrial
firms, those that were over-stimulated by the government.
Models (6) to (10) in Table 4 add an interaction term between bank board members
connected to the Finance Minister’s circle and heavy industrial companies. First, model (9)
shows that, for the bank with no connections to the Finance Minister’s circle, the independent
effect of a heavy industry interlock is a net investment loss, just as expected. Second, model
(7) reveals that, for the bank with no connections to heavy industry, the effect of an interlock
with the Finance Minister’s circle is not statistically significant. This suggests that banks had
to have heavy industry connections to sustain investment losses. Third, the positive and
statistically significant interaction term in model (7) suggests that the effect of having a heavy
industry interlock is strengthened by an interlock with the Finance Minister’s circle.
Models (11) and (12) in Table 4 model the effect of interlocks with both government
and industry. Model (12) reveals that with each additional bank board member connected
simultaneously to the Finance Minister’s circle and heavy industrial companies, the bank
experienced a 9.9 per cent of extra investment losses. As expected, this loss is larger in
magnitude than of an interlock with the Finance Minister’s circle alone, as shown in model
(2). These results are robust to additional tests (see Appendix B).
Univariate models (13) to (16) in Table 5 largely confirm the results discussed in
models (1) to (4) by testing the impact of interlocks on another distress variable, namely the
percentage change in the bank share price over the first year of the crisis.
20
These models
20
Not all banks were publicly traded on the St. Petersburg Stock Exchange; hence the limited sample
size of 26. Henceforth, the models using bank share price are univariate.
30
Table 4: Net investment losses 1899-1902, OLS model
6 7 8 9
10 11 12
Member connected to the government -0.007
(0.006)
Member connected to the Finance Minister’s circle -0.008
(0.023)
Member connected to non-financial firms 0.013
∗∗∗
(0.005)
Member connected to heavy industrial firms 0.023
∗∗∗
(0.008)
Member connected to banks 0.001
Interaction
(0.006)
0.012
∗∗∗
0.012
∗∗∗
0.009
∗∗∗
0.007
∗∗∗
0.011
∗∗∗
(0.001) (0.003) (0.001) (0.001) (0.002)
Baseline controls X X X X X X X
Member connected to gov./non-fin. firms 0.026
(0.023)
Member connected to FM’s circle/heavy industry 0.099
∗∗
(0.039)
Constant 0.051 0.063 0.070
∗
0.052 0.061 0.032 0.024
(0.042) (0.045) (0.040) (0.049) (0.047) (0.054) (0.071)
Observations 39 39 39 39 39 39 39
Adjusted R
2
0.557 0.545 0.656 0.648 0.544 0.048 0.372
Robust standard errors in parentheses. Baseline controls are bank age, locations, leverage, asset growth, and board size. Interaction term is member
connected to the Finance Minister’s circle * member connected to heavy industrial firms.
Member connected to gov.-non/fin. firms refers to a member simultaneously connected to the government and non-financial firms. Member connected
to FM’s circle/heavy industy refers to a member simultaneously connected to the Finance Minister’s circle and heavy industrial firms.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
Table 5
: Bank share price percentage change 1899, OLS model
13 14 15 16 17
Member connected to the government -0.011
(0.010)
Member connected to the Finance Minister’s circle -0.031
∗∗
(0.013)
Member connected to non-financial firms -0.014
∗∗
(0.005)
Member connected to heavy industrial firms -0.021
∗∗
(0.009)
Member connected to banks -0.026
∗∗
Constant
(0.011)
-0.142
∗∗∗
-0.136
∗∗∗
-0.094
∗∗∗
-0.118
∗∗∗
-0.125
∗∗∗
(0.026) (0.021) (0.031) (0.024) (0.024)
Observations 26 26 26 26 26
Adjusted R
2
0.002 0.095 0.157 0.131 0.154
Robust standard errors in parentheses. Univariate model.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0
.01
31
incorporate both direct and indirect interlocks. The models reveal that the presence of a bank
board member connected to the Finance Minister’s circle and separately to heavy industrial
companies translated into an extra reduction in the bank’s share price.
Model (5) in Table 3 shows that the presence of a bank board member connected to
competitor banks did not lead to statistically significant investment losses. However, model
(17) in Table 5 shows that having a board member of this type led to a 2.6 per cent extra
decline in the bank’s stock price. These results suggest that we need further evidence to make
a fully conclusive statement on the role of banking connections.
How much distressed were banks by industrial losses? Official audits revealed that
key banks were practically bankrupt or on the verge of collapse and it was only the massive
assistance from the State Bank that prevented the larger half of the banking system from
failure (Bovykin, 1984).
8. Conclusions
The main finding of my research is that the banks which experienced greater distress in the
crisis of 1899-1902, Imperial Russia’s last financial crisis, had more connections to those
government officials who were closest to the epicentre of policy-making, and to those
companies which had been most-stimulated by state policies to expand production in the
lead-up to the crisis. These findings indicate the negative effect of the national development
policies on bank performance following Russia’s rapid economic growth in the 1890s.
The findings suggest that it is not only banking regulation that can alter banks’
behaviour, but also other policy-conceived incentives. In industrialising Russia, it was
policies targeted at the development of the real economy that enticed banks to expose
themselves financially to new technology companies. These badly-designed incentives
culminated in disastrous bank performance when heavy industry experienced a slowdown. In
32
the end, even the possession of a stringent regulatory and supervisory banking framework did
not safeguard banks from near failure.
Put differently, during the crisis, sound investment and lending opportunities turned
out to be less safe than market participants, bankers and government officials had supposed
before the crisis. History offers many examples of similar cases. What this paper has
attempted to show is that government itself can point to seemingly secure and low-risk
opportunities that eventually turn out to be inept and costly decisions for banks.
The findings also suggest that it is worth considering the possible political
ramifications of economic development policies. The financial crisis aggravated the already
poor working and living conditions for ordinary workers, contributing to the spread of labour
strikes (Gefter, 1955; Gindin, 1950). The First Russian Revolution of 1905, which began only
two and a half years after the conclusion of the crisis, was initiated by workers in the very
industries the government had tried to develop (Vvedensky, 1952).
33
References
Abadie, A., Athey, S., Imbens, G., & Wooldridge, J. (2017). When should you adjust standard errors
for clustering? (MIT working paper).
Acemoglu, D., Johnson, S., Kermani, A., Kwak, J., & Mitton, T. (2016). The value of connections in
turbulent times: Evidence from the United States. Journal of Financial Economics, 121(2), 368-
391.
Adams, R., Hermalin, B, & Weisbach, M. (2010). The role of boards of directors in corporate
governance: a conceptual framework and survey. Journal of Economic Literature, 48(1), 58-107.
Afanasiev, G. (1900). Deneznyj krizis [The money crisis]. Odessa: U. I. Sandomirskiy.
Bernanke, B. (1983). Nonmonetary effects of the financial crisis in the propagation of the great
depression. The American Economic Review, 73(3), 257-276.
Borodkin, L. (2011a). Dobyvajushhaja promyshlennost, toplivno-jenergeticheskij kompleks [Mining
industry, fuel and energy complex]. In Dinamika jekonomicheskogo i socianogo razvitija Rossii v
XIX nachale XX v. [The dynamics of economic and social development of Russia in the XIX - early
XX century] [Data file]. Retrieved from http://www.hist.msu.ru/Dynamics/08_min.htm
Borodkin, L. (2011b). Kurs rublja k evropejskim valjutam, 1814-1914 gg. [The ruble exchange rate to
European currencies, 1814-1914]. In Dinamika jekonomicheskogo i socianogo razvitija Rossii v
XIX-nachale XX v. [The dynamics of economic and social development of Russia in the XIX - early
XX century] [Data file]. Retrieved from http://www.hist.msu.ru/Dynamics/data/12_013.xls
Borodkin, L. (2011c). Obrabatyvajushhaja promyshlennost [Processing industry]. In Dinamika
jekonomicheskogo i socianogo razvitija Rossii v XIX-nachale XX v. [The dynamics of economic and
social development of Russia in the XIX - early XX century] [Data file]. Retrieved from
http://www.hist.msu.ru/Dynamics/07_ind.htm
Bovykin, V. (1967). Zarozhdenie finansovogo kapitala v Rossii [The rise of financial capital in
Russia]. Moscow: Moscow State University.
Bovykin, V. (1984). Formirovanie finansovogo kapitala v Rossii konec XIX v. - 1908 g. [The
formation of financial capital in Russia. The end of the XIX century to 1908]. Moscow: Nauka.
Bovykin, V., & Petrov, U. (1994). Kommercheskie banki Rossijskoj Imperii [Commercial banks of the
Russian Empire]. Moscow: Perspektiva.
Brandt, B. (1902). Torgovo-promyshlennyĭ krizis v Zapadnoi Evropie i v Rossii, 1900-1901 g. Chast 1
[The commercial and industrial crisis in the West Europe and Russia, 1900-1901. Part 1]. Saint
Petersburg: V.O. Kirshvaum.
Brandt, B. (1904, December 9). Sovremennyj torgovo-promyshlennyj krizis v Zapadnoj Evrope [The
modern commercial and industrial crisis in Western Europe]. The bulletin of finance, industry and
trade, 447-455.
Braun, M., & Raddatz, C. (2010). Banking on politics: When former high-ranking politicians become
bank directors. The World Bank Economic Review, 24(2), 234-279.
34
Bugrov, A. (2003). Gosudarstvennyj bank i birzhevye sindikaty v Rossii, 1899-1917 gg. [The State
Bank and stock market syndicates in Russia, 1899-1917]. Jekonomicheskaja istorija: Ezhegodnik,
2002, 35-46. Moscow: Rosspen.
Calomiris, C. (1995). The costs of rejecting universal banking: American finance in the German
mirror, 1870-1914. In N. Lamoreaux and D. Raff (Eds.), Coordination and information: historical
perspectives on the organization of enterprise. Chicago: The University of Chicago Press.
Calomiris, C., & Haber, S. (2014). Fragile by design: The political origins of banking crises and
scarce credit. Princeton: Princeton University Press.
Cameron, R. (1967). Banking in the early stages of industrialization: A study in comparative
economic history. Oxford: Oxford University Press.
Cassis, Y. (2002). Finance and financiers in European history 1880-1960. Cambridge, UK:
Cambridge University Press.
Colvin, C., de Jong, A., & Fliers, P. (2015). Predicting the past: understanding the causes of bank
distress in the Netherlands in the 1920s. Explorations in Economic History, 55(1), 97-121.
Colvin, C. (2014). Interlocking directorates and conflicts of interest: the Rotterdamsche
Bankvereeniging, Muller & Co. and the Dutch financial crisis of the 1920s. Business History, 56(2),
314-334.
Complete Collection of laws of the Russian Empire. (1899a). Collection 3, volume 15. Retrieved from
http://www.nlr.ru/e-res/law_r/search.php
Complete Collection of laws of the Russian Empire. (1899b). Collection 3, volume 16. Retrieved
from http://www.nlr.ru/e-res/law_r/search.php
Complete Collection of laws of the Russian Empire. (1900). Collection 3, volume 17. Retrieved from
http://www.nlr.ru/e-res/law_r/search.php
Complete Collection of laws of the Russian Empire. (1901). Collection 3, volume 18. Retrieved from
http://www.nlr.ru/e-res/law_r/search.php
Complete Collection of laws of the Russian Empire. (1903). Collection 3, volume 21, law # 20874.
Retrieved from http://www.nlr.ru/e-res/law_r/search.php
Connelly, B., Johnson, J., Tihanyi, L., & Ellstrand, A. (2011). More than adopters: Competing
influences the interlocking directorate. Organization Science, 22(3), 688-703.
Council of Congresses of Representatives of Industry and Trade. (1914). Doklad Coveta Cezdov o
sovremennom polozhenii promyshlennosti i torgovli i vidah na budushhee v svjazi s
predprinimatelskoj dejatelnostju kazny [The Report of the Council of Congresses on the current
situation in industry and trade and the future prospects of the entrepreneurial activity of the state].
Typography of Stoykova.
Crisp, O. (1976). Studies in the Russian economy before 1914. New York: Barnes & Noble Books.
Dmitriev-Mamonov, V. (1903). Ukazatel deistvuiushchikh v imperii aktsionernykh predpriiatii i
torgovikh domov [The directory of corporations and partnerships in the Russian Empire]. St.
Petersburg: Isidor Goldberg.
35
Duchin, R., & Sosyura, D. (2012). The politics of government investment. Journal of Financial
Economics, 106, 24-48.
Faccio, M. (2006). Politically Connected Firms. The American Economic Review, 96(1), 369-386.
Fohlin, C. (1999). The rise of interlocking directorates in imperial Germany. Economic History
Review, 52(2), 307-333.
Friedman, M., & Schwartz, A. (1963). A monetary history of the United States, 1867‐1960. Princeton:
Princeton University Press.
Gefter, M. (1955). Jekonomicheskie predposylki pervoj russkoj revoljucii [Economic preconditions
for the First Russian Revolution]. Doklady i soobshhenija [Reports and messages], 6, 3-26.
Goetzmann, W., Cabolis, C., & Radchenko, P. (n. d.a). St. Petersburg Stock Exchange Index. SPSE
and NYSE comparison [Data set]. Retrieved from
http://som.yale.edu/sites/default/files/files/SPSENYSE18651917ind.xls
Goetzmann, W., Cabolis, C., & Radchenko, P. (n. d.b). St. Petersburg Stock Exchange Index.
Individual securities data [Data set]. Retrieved from
http://som.yale.edu/sites/default/files/files/St_%20Petersburg%20data%20Revised%20082010%28
1%29.xlsx
Gerschenkron, A. (1962). Economic backwardness in historical perspective: A book of essays.
Cambridge: Harvard University Press.
Gertsenshtein, M. (1903). Harkovskij krah. Po povodu processa o zloupotreblenijah v Harkovskom
zemelnom i torgovym bankah [Kharkov collapse. Regarding the process of misuses at the Kharkov
Mortgage and Commercial banks]. St. Petersburg: Slovo.
Gindin, I. (1950). Neustavnye ssudy Gosudarstvennogo banka i ekonomiceskaja politika carskogo
pravitelstva [Extraordinary loans of the Bank and economic policy of the tsarist government].
Istoricheskie zapiski [Historical Notes], 35, 87-126.
Gindin, I. (1957). Trebovanija dvorjanstva i finansovo-jekonomicheskaja politika carskogo
pravitelstva v 1880-1890-h godax [The requirements of the nobility and the financial and economic
policy of the tsarist government in the 1880-1890's]. Istoricheskij arhiv [The Historical Archive], 4,
122-155.
Gindin, I. (1958). Moskovskie banki v period imperializma [Moscow banks during the period of
Imperialism]. Istoricheskie zapiski [Historical Notes], 58, 38-106.
Gindin, I. (1959). Ob osnovah jekonomicheskoj politiki carskogo pravitelstva v konce XIX - nachale
XX vv. [About the foundation of the economic policy of the Czarist government in the late XIX -
beginning of XX centuries]. Materialy po istorii SSSR [Materials on the history of the USSR], 6,
173-195.
Gindin, I. (1960). Gosudarstvennyj bank i jekonomicheskaja politika carskogo pravitelstva, 1861-
1892 [The State Bank and the economic policy of the tsarist government, 1861-1892]. Moscow:
Gosfinizdat.
Gindin, I. (1996). Jekonomicheskij krizis 1900-1903 goda i antikrizisnaja politika pravitelstva [The
economic crisis of 1900-1903 and the state anti-crisis policy]. Iz istorii jekonomicheskoj mysli i
narodnogo hozjajstva Rossii [From the history of economic thought and the economy of Russia], 2,
1-153.
36
Gindin, I. (2007a). Gosudarstvo i jekonomika v gody upravlenija S. Ju. Witte [The state and economy
in the years of S. U. Witte’s governance]. Voprosy Istorii [Questions of History], 1, 78-93.
Gindin, I. (2007b). Gosudarstvo i jekonomika v gody upravlenija S. Ju. Witte [The state and economy
in the years of S. U. Witte’s governance]. Voprosy Istorii [Questions of History], 4, 70-97.
Gindin, I. (2007c). Gosudarstvo i jekonomika v gody upravlenija S. Ju. Witte [The state and economy
in the years of S. U. Witte’s governance]. Voprosy Istorii [Questions of History], 6, 80-106.
Gindin, I. (2007d). Gosudarstvo i jekonomika v gody upravlenija S. Ju. Witte [The state and economy
in the years of S. U. Witte’s governance]. Voprosy Istorii [Questions of History], 9, 79-110.
Gregory, P. (1991). The role of the state in promoting economic development: The Russian case and
its general implications. In R. Sylla and G. Toniolo (Eds.), Patterns of European industrialization:
The nineteenth century. London: Routledge.
Gregory, P. (1994). Before command: An economic history of Russia from emancipation to the first
five-year plan. Princeton: Princeton University Press.
Gregory, P. (2003). Jekonomicheskij rost Rossijskoj imperii (konec XIX - nachalo XX v.): Novye
podschety i ocenki [The economic growth of the Russian Empire (the end of the XIX – beginning of
the XX century): New calculations and estimates]. Moscow: Rossen.
Grossman, R., & Imai, M. (2016). Taking the lord’s name in vain: The impact of connected directors
on 19th century British Banks. Explorations in Economic History, 59(1), 75-93.
Hall, R. (2010). Why does the economy fall to pieces after a financial crisis? Journal of Economic
Perspectives, 24(4), 3-20.
Hammond, B. (1957). Banks and politics in America from the revolution to the Civil War. Princeton:
Princeton University Press.
Helwege, J., & Zhang, G. (2016). Financial firm bankruptcy and contagion. Review of Finance, 20(4),
1321-1362.
Kanel, V. (1906). Samoderzhavie i narodnoe hozjajstvo [Autocracy and the economy]. In Voprosy
dnja. Sbornik statej, 1-29. Moscow: I. N. Holchev & Co.
King, G., & Wilson, P. (2011). The resurrection of the Romanovs: Anastasia, Anna Anderson, and the
world’s greatest royal mystery. Hoboken: John Wiley & Sons.
Krugman, P. (1998). What happened to Asia? (Unpublished conference paper).
Laeven, L. (2001). Insider lending and bank ownership: The case of Russia. Journal of Comparative
Economics, 29(2), 207-229.
Lamoreaux, N. (1996). Insider lending: banks, personal connections, and economic development
industrial New England. Cambridge: Cambridge University Press.
La Porta, R., Lopez-de-Silanes, F., & Zamarripa, G. (2003). Related lending. The Quarterly Journal
of Economics, 118(1), 231-268.
League of Nations (1945). Industrialization and foreign trade. Geneva: League of Nations.
37
Lebedev. S. (2003). S.-Peterburgskij Mezhdunarodnyj Kommercheskij Bank vo vtoroj polovine XIX
veka [The St. Petersburg International Commercial Bank in the second half of the XIX century].
Moscow: Rosspen.
Lisbach, B., & Meye, V. (2013). Linguistic identity matching. Freiburg: Springer Vieweg.
Lyaschenko, P. (1949). History of the national economy of Russia to the 1917 Revolution. New York,
NY: Macmillan.
Machlai, M. (2011). Deloproizvodstvennye dokumenty centralnyh gosudarstvennyh uchrezhdenij kak
istochnik po istorii kazennoj promyshlennosti ... [Business correspondence documents of the central
government offices as a source on the history of state industry in Russian Empire at the end of XIX
–beginning of XXth centuries]. The Bulletin of Udmurt State University, 1, 19-31.
Matveeva, L. (1897). Socialno-jekonomicheskie predposylki Velikoj Oktjabrskoj socialisticheskoj
revoljucii [The socio-economic background of the Great October Socialist Revolution]. The
National Academy of Ukraine.
Migulin, P. (1907). Russkij gosudarstvennyj kredit (1769-1906). Opyt istoriko-kriticheskogo obzora.
Tom 3 [Russian state credit (1769-1906). The experience of historical and critical review. Volume
3] Kharkiv: Tipolithography Pechatnoe Delo.
Newton, L. (1996). Regional bank-industry relations during the mid-nineteenth century: links between
bankers and manufacturing in Sheffield. Business History, 38(3), 64-83.
Ol, P. (1925). Inostrannye kapitaly v narodnom hozjajstve dovoennoj Rossii [Foreign capital in the
economy of pre-war Russia]. Leningrad: Press of the All-Union Academy of Sciences.
Ozerov, I. (1905). Jekonomicheskaja Rossija i ejo finansovaja politika na ishode XIX i v nachale XX
veka [Economic Russia and her financial policy at the turn of the XIX and early XX century].
Moscow: Tipolithography of I. N. Kushnereva & Co.
Rajan, R. (2011). Fault Lines: how hidden fractures still threaten the world economy. Princeton:
Princeton University Press.
Raskin, D. (1998). Vysshie i centralnye gosudarstvennye uchrezhdenija Rossii. 1801-1917gg. Tom 1-
4. [The supreme and central state institutions in Russia, 1801-1917. Volumes 1-4]. St. Petersburg:
Nauka.
Rudjuk, N. (2005). Bankovskie dejateli Rossii 1864-1914 gg.: Istocniki i metody izucenija
biograficeskih dannyh [Banking leaders of Russia 1864-1914: Sources and methods of studying
biographical data] (Doctoral dissertation). Moscow: Moscow State University.
Siegel, J. (2015). For Peace and Money: French and British finance in the service of tsars and
commissars. New York, NY: Oxford University Press.
Shepelev, L. (1981). Carizm i burzhuazija vo vtoroj polovine XIX veka [Czarism and the bourgeoisie
in the second half of the 19th century]. Leningrad: Nauka.
Sheppard, D. (2006). The growth and role of U.K. financial institutions 1880-1962. London:
Routledge.
Solovyov, Y. (2003). Ministerstvo finansov Rossijskoj imperii v 1858 - 1903 gg.: Organizacija i
funkcionirovanie [Ministry of Finance of the Russian Empire in 1858-1903: Organization and
functioning] (Doctoral dissertation). Moscow: Moscow Regional State University.
38
Stern, G., & Feldman, R. (2004). Too Big to Fail: The hazards of bank bailouts. Washington, D.C.:
Brookings Institution Press.
The Economist (newspaper). (1901, June 29). Distressful Russia.
Thomas, R. & Williamson, S. (2018). What was the U.K. GDP then? Retrieved from
http://www.measuringworth.com/ukgdp/
Turner, J. (2014). Banking in crisis: The rise and fall of British banking stability, 1800 to the present.
Cambridge: Cambridge University Press.
Vvedensky, B. (1952). Devjatoe janvarja 1905 [The ninth of January 1905]. In Bolshaja sovetskaja
jenciklopedija [The Great Soviet Encyclopedia]. (Vol. 13, pp. 539-540). Moscow: Bolshaja
Sovetskaja Jenciklopedija.
Von Laue, T. (1954). A secret memorandum of Sergei Witte on the industrialization of Imperial
Russia. The Journal of Modern History, 26(1), 60-7.
Witte, S. (1898a). Rech ministra finansov 6 marta na zasedanie kommissii po peresmotru
dejstvujushhih zakonopolozhenij o birzhah i akcionernyh kompanijah [Speech of the Minister of
Finance on the March sixth meeting of the commission on the revision of the existing regulation
relating to exchanges and joint-stock companies]. In N. Figurovskaya (Ed.), S. Ju. Witte. Sobranie
sochinenij i dokumentalnyh materialov: v pjati tomah. Tom 4: Promyshlennost, torgovlja i selskoe
hozjajstvo Rossii. Kniga 1. Moscow: Nauka.
Witte, S. (1898b). Vsepoddanejshaja dokladnaja zapiska S. Ju. Witte ‘Ob ugnetjonnom polozhenii
Evropejskih denezhnyh rynkov’ [The most loyal memo of S. U. Witte to the Czar ‘On the depressed
situation in European money markets’]. In N. Figurovskaya (Ed.), S.Ju. Witte. Sobranie sochinenij i
dokumentalnyh materialov: v pjati tomah. Tom 2: Nalogi, bjudzhet i gosudarstvennyj dolg Rossii.
Kniga 2. Moscow: Nauka.
Official publications
Address-Calendar: A general list of governing and other functionaries at all the offices in the Russian
Empire for the year 1898 [Adres-Kalendar. Obshhaja rospis nachalstvujushhih i prochih
dolzhnostnyh lic po vsem upravlenijam v Rossijskoj Imperii na 1898 god]. (1898). St. Petersburg:
Printing Office of the Senate.
Golubev, A. (1899). Russkie banki. Spravochnye i statisticheskie svedenija o vseh dejstvujushhih v
Rossii gosudarstvennyh, chastnyh i obshhestvennyh kreditnyh uchrezhdenijah [Russian banks.
Reference and statistical information about all active government, private, and public credit
institutions in Russia]. St. Petersburg: Skoropechatnya P. O. Yablonsky.
Golubev, A. (1905). Statistika kratkosrochnogo kredita. Operacii akcionernyh bankov
kommercheskogo kredida v 1894-1900 gg. Tom 1 [Statistics on short-term credit. Operations of
joint-stock commercial banks between 1894 and 1900. Volume 1] St. Petersburg: Printing house of
V. Kirschbaum.
Kazan Merchant Bank. (1900, 1901, 1902). Annual report for the year 1899, 1900, 1901. Kazan.
Kiev Private Commercial Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900,
1901, 1902. Kiev.
39
Lodz Merchant Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
Lodz.
Ministry of Finance. (1896a, 1900a). Vsepoddannejshij doklad ministra finansov o
gosudarstvennoj rospisi dohodov i rashodov na 1894, 1897, 1901 g. [The most loyal report by the
Minister of Finance to the Czar on state revenues and expenditures for the year 1894, 1897, 1901].
St. Petersburg.
Ministry of Finance. (1900-1907b). Ezhegodnik Ministerstva Finansov. Vypusk 1899-1907 [The
Yearbook of the Ministry of Finance. Issue 1899-1907]. St. Petersburg: V. Kirschbaum.
Ministry of Finance (1902-1903c). Vestnik finansov, promyshlennosti i torgovli. Otchety kreditnyh
uchrezhdenij, torgovyh i promyshlennyh predprijatij [The bulletin of finance, industry and trade.
Financial statements of credit, trade, and industrial enterprises]. St. Petersburg: V. Kirschbaum.
Ministry of Finance. (1902d). Obshhaja gosudarstvennaja rospis dohodov i rashodov na 1903 god
[Government revenues and expenditures for the year 1903]. St. Petersburg.
Ministry of Railways (1901). Statisticheskij sbornik Ministerstva Putej Soobshhenija. Svedenija o
zheleznyh dorogah za 1899 g. [The statistical yearbook of the Ministry of Railways. Information on
railways for 1899]. St. Petersburg: The Ministry of Railways Press.
Minsk Commercial Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901,
1902. Minsk.
Moscow Discount Bank. (1900, 1902, 1903). Annual report for the year 1899, 1901, 1902. Moscow.
Moscow Merchant Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901,
1902. Moscow.
Moscow Trade Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
Moscow.
Nizhny Novgorod Merchant Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900,
1901, 1902. Nizhny Novgorod.
Odessa Discount Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
Odessa.
Orlovsky Commercial Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901,
1902. Orlov.
Pushkin, N. (1897). Statistika akcionernogo dela v Rossii. Sostav direktorov pravlenij na 1897 g.
Vypusk pervyj [Statistics on joint-stock businesses in Russia. The composition of the boards of
directors in 1897. Volume 1]. Moscow: I. N. Kushnerev and Co.
Riga Commercial Bank. (1902, 1903). Annual report for the year 1901, 1902. Riga.
Russian for Foreign Trade Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900,
1901, 1902. St. Petersburg.
Russian for Trade and Industry Commercial Bank. (1902, 1903). Annual report for the year 1901,
1902. St. Petersburg.
40
Russo-Chinese Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
St. Petersburg.
Russian State Historical Archive (RSHA). Collection 587, inventory 33, file 98: State Bank (n. d.),
Vedomost o dvizhenii ssud na osobyh osnovanijah v 1899 godu [A statement on the movement of
non-statutory loans in 1899]. St. Petersburg.
Russian State Historical Archive (RSHA). Collection 587, inventory 33, file 101, pp. 12-13: State
Bank (n. d.), Objasnitelnaja zapiska k otchjotu Gosudarstvennogo Banka za 1900 god [Explanatory
note to the report of the State Bank for the year 1900]. St. Petersburg.
Russian State Historical Archive (RSHA). Collection 588, inventory 1, file 247: State Bank (n. d.),
Kniga specialnyh tekushhih schetov, otkrytyh chastnym licam, bankam, obshhestvam i drugim
uchrezhdenijam pod obespechenie vekselej i tovarov za 1899 g. [A ledger of special current
accounts collateralized with bills of exchange and goods, opened to individuals, banks, companies,
and other institutions in the year 1899]. St. Petersburg.
Russian State Historical Archive (RSHA). Collection 588, inventory 1, file 277: State Bank (n. d.),
Kniga specialnyh tekushhih schetov, otkrytyh chastnym licam, bankam, obshhestvam i drugim
uchrezhdenijam pod obespechenie procentnyh bumag za 1898 g. [A ledger of special current
accounts collateralized with equity and bond securities, opened to individuals, banks, companies,
and other institutions in the year 1898]. St. Petersburg.
Russian State Historical Archive (RSHA). Collection 588, inventory 1, file 593: State Bank (n. d.),
Spiski lici firm, predjavivshih k uchetu vekselja za 1897-1900 g. [Records of individuals and
companies that submitted bills of exchange for discounting in 1897-1900]. St. Petersburg.
Russo-Chinese Bank. (1900-1903). The report on operations for the Russo-Chinese Bank for the year
1899, 1900, 1901, 1902. St. Petersburg.
Siberian Trade Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
Yekaterinburg.
South-Russian Industrial Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900,
1901, 1902. Moscow.
State Bank of the Russian Empire. (1899). Otchet Gosudarstvennogo banka za 1898 [Annual report of
the State Bank for 1898]. St. Petersburg.
State Comptroller. (1902). Vsepoddanneyshiy otchet gosudarstvennogo kontrolera za 1901 [The
report of the State Comptroller for the year 1901]. St. Petersburg: State Press.
St. Petersburg Discount and Loan Bank. (1900, 1901, 1902). Annual report for the year 1899, 1900,
1901. St. Petersburg.
St. Petersburg International Commercial Bank. (1900, 1901, 1902, 1903). Annual report for the year
1899, 1900, 1901, 1902. St. Petersburg.
St. Petersburg Muscovy Commercial Bank. (1902, 1903). Annual report for the year 1901, 1902. St.
Petersburg.
St. Petersburg Private Commercial Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899,
1900, 1901, 1902. St. Petersburg.
41
Trade Bank in Lodz. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901, 1902.
Lodz.
Vilnius Private Commercial Bank. (1901, 1902, 1903). Annual report for the year 1900, 1901, 1902.
Vilnius.
Volzhsko-Kamski Commercial Bank. (1900, 1901, 1902). Annual report for the year 1899, 1900,
1901. St. Petersburg.
Warsaw Discount Bank. (1900, 1901, 1902, 1903). Annual report for the year 1899, 1900, 1901,
1902. Warsaw.
42
Appendix A. Summary statistics
Table A1: List of banks
Bank name Headquarters Total assets Branches Bank age Annual report
St. Petersburg 173,402 22 28 1900-1902
145,697 29 29 1900-1903
108,231 14 27 1900-1903
96,715 24 2 1900-1903
Volzhsko-Kamski Commercial Bank
St. Petersburg International Commercial Bank
Russian for Foreign Trade Bank
Russo-Chinese Bank
St. Petersburg Discount and Loan bank
St. Petersburg
St. Petersburg
St. Petersburg
St. Petersburg 73,734 2 29 1900-1902
Russian for Trade and Commerce Commercial Bank St. Petersburg 67,437 17 8 1902-1903
St. Petersburg 48,878 3 34 1900-1903
45,127 9 11
28,272 1 14 1902-1903
28,203 3 7
117,016 4 32 1900-1903
60,701 30 25
32,863 1 27 1900-1903
23,849 2 28 1900-1903
15,993 8 27 1900-1903
81,522 28 27
55,574 9 28
31,476 13 26 1900-1903
28,954 6 26 1902-1903
25,819 4 26 1900-1903
25,527 22 26 1900-1903
25,152 3 19 1900-1903
24,355 4 26
15,959 6 26 1901-1903
15,008 1 30 1900-1903
12,143 1 30
11,414 12 25 1900-1903
10,745 1 27 1900-1903
10,183 5 25
5,497 4 25
5,242 1 28 1900-1903
3,981 1 25 1900-1902
3,948 2 27
3,444 1 1 1900-1903
3,308 1 26
2,827 1 1
2,338 1 5
2,236 1 1
St. Petersburg Private Commercial Bank
St. Petersburg-Azovcky Commercial Bank
St. Petersburg Muscovy Commercial Bank
Credit Lyonnais
Moscow Merchant Bank
Moscow International Trade Bank
Moscow Trade Bank
Moscow Discount Bank
South-Russian Industrial Bank
Azovsko-Donskoi Commercial Bank
Commercial Bank in Warsaw
Siberian Trade Bank
Riga Commercial Bank
Trade Bank in Lodz
Orlovsky Commercial Bank
Odessa Discount Bank
Tiflis Commercial Bank
Vilnius Private Commercial Bank
Kiev Private Commercial Bank
Kharkov Trade Bank
Minsk Commercial Bank
Warsaw Discount Bank
Pskov Commercial Bank
Voronezh Commercial Bank
Nizhny Novgorod Merchant Bank
Kazan Merchant Bank
Commercial Bank in Kostroma
Lodz Merchant Bank
Ekaterynoslavsky Commercial Bank
Commercial Bank in Bialystok
Rostov-on-Don Merchant Bank
Baltic Commerce and Industry Bank
Central Asian Commercial Bank
St. Petersburg
St. Petersburg
St. Petersburg
Moscow
Moscow
Moscow
Moscow
Moscow
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces
The provinces 932 1 17
Notes: sorted by location of headquarters and then by total assets. Total assets are in thousands of rubles.
Sources: Golubev (1899, 1905)
43
Table A2: Correlation
of outcome and bank-specific predictor variables
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
(1) Net investment losses 1.00
(2) Share price -0.46
∗∗
1.00
(3) Net investment profits 0.72
∗∗∗
-0.41
∗∗
1.00
(4) Bank size 0.47
∗∗
-0.50
∗∗
0.37
∗
1.00
0.02 0.31 0.20 -0.05 1.00
0.01 -0.29 -0.11 0.60
∗∗∗
-0.06 1.00
0.06 0.13 0.18 0.31 0.10 0.06 1.00
0.26 0.07 -0.02 0.22 -0.26 0.20 -0.07 1.00
-0.02 0.07 -0.02 -0.37
∗
0.07 0.01 -0.43
∗∗
-0.22 1.00
(5) Bank age
(6) Number of locations
(7) Leverage
(8) Liquidity
(9) Asset growth
(10) Board size 0.34 0.00 0.03 0.44
∗∗
0.12 0.18 0.22 0.08 -0.08 1.00
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
Table A3: Correlation of interlock predictor variables
(1) (2) (3) (4) (5)
(1) Member connected to the government 1.00
(2) Member connected to the Finance Minister’s circle 0.61
∗∗∗
1.00
(3) Member connected to non-financial firms 0.66
∗∗∗
0.55
∗∗∗
1.00
0.42
∗∗∗
0.66
∗∗∗
0.81
∗∗∗
1.00(4) Member connected to heavy industrial firms
(5) Member connected to banks 0.70
∗∗∗
0.42
∗∗∗
0.65
∗∗∗
0.39
∗∗
1.00
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
44
Table A4: Distribution of bank board member connections
Location
54 29 12 9 12
8 6 8 6 5
62 35 20 15 17 120
56%
75%
Moscow 30 4 14 1 15
Moscow 1 0 7 2 4
Moscow 31 4 21 3 19 72
Moscow 13%
Moscow
14%
Provincial
39 13 4 1 3
Provincial
10 3 4 3 2
Provincial
49 16 8 4 5 224
Provincial
33%
Provincial
50%
All banks 142 55 49 22 41 416
Non-
financial
firms
Heavy
industry
firms
The
govern-
ment
The
finance
minister's
circle
Other
banks
Total
bank
board
members
Bank board member connected to:
Direct interlock St. Petersburg
Indirect interlock St. Petersburg
Total St. Petersburg
Interlocks with heavy industry firms as % of those with all non-financial firms St. Petersburg
Interlocks with the Finance Minister’s circle as % of those with the government St. Petersburg
Direct interlock
Indirect interlock
Total
Interlocks with heavy industry firms as % of those with all non-financial firms
Interlocks with the Finance Minister’s circle as % of those with the government
Direct inter
lock
Indirect interlock
Total
Interlocks with heavy industry firms as % of those with all non-financial firms
Interlocks with the Finance Minister’s circle as % of those with the
government
Notes: Direct interlock occurs when a bank board member holds once in the government or at a company. Indirect interlock occurs when a bank board member is connected
to his brother either in the government or at a company. Location is the location of the bank’s headquarters. Member connected to non-financial firms is the number of bank
board members interlocked with non-financial firms. Member connected to heavy industrial firms is the number of bank board members interlocked with heavy industrial rms.
Member connected to the government is the number of bank board members interlocked with all types of government officials. Member connected to the Finance Minister’s
circle is the number of bank board members interlocked with government officials who were in close contact with the Minister of Finance. Member connected to other banks
is the number of bank board members interlocked with competitor banks. Total bank board members is the total number of bank board members at all banks.
45
Appendix B. Robustness
Models (18) to (22) in Appendix Table B1 show that personal connections brought
significant value to banks in the years before the crisis. In particular, Model (19) portrays that
having a bank board member with both direct and indirect connections to the Finance
Minister’s circle resulted in 9.2 per cent of extra investment profits between 1895 and 1898.
Model (21) shows that connections with heavy industrial firms brought rewards as well,
allowing the bank to outperform by 6.5 per cent. These were considerable gains because in
this period the stock market appreciated by a mere 6.9 per cent.
21
Models (23) to (27) in Appendix Table B2 and models (28) to (32) in Appendix Table
B3 use a different set of interlock variable, capturing only direct interlocks. These models
confirm the results of the core models (1) through (5).
Models (33) to (37) in Appendix Table B4 and models (38) to (42) in Appendix Table
B5 use yet another set of interlock variable, capturing only indirect interlocks. These models
confirm the results of core models (1) through (10), except that the effect of connections with
the Finance Minister’s circle is not fully conclusive, as seen in model (34).
Also note that in models (35) and (36) the coefficients on indirect interlocks with non-
financial and heavy industrial companies are higher than on direct interlocks, as shown in
models (25) and (26). Intuitively, we would expect direct interlocks to bring more distress.
One potential explanation for the opposite result is that direct interlocks were not the main
decision-makers, while bank board members with indirect connections were.
Models (43) to (47) in Appendix Table B6 and models (48) to (52) in Appendix
Table B7 use a different set of interlock variable, or the share of bank board members with
both direct and indirect connections of all board members at a bank. These models confirm
the results of the core models.
21
Calculated from data in Goetzmann et al. (n. d.).
46
Table B1: Net investment profits 1895-1898, OLS model
18 19 20 21 22
Member connected to the government 0.020
(0.019)
Member connected to the Finance Minister’s circle 0.092
∗∗
(0.041)
Member connected to non-financial firms 0.047
∗∗∗
(0.014)
Member connected to heavy industrial firms 0.065
∗∗∗
(0.020)
Member connected to banks 0.021
(0.016)
Baseline controls X X X X X
Constant 0.158 0.179
∗
0.178
∗
0.133 0.153
(0.107) (0.104) (0.091) (0.102) (0.099)
Observations 36 36 36 36 36
Adjusted R
2
-0.125 0.110 0.317 0.256 -0.120
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, asset growth, and board size.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0.01
47
Table B3: Bank share price percentage change 1899, OLS model
28 29 30 31 32
Member connected to the government (direct interlock) -0.013
(0.016)
Member connected to the Finance Minister’s circle (direct interlock) -0.041
(0.028)
Member connected to non-financial firms (direct interlock) -0.013
∗∗
(0.005)
Member connected to heavy industrial firms (direct interlock) -0.022
∗∗
(0.010)
Member connected to banks (direct interlock) -0.024
∗
Constant
(0.012)
-0.147
∗∗∗
-0.142
∗∗∗
-0.104
∗∗∗
-0.122
∗∗∗
-0.136
∗∗∗
(0.029) (0.024) (0.029) (0.025) (0.025)
Observations 26 26 26 26 26
Adjusted R
2
-0.020 0.036 0.108 0.097 0.083
Robust standard errors in parentheses. Univariate model.
∗
p < 0
.10,
∗∗
p
< 0
.
05,
∗∗∗
p
< 0
.
01
Table B2: Net investment losses 1899-1902, OLS model
23 24 25 26 27
Member connected to the government (direct interlock) 0.013
(0.019)
Member connected to the Finance Minister’s circle (direct interlock) 0.088
∗∗
(0.036)
Member connected to non-financial firms (direct interlock) 0.021
∗∗∗
(0.007)
Member connected to heavy industrial firms (direct interlock) 0.047
∗∗∗
(0.011)
Member connected to banks (direct interlock) 0.005
(0.011)
Baseline controls X X X X X
Constant 0.015 0.015 0.046 0.051 0.005
(0.059) (0.074) (0.054) (0.050) (0.059)
Observations 39 39 39 39 39
Adjusted R
2
-0.050 0.262 0.213 0.462 -0.069
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, leverage, liquidity, asset growth, board size.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0.01
48
Table B5: Bank share price percentage change 1899, OLS model
38 39 40 41
42
Member connected to the government (indirect interlock) -0.033
(0.022)
Member connected to the Finance Minister’s circle (indirect interlock) -0.058
∗∗
(0.023)
Member connected to non-financial firms (indirect interlock) -0.042
∗
(0.023)
Member connected to heavy industrial firms (indirect interlock) -0.058
(0.037)
Member connected to banks (indirect interlock) -0.081
∗∗
Constant
(0.030)
-0.141
∗∗∗
-0.140
∗∗∗
-0.132
∗∗∗
-0.140
∗∗∗
-0.127
∗∗∗
(0.023) (0.020) (0.023) (0.021) (0.021)
Observations 26 26 26 26 26
Adjusted R
2
0.028 0.103 0.094 0.078 0.158
Robust standard errors in parentheses. Univariate model.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
Table B4: Net investment losses 1899-1902, OLS model
33 34 35 36 37
Member connected to the government (indirect interlock) 0.023
(0.037)
Member connected to the Finance Minister’s circle (indirect interlock) 0.071
(0.044)
Member connected to non-financial firms (indirect interlock) 0.054
∗∗
(0.025)
Member connected to heavy industrial firms (indirect interlock) 0.088
∗∗
(0.035)
Member connected to banks (indirect interlock) 0.091
∗∗
(0.039)
Baseline controls X X X X X
Constant 0.009 -0.005 -0.016 -0.026 0.016
(0.061) (0.078) (0.054) (0.071) (0.048)
Observations 39 39 39 39 39
Adjusted R
2
-0.046 0.117 0.170 0.210 0.220
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, leverage, liquidity, asset growth, and board size.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
49
Table B6: Net investment losses 1899-1902, OLS model
43 44 45 46 47
Share of board connected to the government 0.211
(0.195)
Share of board connected to the Finance Minister’s circle 0.545
∗
(0.282)
Share of board connected to non-financial firms 0.184
∗∗∗
(0.047)
Share of board connected to heavy industrial firms 0.371
∗∗∗
(0.089)
Share of board connected to banks 0.251
(0.192)
Baseline controls X X X X X
Constant -0.016 -0.045 -0.051 -0.035 -0.015
(0.072) (0.083) (0.055) (0.062) (0.058)
Observations 39 39 39 39 39
Adjusted R
2
0.010 0.189 0.272 0.369 0.049
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, leverage, liquidity, asset growth,
and board size.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0.01
Table B7: Bank share price percentage change 1899, OLS model
48 49 50 51 52
Share of board connected to the government -0.257
(0.171)
Share of board connected to the Finance Minister’s circle -0.531
∗∗∗
(0.171)
Share of board connected to non-financial firms -0.191
∗∗∗
(0.066)
Share of board connected to heavy industrial firms -0.294
∗∗
(0.128)
Share of board connected to banks -0.202
Constant
(0.129)
-0.132
∗∗∗
-0.129
∗∗∗
-0.078
∗∗
-0.109
∗∗∗
-0.137
∗∗∗
(0.028) (0.022) (0.030) (0.027) (0.028)
Observations 26 26 26 26 26
Adjusted R
2
0.038 0.143 0.172 0.155 0.020
Robust standard errors in parentheses. Univariate model.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0
.01
50
Models (53) to (57) in Appendix Table B8 and models (58) to (62) in Appendix
Table B9 use yet another set of interlock variable, or the aggregate number of both direct
and indirect connections at a bank. These models support the results of the core models.
Models (63) to (67) in Appendix Table B10 and models (68) to (72) in Appendix
Table B11 use another method to calculate robust standard errors. I employ the iteratively
reweighted least squares algorithm which assigns a weight to each observation. The results of
these models confirm those of the core models. Note that, following Abadie et al. (2017), I do
not cluster standard errors, because I use the entire population of banks in my analysis.
Models (73) to (77) in Appendix Table B12 show the ninth and final robustness check
which uses Tobit models to address censoring issues. These issues arise because the
dependent variable, net investment losses, is censored at zero for banks which made net
investment profits during the crisis. These models confirm the results of the core models.
51
Table B8: Net investment losses 1899-1902, OLS model
53 54 55 56 57
Interlocks with the government 0.005
(0.010)
Interlocks with the Finance Minister’s circle 0.035
∗
(0.020)
Interlocks with non-financial firms 0.008
∗∗∗
(0.002)
Interlocks with heavy industrial firms 0.017
∗∗∗
(0.005)
Interlocks with banks 0.015
(0.013)
Baseline controls X X X X X
Constant -0.005 -0.007 0.031 0.031 0.008
(0.058) (0.080) (0.039) (0.051) (0.050)
Observations 39 39 39 39 39
Adjusted R
2
-0.038 0.175 0.491 0.382 0.028
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, leverage, liquidity,
asset growth, and board size.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0.01
Table B9: Bank share price percentage change 1899, OLS model
58 59 60 61 62
Interlocks with the government -0.009
(0.009)
Interlocks with the Finance Minister’s circle -0.021
∗
(0.012)
Interlocks with non-financial firms -0.004
∗∗∗
(0.001)
Interlocks with heavy industrial firms -0.011
∗∗∗
(0.004)
Interlocks with banks -0.025
∗∗
Constant
(0.009)
-0.143
∗∗∗
-0.141
∗∗∗
-0.103
∗∗∗
-0.113
∗∗∗
-0.126
∗∗∗
(0.026) (0.022) (0.027) (0.023) (0.024)
Observations 26 26 26 26 26
Adjusted R
2
-0.001 0.044 0.195 0.194 0.156
Robust standard errors in parentheses. Univariate model.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0.01
52
Table B10: Net investment losses 1899-1902, OLS model
63
64
65 66 6
7
Member connected to the government -0.013
∗∗
(0.005)
Member connected to the Finance Minister’s circle 0.063
∗∗∗
(0.016)
Member connected to non-financial firms 0.008
∗∗
(0.003)
Member connected to heavy industrial firms 0.018
∗∗∗
(0.004)
Member connected to banks -0.002
(0.006)
Baseline controls X X X X X
Constant 0.026 0.014 0.046 0.003 0.041
(0.032) (0.064) (0.033) (0.030) (0.038)
Observations 39 39 39 39 39
Adjusted R
2
0.276 0.286 0.171 0.375 0.072
Robust standard errors in parentheses, calculated using iteratively reweighted least squares.
Baseline controls are bank age, number of locations, leverage, liquidity, asset growth, and board size.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0
.01
Table B11: Bank share price percentage change 1899, OLS model
68
69 70 71 72
Member connected to the government -0.012
(0.012)
Member connected to the Finance Minister’s circle -0.032
∗
(0.017)
Member connected to non-financial firms -0.014
∗∗
(0.006)
Member connected to heavy industrial firms -0.021
∗
(0.010)
Member connected to banks -0.030
∗∗
Constant
(0.011)
-0.139
∗∗∗
-0.132
∗∗∗
-0.090
∗∗
-0.115
∗∗∗
-0.113
∗∗∗
(0.027) (0.024) (0.036) (0.029) (0.023)
Observations 26 26 26 26 26
Adjusted R
2
0.002 0.088 0.132 0.112 0.207
Robust standard errors in parentheses, calculated using iteratively reweighted least squares. Univariate model.
∗
p < 0.10,
∗∗
p < 0
.05,
∗∗∗
p < 0
.01
53
Table B12: Net investment losses 1899-1902, Tobit model
73 74 75 76 77
Member connected to the government 0.010
(0.016)
Member connected to the Finance Minister’s circle 0.061
∗∗∗
(0.021)
Member connected to non-financial firms 0.031
∗∗∗
(0.008)
Member connected to heavy industrial firms 0.046
∗∗∗
(0.009)
Member connected to banks 0.015
(0.013)
Baseline contols X X X X X
Constant -0.028 -0.001 0.012 0.002 -0.034
(0.078) (0.087) (0.058) (0.067) (0.073)
Observations 39 39 39 39 39
Robust standard errors in parentheses. Baseline controls are bank age, number of locations, leverage, liquidity, asset
growth, and board size. Dependent variable is zero for banks that made net investment profits over 1899-1902.
∗
p < 0.10,
∗∗
p < 0.05,
∗∗∗
p < 0.01
54
Appendix C. Government banking policy
C1. Government expectations
Given that it was crucial for Russia to industrialise, it is likely that government expected the
banking sector to finance industrial development. This was the second channel of government
influence. However, there are several reasons to believe that the government’s direct
influence, assuming that it was indeed present, probably had a minimal impact on individual
banks’ lending and investment decisions.
The following studies point that government influence was no more than an
expectation and surely not a requirement, because authorities have never explicitly told banks
how to conduct their business and what sectors or individual companies to finance. A detailed
study of financing deals by the St. Petersburg International Commercial Bank, a key banker
to the government, found no evidence of the bank being exploited as an instrument of the
state (Lebedev, 2003).
It is still possible that the government’s expectation carried considerable weight in
banks’ decisions, since the Ministry of Finance was both their regulator and supervisor. For
example, the ministry could decline a bank’s petition to open a new branch or deprive a bank
of its securities trading license by modifying its statute. This in effect would turn an
innocuous expectation into a powerful requirement. To address this issue, I examined banks’
statutes between 1895 and 1898 and found that that no statute was curtailed.
22
Also, an
archival study of the ministry’s policies vis-a-vis banks yielded no indication that the
ministry abused its regulatory power (Gindin, 1960).
It is also possible that the government had substantial leverage on the decisions of a
few banks which were bankers to the government, because it could threaten the loss of
lucrative government business. I conclude that the government’s expectation probably did
22
Banks’ statutes are from Complete Collection of Laws (1899a, 1899b, 1900, 1901).
55
have a meaningful impact on these banks.
23
However, given that heavy industrial firms were
highly profitable and banks’ profitability was never at immediate risk, the question of how to
balance industry financing without damaging the relationship with authorities was probably
not an issue these bankers had to take seriously. The other banks were not bankers to the
government and thus had no risk of losing government business.
C2. The State Bank’s policy
The third channel through which the government affected banks’ decisions was the
protectionist policy of the State Bank regarding banks. I argue that this policy, too, had a
minimal impact on inducing banks to finance industry.
It is known that the State Bank was highly protective of banks under stress ever since
the first bank failure in 1875 (Gindin, 1960). Indeed, only nine banks were liquidated after
the establishment of the first commercial bank in 1864 (Gindin, 1960). Keeping banks stable
was important for another reason: on the eve of the crisis, banks supplied well over half of
industrial financing. Modern economists point out that the banks which expect to receive a
bailout tend to increase their risk-taking activities (Stern & Feldman, 2004).
If the banks were indeed counting on a bailout, then we should expect to have seen
them hastily extending new loans in the first months of the crisis, seeking to tie themselves to
the industrial companies which were so important for industrialisation. To test whether this
ever happened, I track the change in the bank loans that were most indicative of heavy
industry financing.
24
I find that the eight banks which were most likely to be bailed out by the
State Bank reduced their lending by 27 per cent on average between January and December
23
How far each bank provided financial services to the government is unknown, and thus this factor
cannot be controlled for in regression analysis.
24
Namely, I calculate the change in term loans backed by corporate securities and call loans backed
by government and corporate securities. Loan quantities are from Golubev (1905).
56
1899.
25
This suggests that these banks did not engage in last-minute gambling for resurrection
and, therefore, had not been counting on a bailout.
Furthermore, even in non-crisis times, the State Bank provided credit to nearly all
banks and rediscounted their bills in order to facilitate their functioning as credit suppliers to
the economy (Gindin, 1960). Thus, the banks constantly receiving more financial assistance
from the State Bank may have been more inclined to engage in industry financing. To test
whether this even happened, I examine the allocation of State Bank loans to banks. In 1898,
these loans amounted to merely 6.1 per cent of banks’ lending to the economy.
26
If these
loans were evenly distributed between banks, then the effect of the State Bank can be
considered insubstantial. I estimate that most of the State Bank’s assistance went to banks in
the provinces.
27
This suggests that the impact of the State Bank’s lending programmes did not
much matter; otherwise, we would have seen provincial banks incurring large losses during
the crisis, but only 4 out of 25 provincial banks experienced sizable distress.
Appendix D. The choice of dependent variables
I have chosen net investment losses and bank share price as my distress measures for four
reasons. First, given that both measures were determined by stock market participants, these
measures were least affected by banks’ fraudulently understating losses on income statements
(Bovykin, 1984). Although banks did tend to overestimate their investment assets (Lebedev,
25
I consider the banks which sustained ten per cent or greater investment portfolio losses over the
crisis to be the most likely candidates for bailout.
26
Bank credit data are from Golubev (1905). The State Bank’s loans are from the State Bank (1899).
27
As of December 1898, out of 46.3 million rubles of total assistance to banks, 10.8 million was
distributed by the State Bank’s St. Petersburg office (State Bank, 1899), which almost exclusively
assisted banks based in St. Petersburg and Moscow (RSHA, 588, 1, 247; 588, 1, 277; 588, 1, 593),
while 35.5 million was given out by Moscow and provincial offices, which lent to banks based in
Moscow and in provincial centres.
57
2003), there was a mandatory requirement to mark investment portfolios to market every
year.
28
This mark-to-market requirement forced banks sooner or later to document investment
losses during the three-year-long stock market downturn.
Second, net investment losses capture the reintermediation of collateral onto banks’
balance sheets throughout the crisis as heavy industrial companies defaulted on their loans.
As a result of this reintermediation, banks’ portfolio holdings increased by an estimated 55
per cent between January and December 1899 alone.
29
Indeed, an annual report of the St.
Petersburg International Commercial Bank (1901) documented the occurrence of this
takeover of collateral. This massive reintermediation was one of the primary reasons why
banks sustained heavy losses. Thus, using this variable, I can capture banks’ involvement in
heavy industry financing – exactly the type of data that would have been ideal for capturing
exposure to government policies.
Third, both measures of distress were less affected by the substantial anti-crisis
assistance provided to banks by the State Bank, than by other distress measures, such as net
income. In particular, the share price variable was probably least affected because a state-
funded investment fund, introduced to support the value of banking and industrial shares,
began to acquire shares mainly in 1900, after the period over which the variable is
calculated.
30
28
The statute of each individual bank specified the annual mark-to-market requirement.
29
Own estimation, calculated as the increase in corporate securities on banks’ balance sheets, in a
rapidly and constantly declining securities market.
30
Although the investment fund was established as early as October 1899, it seems to have begun
acquiring large bundles of shares only in 1900. I determine this by using Bugrov’s (2003) data on the
composition of the fund’s portfolio in February 1901 and the price for which the shares were acquired
by the fund. I then match this price with the actual market price, drawn from Goetzmann et al. (n.d.b),
to identify the month and year when the acquisitions were probably made. This reveals that the fund
bought shares of 20 companies at some point in 1900 or later, and of three companies in 1900 or
possibly in 1899.
58
Fourth, net investment losses, which are censored at zero for banks which made a net
profit, are preferable to net investment profits, which would not be censored at zero. From
archival evidence we know that some banks simply deemed it inappropriate to finance heavy
industry. They made net investment profits over the crisis, despite having an above-average
number of board members connected to government and the Finance Minister’s circle
(Gindin, 1958; Gindin, 1960).
31
Therefore, using net profits in regression analysis would not
capture the true relationship between the presence of interlocks and bank distress.
31
The Moscow International Trade Bank and the Central-Asian Commercial Bank are examples.
59
