Data Sources on the 19th and Early 20th Century German Capital Market: Challenges and Opportunities

1 Introduction

Ever since Gerschenkron (1962)’s seminal work on German industrialization, Germany has been viewed as a model for a bank-based financial system, where universal banks played a pivotal role in its economic rise. This discourse often contrasts the German financial system with the American one, characterized by a well-developed stock market (Calomiris 1995). However, despite the prominence of Germany’s banking sector in the 19th Century, the country also reaped significant benefits from its large and efficient Berlin stock market, which boasted a market capitalization above the world average (Rajan and Zingales 2003) and rivaled other major stock exchanges (Lehmann-Hasemeyer and Streb 2016). Beyond Berlin, Germany was home to numerous regional stock exchanges until 1934 (Lehmann-Hasemeyer and Burhop 2014).

The data on the German capital market is particularly rich for the period following the Stock Exchange Act of 1896, which remains in effect today and mandates comprehensive publication requirements for stock corporations. This act significantly enhanced the transparency and availability of financial information, laying the groundwork for modern financial reporting standards in Germany. In this article, we provide an overview of the data and sources available for the German stock market from the long 19th Century and the Weimar Republic. Despite the extensive literature on German stock corporations and stock exchanges, much of the data and sources on this topic have been recorded and digitized only in fragments. This is due to the sheer volume of sources and data, much of which exists in text format. Consequently, indexing and utilizing this data has been complex and costly. However, with the rapid development of AI-driven data collection methods, a significant portion of this data is expected to become available in a structured format in the next years. Therefore, it is crucial to understand the distinctive nature and limitations of this data to utilize it effectively on a large scale.

We aim to provide a comprehensive overview. However, we cannot discuss all sources in detail, but have to make a selection. We have concentrated on our own research topics, as this is where we have the greatest expertise. This survey article will therefore focus on company data, initial public offerings (IPOs) and price data for traded companies. The scope of this overview does not extend to data on bonds and other securities traded on German stock exchanges. Along with the data description, we review literature that has utilized this data, noting that these papers are exemplary rather than exhaustive in covering the related research.

This article is structured as follows: We begin with an examination of sources that provide firm-specific information on both listed and non-listed firms. Here, we discuss the peculiarities, content, and limitations of handbooks such as the Saling Börsenhandbuch and the Handbuch der Deutschen Aktiengesellschaften, and offer guidance on where to access structured fragments of this data. We dedicate the subsequent section to the discussion of the data of IPOs. Next, we briefly discuss selected data sources that can be found in company, bank and economic archives. This is essential, because we consider it important that a researcher dealing with the German capital market not only analyses official data, but also studies the conditions of the market and how the data is generated in a thorough manner. This presupposes that one analyses capital market transactions, ownership structures and investment decisions in detail and also takes a close look at the primary sources. We discuss the availability and particular importance of price data as a measure of information and investor sentiment before reviewing available historical stock market indices and presenting a novel daily stock market index for the banking sector based on banks listed on the Berlin Stock Exchange between 1896 and 1930. The concluding section presents a synopsis of the main points.

2 Firm-Specific Information for Joint Stock Firms

A good starting point for company-specific data on public stock corporations listed on a German stock exchange is Salings’ Börsenhandbuch. In this series, the first handbook that was published was titled Die Norddeutschen Börsen-Papiere. It contains information on all domestic and foreign government and premium bonds, mortgage and annuity bonds, bank, industrial, and railway securities listed on the Berlin Stock Exchange 1868–1869. Salings’ Börsenhandbuch was published annually (with gaps) from 1870 by the Berlin publishing house Haude und Spener and later by the Verlag für Börsen- und Finanzliteratur, Berlin/Leipzig/Hamburg. In 1935, it became the Saling Aktienführer, which was published annually until 1995 and subsequently became Hoppenstedt Aktienführer. This handbook is the most comprehensive collection of information on listed companies, known for its reliability and the breadth of data it covers, making it a crucial resource for financial analysts and historians.

The Saling Aktienführer and its predecessors consists of several series, not all of which cover the same period. Part 1 contains general information on the mode of operation of the stock exchange (1870–1932). Part 2 contains information on special features and securities traded on the Berlin Stock Exchange (1870–1935), while Part 3 contains information on special features and securities traded on the German local stock exchanges. This series existed only from 1900 to 1932. This series is particularly interesting because although Germany is famously characterised as a bank-based system, it also had a large number of local stock exchanges. Most importantly, the number and location of local stock exchanges varied over time. On the eve of the First World War, there were 22 local stock exchanges operating in addition to the main exchange in Berlin. In December 1934, the National Socialists significantly reduced the number of stock exchanges to nine under the “Law on Securities Trading” (for further information see (Lehmann-Hasemeyer and Burhop 2014)). It is unfortunate that the series of handbooks covering local stock exchanges ends in 1932, as the significant developments in regional stock exchanges that took place in 1934 are not covered.

Part 4 is distinct in both style and content from the other parts of the handbook. It contains information and addresses of banks and bank branches available for the years 1928–1931. Unfortunately its publication stops just before the banking crisis of 1931. It is noteworthy that this volume does not merely cover listed banks, but provides a comprehensive coverage of all banks.^[1]

Part 5 of the handbook contains information on companies listed on the important international stock exchanges (1928–1931), i.e. Amsterdam, Brussels, Budapest, London, Milan, New York, Paris, Prague, Stockholm, the Swiss exchanges and Vienna. As the editors noted in the preface to the 1928 volume, this was in response to German investors’ demand during the inter-war period for information on foreign stock exchanges and the securities listed there.

After 1935, when the handbook became the Saling Aktienführer, there was only one series. In the early years it contained only information on companies listed in Berlin. During the Second World War there was a short interruption in its publication, but it was resumed in 1953, now containing information on all listed companies, not just those listed in Berlin. The University Library in Mannheim has already digitized a large part of these volumes. Earlier volumes are available in PDF format, while later editions have already been transformed into structured data.^[2] This digitization effort, involving significant work by librarians and researchers, ensures that historical financial data is preserved and accessible for modern analysis. The volumes contain detailed information on the company’s name and sector, headquarters and address, CEO, board of directors, date of incorporation, share capital, current balance sheet, dividends, listing information, as well as a brief history and information for the current year. However, this additional information varies considerably from company to company, reflecting inconsistencies in historical reporting standards and practices.

Despite its comprehensiveness, Salings’ Aktienführer and its predecessors do have some shortcomings: There are gaps in the data, especially for smaller companies or those that were listed for shorter periods and during the war. Additionally, the quality and completeness of the information can vary, making it sometimes challenging to compare companies directly. The crucial limitation, however, is that the Salings’ Aktienführer, with the exception of the banking series, only contains information on listed companies.

However, this gap is filled by Handbuch der Deutschen Aktiengesellschaften (HDAG). It was first published in 1896 and continued to be published annually (with gaps during the Second World War) until 1999. The content of the handbook is comparable to that of Salings’ Aktienführer, and according to the editors, more reliable, although it is usually less detailed for each company: In the foreword to the 1953 edition, the editors state that the Saling Aktienführer is of particular relevance to investors currently active in the capital market, whereas the HDAG provides in-depth reports on all German stock corporations and remains the most reliable source of information for those seeking comprehensive insight into the operational and financial performance of a stock corporation. As stated before, its largest advantage it that it also covers all joint stock companies that were not listed on a stock exchange. However, it is strongly recommended that researchers consult both handbooks in order to verify the data used in their research projects involving listed companies. The volumes of the HDAG are also available from the University Library in Mannheim in PDF format.^[3] At the time of writing this article, the volumes up to 1950 were available there. However, the library plans to continue digitizing and making available later volumes.

Overall, there is reliable annual information on joint-stock companies for the period after 1870. Some of this information has already been digitised and used in various projects. However, the information on individual companies varies and is mainly in text format, making it time-consuming and costly to extract into a format suitable for quantitative analysis. The SAFE Institute in Frankfurt is undertaking a data project to collect all this information, along with price data.^[4] So far, however, most researchers only extract information for subsamples of benchmark years.

However, a few papers have been published that collect all listed joint stock companies at a specific point in time. Burhop and Lehmann-Hasemeyer (2016), for instance, collected the number of firms listed on any of the German stock exchanges, their market value and industry characteristics from the Saling Aktienführer for the benchmark year 1913 to understand the geography of stock market listings.^[5] At the end of 1913, there were 1,579 German firms listed on a German stock exchange with a total market value of 20.27 billion Marks, which equaled 38 percent of the German net national product. In addition, 22 foreign companies were traded (Burhop and Lehmann-Hasemeyer 2016, p. 435). In another paper, the same authors further collected the benchmark year 1937 from the Saling Aktienführer and compared it with the 1913 data (Lehmann-Hasemeyer and Burhop 2014). In this year, a total of 880 German companies with a share capital of 10.6 billion Reichsmarks (RM) and a market value of 14.1 billion RM were traded on the remaining nine stock exchanges. No foreign companies were traded on a German stock exchange at the time. The market value corresponded to approx. 14 % of the net national product, less than half of what it was in 1913. In preparation for this article, the two benchmark years were published via Emporion and are now available online.^[6]

Another paper worth mentioning in this context is Lehmann-Hasemeyer and Opitz (2019). They aim at studying the impact of political connections of firms listed on the Berlin stock exchange in the 1920s via current members of parliament on supervisory boards and board of directors. The authors collected the information on the firms listed in Berlin as well as their board members at the end of 1924 from the above discussed handbooks. In this year, they count 1,064 companies, with a nominal share value of 9,759 million Reichsmark and a market value of about 7,000 million Reichsmark. This corresponds to a market capitalisation of between 11 and 15 per cent of GDP, which is about half the share estimated by Rajan and Zingales (2003). The average firm had a share capital of about RM 9.3 million, was about 33.6 years old and had a board of 9.4 members. The most dominant group was heavy industry, which accounted for about 23 per cent of all firms, followed by light industry. To get a second reference year close to the May 1928 election, they checked which of the sample firms were still listed in December 1928 and concluded that the total numbers remained relatively constant.

3 Initial Public Offerings (IPOs)

An IPO is when a private company sells shares of its stock for the first time to the public and thus becomes a public company. In contrast, the Seasoned Equity Offering or SEO are the follow on offering issue of stocks for sale by a company. According to the Stock Exchange Act of 1896, every newly approved security had to be published in the “Vierteljahreshefte zur Statistik des Deutschen Reiches”. Figure 1 shows a sample page. The table on the sample page includes the exchange on which the shares were issued as a subheadline (here Berlin). From left to right, the table provides the date of the issue, the name of the firm (or the securities – as not only firm shares are included), the value of the shares, the underwriter, the issuing price, the first market price (important for calculating the initial return on the market, i.e. underpricing), and notes. The notes indicate whether the shares were also issued on another stock exchange and the number of such issue. For the period 1897 to 1913, this source allows for the easy collection of data on IPOs and SEOs across all German stock exchanges.

Figure 1:

IPOs in the Vierteljahreshefte zur Statistik des Deutschen Reiches – sample page. Sources: Vierteljahreshefte zur Statitik des Deutschen Reiches, 1903 (1), page 215.

However, there are a few things to consider: First, distinguishing IPOs from SEOs is not always straightforward. Generally, if it says “New Shares” (Neue Aktien, line 3 in the table), this indicates an SEO rather than an IPO. Second, if a share is issued on more than one stock exchange, it also appears in the list of the other exchanges. Simply extracting all issues from the table would lead to double counting. It is also useful to check whether the issue on the other exchange actually took place on the same date. Often we see a clear IPO on one exchange and then the shares start trading on another exchange a few weeks later. The third and probably most important point is that the share capital indicated here is not necessarily the capital that was actually sold on the market. A comparison of the amount given in the Vierteljahrshefte with the share capital in the stock exchange handbooks shows that it corresponds to the total share capital and that it is unlikely that the total share capital was offered in each issue. This is problematic when studying the initial return on the market, as it makes a significant difference whether the entire capital was sold or just a negligible portion of it. Unfortunately, this issue persists since this information is not disclosed elsewhere.

Getting data for periods before 1896 is much more difficult. To illustrate the effort, see for example Burhop (2013), who has studied the underpricing of IPOs in the period 1870 to 1896. He collected data from several sources: For the period 1870–81, the names of newly listed companies and their year of listing were available from Van der Borght (1883). Information on the initial capital, the name of the lead underwriter, the issue price and the IPO date was collected from the Berliner Börsenzeitung, Meyer (1873) and Salings’ Börsenhandbuch (1875, 1882/83). Burhop could not rely on the Salings’ Börsenhandbuch alone because it does not always clearly indicate when the companies actually started trading on the stock exchange.

There are several papers that have studied historical IPOs in Germany, for sub-periods and some of this data is available online. Burhop (2013), Lehmann (2014) and Lehmann-Hasemeyer and Streb (2016),^[7] for instance have studied the listing decisions of IPOs in the 19th Century and the role of Berlin and other stock markets as a source of capital. Burhop, Chambers, and Cheffins (2018) and Kozik (2021) have studied IPOs until World War II. Lehmann-Hasemeyer, Dwenger, and Ehrhardt (2024) are currently working on a research paper that will present a comprehensive analysis of the IPOs of all these previous subsamples and extend them to 2016.^[8] This data will be made available via Emporion once the paper is accepted for publication.

4 Archival Sources

In addition to the official statistics mentioned above, researchers should also consult historical archives. The wealth of sources contained in these archives allows researchers to better assess the data from official statistics, for example by learning more about the procedures of an IPO, pricing strategies, but also more detailed information about the shareholders or the company. This section presents just a few of these sources as examples to show how valuable and rich the information they contain can be.

To begin with, archives can provide information on the ownership structure of companies. Ownership data for Germany is generally scarce as most shares in German companies were bearer shares and traded anonymously. Only those shareholders who attended the company’s annual general meeting (AGM) were required to register their names and shares. With the Stock Exchange Act of 1896, companies were legally obliged to submit information on shareholders attending general meetings to the stock exchange on which their shares were listed (Franks, Mayer, and Rossi 2008). It is therefore possible to collect these lists and thus information on the structure of attending shareholders after 1896. For the period before 1896, we have to rely on voluntary information. Even after 1896, when the meetings were reported every year, the archives only kept a sample of the lists. The data on the meetings can be found in various archives, such as the Hessian Economic Archive, the Bavarian Economic Archive, the Baden-Württemberg Economic Archive, the Historical Archive of Deutsche Bank AG, the Historical Archive of Commerzbank AG as well as the Bundesarchiv in Berlin (listing office – Zulassungsstelle). Of course, the information only covers owners who attended the meeting. However, these are the investors who have made the effort to attend the meetings in order to influence the future of a company and are therefore most likely to be those with the largest shareholdings. Based on a sample of lists, Lehmann-Hasemeyer and Neumayer (2022) show that on average about 60 % of the capital was present at the meetings.

Lehmann-Hasemeyer and Neumayer (2022) examine the social structure of ownership in German joint-stock companies from 1869 to 1945 using attendance lists of annual general meetings. They confirm that, despite changes in the economic and political environment, ownership remained predominantly in the hands of a few male, insider investors, and find only a minimal influence of women and lower class investors. They also identify bankers at the meeting and quantified their influence on the decision-making process.

Although the attendance lists are incomplete and probably not fully representative, they provide a good insight into the structure of shareholders who exercised their voting rights at annual general meetings. But again, caution must be exercised in interpreting them. These lists are not standardised. Sometimes one can extract the full name and profession of attendants, and also their function, such as “bank representative”, as well as the corresponding capital or votes. Sometimes the lists contain just names with no further information. Often but not always the place of residence is provided. In previous studies, the fact that the share of local investors is disproportionally high was interpreted as a so-called home bias, i.e. that investors also prefer to invest in companies in their place of residence (see for instance Burhop and Lehmann-Hasemeyer 2016). However, Neumayer (2018) has shown that this probably has more to do with the cost and effort of attending. If the AGM is held elsewhere than at the company’s headquarters, the majority of investors attending are still likely to be local, even if the company’s headquarters are further away. Home bias therefore relates to travel, not investment.

Furthermore, shareholders of German companies had the option of appointing banks or other third parties to exercise their voting rights on their behalf. This practice, designated as proxy voting, has the potential to distort measures of ownership concentration. For example, if we observe a banker or a bank with a significant proportion of the votes, it is possible that although the shares are counted as being in the hands of one investor, they may be dispersed because the banker acts as a proxy for a large number of smaller investors. Furthermore, proxy voting permitted banks to exert control over companies without actually owning their shares. Another challenge in identifying the impact of banks is that the bankers often appear only as names on the list, with no information about their occupation or whether they are acting on behalf of the actual capital owners. Lehmann-Hasemeyer and Neumayer (2022) have carefully reviewed and compared the names on the attendance list with lists of bankers and bank CEOs to identify and assess this potential bank influence.

Moreover, analysing archival documents and records can help shed light on the decisions of investors. Occasionally, portfolios of individual shareholders can be found. Lehmann-Hasemeyer and Neumayer (2019), for instance, have studied the portfolio choices of the private banker Joseph Frisch. They find that the preference for local shares was highest in times of insecurity, low returns and reduced investment activity. With higher returns, a stable and growing economy and more experience, the preference for local shares decreased.

The Law of Deposits (Depotgesetz) of 1896 required banks to keep deposit books. The deposit books record, for each client, every transaction, and after each transaction, the holdings in the respective security. The Law of Deposits ensured that the information on transactions and holdings in these books is comprehensive. Furthermore, the deposit books provide insight into various investor characteristics, including the clients’ place of residence and whether they hold accounts at other financial institutions. However, the quality of the information of these deposit books vary and archives often just keep a small sample of these books. Recently, Braggion, Von Meyerinck, and Schaub (2023) were fortunate to obtain deposit books from a German bank. These books encompass information on approximately 3,000 private clients with a security portfolio during the specified sample period 1920–1924. Unfortunately the bank did not permit the researchers to reveal the name of the providing bank or share the data, which may have been of interest to other researchers. This is possible because banks are not subject to the usual time limits for archiving. Banks are free to withhold information even if there is no legal obligation to do so. This is particularly noteworthy in the context of historical data, where the customers in question may no longer be alive and therefore not subject to data protection regulations.

5 Share Prices and Market Reactions

Price data for listed firms are regularly published on a daily basis in the evening edition of the Berliner Börsenzeitung. This newspaper, available in PDF format via the State Library in Berlin from 1872 to 1930 without any gaps, serves as a comprehensive source for historical prices at the Berlin stock exchange. Researchers can access both stock prices and dividend information through this newspaper.^[9]

However, there are challenges using this data. The Berliner Börsenzeitung often uses abbreviated or non-standard names for firms, complicating efforts to match data with data from the sources discussed above. For example, the pencil producer “Faber Castell” is listed under the name “Bleistiftfabrik” (pencil factory) in certain years. Moreover, there are occasional gaps in the reported data caused by the fact that the shares were not traded on that day. While earlier editions of the Berliner Börsenzeitung explicitly noted these gaps, later editions simply listed firms for which prices were available on a given day.

A significant limitation is that the Berliner Börsenzeitung only covers firms listed on the Berlin Stock Exchange. Researchers interested in firms from other exchanges must consult local newspapers, which do not consistently report prices. Consequently, much of the research on German capital markets focuses predominantly on the main exchange in Berlin. Information on how the prices were calculated can be found in Gelman and Burhop (2008) and Burhop and Gelman (2022).

To use these prices effectively and not misinterpret them, researchers must possess a solid understanding of the workings of the German capital market and cross-validate their data with other reliable sources. To give an example, Fohlin (2010) examines initial returns of IPOs and SEOs on the Berlin Stock Exchange from 1882 to 1892, arguing that underwriters significantly benefited from purchasing new issues below market value, termed the “underwriter premium.” She calculates this premium as the percentage difference between the offering price and the par value of a share (Fohlin 2010, 642). However, Burhop (2013) has rightly criticised the results, showing on the basis of underwriters’ contracts that underwriters generally paid well above par, thus disproving Fohlin’s assumption. This correction leads to a significant downward revision of the estimated risk premium. This example illustrates the pitfalls that researchers can encounter when interpreting historical price data.

With prices from different exchanges, information flows (Baltzer 2006) and market integration (Weigt 2005) can be analysed. Moreover, price data is particularly useful in the historical context, as it can be used to identify investors’ expectations. Currently, economists and politicians can conduct surveys to gain insight into the preferences of various economic actors. This is not possible in a historical context, as many of the key actors are no longer alive. Consequently, there are limited methods for assessing or quantifying expectations, moods, and reactions in a historical context. In economic history, the principle of “revealed preferences” is often employed, whereby actions provide insights into expectations.

In principle, the process is simple. Prices provide information about investor demand and are available on a daily basis. By analysing the price changes of particular companies in response to events, using the event study method,^[10] it is possible to infer how investors in these companies have valued these events. This method can be applied to any context including company-specific events like mergers, as discussed by Lübbers (2008), as well as historical political events, as in Lehmann-Hasemeyer, Hauber, and Opitz (2014). The latter study seeks to understand the economic effects of democratisation at the end of the 19th Century by studying investors’ reactions to the extension and restriction of the suffrage in the Kingdom of Saxony in 1896 and 1909. This was done by linking their investment behavior, i.e. the prices of firm based in Saxony traded on the Berlin Stock Exchange, to political events in Saxony. The study by Lehmann-Hasemeyer, Hauber, and Opitz (2014) shows that investors on the Berlin stock exchange did indeed react negatively to the extension of the suffrage in the Kingdom of Saxony.

The event study method can also be used to analyse how investors with different information react to the same event by comparing the effects on prices on different exchanges. This is done, for example, by Opitz (2018). He examines the impact of the Russian Revolution of 1905 on contemporary capital markets in Germany and in Russia by applying an event study to Russian government bonds in Berlin and Saint Petersburg. This paper also provides information on the degree of integration of the two markets.

In order to use the event study methodology effectively, two factors need to be considered. First, the event to be studied must be truly unexpected. If the effects of the event are anticipated, they are likely to have been ‘priced in’ by the market. This anticipation reduces the methodological effectiveness, as the event study aims to measure the market’s reaction to new information. Ensuring that the event is unexpected allows for a more accurate assessment of its impact. To determine the exact moment when investors were informed of the event, one must examine their sources of information. This could be, for example, articles from major contemporary newspapers. These newspapers are particularly useful because they are published daily, allowing the timing of the event to be isolated. In addition, clear and prominent discussions in these newspapers ensure the importance of the event for investors, leading to a measurable effect.

Second, selecting a meaningful control group is also essential. This group should remain unaffected by the event but share a similar investor base or fall within the scope of potential investors. This ensures that the control group accurately reflects market conditions and investor behavior, providing a reliable benchmark for comparison.

6 Performance Indices

6.2 Introducing a Daily Bank-Performing Index

We began our brief overview with the much-discussed importance of German banks and conclude with a daily stock market index of banks listed on the Berlin Stock Exchange. We will use this index to illustrate the steps required to construct such an index. In constructing the bank performance index, we follow the methodology outlined by Gelman and Burhop (2008). In a first step, we collected the pdf pages for every day from the Berliner Börsenzeitung from January 1920 until June 1931, shortly before the banking sector collapsed (Straumann 2020). From these pages, we collected the prices of all listed banks.^[14] To ensure our index includes only actively traded banks, we exclude institutions that lack price quotations for at least one-third of the time period, following the recommendations of Gelman and Burhop (2008). In addition, we exclude banks that were not continuously listed on the Berlin stock exchange until the end of the sample period. As a result, Danatbank and its predecessors, Darmstädter Bank and Nationalbank für Deutschland were removed from the index. The index is based on a total of 38 banks and approximately 80,000 prices.

Based on the prices, we then calculated returns, which serve as the basis for the index. The banks’ end-of-year stock market valuations are employed as weights in a Laspeyres index weighting scheme. Consequently, the influence of the various institutes differs over time.

Moreover, to calculate a performance index from observed price data, several corrections were necessary. First, we adjusted performance for dividend payments. Second, alterations to the share capital, in addition to dividend payments, have a direct impact on the stock price. It is important to note that in the past the financial year varied more from company to company than it does today, so it is essential to look at when the dividend was paid. To address this, data on the amount of dividend payments and changes in share capital, along with their exact timing, were collected from the Salings’ Börsenhandbuch and included in the returns. Unlike the weekly and monthly indices, the daily frequency of prices allows for precise adjustments to be made for dividends on their payment dates.

When visualising the index, we distinguish between the period before and after the hyperinflation, as the exploding prices make a useful graph difficult. Figure 2 plots the index together with the DAX-30 blue-chip index from Ronge (2002) up to 1922, and Figure 3 plots the index after 1924. The figures show that before 1922, German banks performed quite similarly to the rest of the market. After the hyperinflation, the situation is different. While the Ronge index is more or less stagnant, the banks perform disproportionately well, as can be seen in Figure 3. We assume that the losses in value of banks were particularly high in 1923 or as a result of the currency reform because banks did not own any inflation-hedged physical assets.

Figure 2:

Banking index and Ronge index (2. Jan. 1920–30. Nov. 1922). Sources: Banking index: own calculations, Ronge index: Ronge (2002), see also Table 1.

Figure 3:

Banking index and Ronge index (4. Jan. 1924–11 July 1931). Sources: Banking index: own calculations, Ronge index: Ronge (2002), see also Table 1.

The fact that we have collected the prices for all banks allows a further differentiation between the large universal banks, the so-called “D-banks”,^[15] and the other banks. Figure 4 shows the performance of the two groups in the early 1920s. While performing quite uniformly at first, D-banks outperformed the smaller one as the inflation approaches. Even in the aggregate index form, prices may well reveal preferences: investors may have seen the larger banks more capable of dealing with the difficult economic situation.

Figure 4:

Banking index for different bank types (2. Jan. 1920–30. Nov. 1922). Sources: Banking index: own calculations, see also Table 1.

Figure 5 shows the situation after the hyperinflation. There is a general upward trend, consistent with the historical situation in the late “golden” 1920s. This time, however, it is the smaller banks that have outperformed the big banks for almost half a decade. Investors may have lost confidence in the D-banks during the crisis, or the smaller institutions may have been able to adapt more quickly to the new economic circumstances. More in-depth research is needed to clarify this aspect.

Figure 5:

Banking index for different bank types (4. Jan. 1924–11 July 1931). Sources: Banking index: own calculations, see also Table 1.

7 Summary

This article offers an overview of the most prominent sources for analyzing the German capital market in the period 1870 to about 1945. Our review includes handbooks such as Saling Aktienführer and its predecssors and the Handbuch der Deutschen Aktiengesellschaften (HDAG), which provide comprehensive annual information on listed and non-listed joint stock companies. These sources are indispensable for financial analysts and historians, despite certain gaps and inconsistencies.

We also discuss data on IPOs, primarily from the V. zur S. des Deutschen Reiches, emphasizing the challenges and nuances of studying IPOs. Additionally, we highlight the importance of archival sources, such as ownership records and deposit books, for gaining insights into the historical market structure and behavior of investors.

A significant contribution of this paper is the introduction of a novel daily stock market index for banks listed on the Berlin Stock Exchange, based on data from 1920 to 1931. This index highlights the potential of detailed price data for market analysis, demonstrating how such data can be used to understand investor sentiment and market development. Table 1 provides an overview of the data sources available online that we have discussed in this survey article.