Price Quotations in Early U.S. Securities Markets, 1790-1860: Description of the Data Set

Compilers:

Richard E. Sylla, Stern School, New York University,

Jack Wilson, North Carolina State University,

Robert E. Wright, Winthrop Group,

Abstract:

This document describes a large data set of prices of public securities (equities and bonds) that traded in 9 U.S. securities markets and London between the end of the Revolution and the Civil War. The description includes a basic orientation to the data, its electronic formats, its limitations, and two examples of its potential use. It also includes a brief bibliography and a full list of the periodicals consulted.

Acknowledgment:

Funding for this data set was provided by the National Science Foundation, grant no. SES-9730692, “America’s First Securities Markets, 1787-1836: Emergence, Development, Integration.”

Table of Contents:

Project Overview 2

Geographical, Temporal, and Sectoral Scope of the Data Set 2

Electronic Format of the Data Set 4

Data Limitations 7

Data Interpretation Examples 10

For Further Guidance 13

Appendix: Periodicals Consulted 14

Project Overview

Over a period of several years, members of the Project, which was composed of two P.I.s (Richard Sylla and Jack Wilson), a postdoctoral associate (Robert E. Wright), a database assistant (Namsuk Kim), and several undergraduate research assistants, scanned early U.S. periodicals (1786-1862) for the prices of public securities -- public and corporate debt and corporate equities. (Details of the sources consulted, the years searched, and the general usefulness of each consulted source are provided in the Appendix.)

Not every extant newspaper for every year was searched because the purpose of the project was not to find every extant securities quotation, but rather to find “runs” of securities quotations (daily, weekly, or monthly) in each of the major securities markets (Boston, New York, Philadelphia, Baltimore, Charleston, S.C., New Orleans, London, and, to a lesser extent, Alexandria, Norfolk, and Richmond, Va.), over the core period under study, 1790 to 1850. (Quotations before and after that period were also recorded where time, financial resources, and primary sources allowed. See below for details.) Quotations of inland and foreign exchange, coins and specie, and non-interest bearing cash instruments like bills of credit and banks notes were not included in the data set.

Geographical, Temporal, and Sectoral Scope of the Data Set

The following table provides an overview of the geographical and temporal coverage of the data set:

Market / Earliest Date / Latest Date / Number of Weeks of Prices Recorded
Alexandria / 1806.0929 / 1814.0528 / 246
Baltimore / 1803.0214 / 1862.0531 / 2,912
Boston / 1789.0303 / 1859.1214 / 3,792
Charleston, S.C. / 1787.0315 / 1860.1217 / 2,639
London / 1792.0606 / 1843.1229 / 1,363
New Orleans / 1810.1020 / 1858.1002 / 574
New York / 1790.0331 / 1853.1228 / 3,493
Norfolk / 1804.0804 / 1819.0506 / 250
Philadelphia / 1786.0816 / 1850.1228 / 3,148
Richmond / 1799.0914 / 1859.1230 / 18,417
Total / 1786.0816 / 1862.0531 / 8,649

Dates are rendered above as in the database, i.e., in the format of year.monthday, where the year is a four digit number to the left of the decimal, months are rendered by the two digits immediately to the right of the decimal, and days find representation in the final two digits. Thus, 1799.0914 indicates a date of 14 September 1799 and 1834.1001 represents 1 October 1834. A date of 1800.0100 represents an unknown day in January 1800.

The data is divided first by market, then by issuer sectors or types. The types include the following:

Issuer Type / Number of Issues
Banks and Trusts (Equities) / 300
Insurance Companies (Equities) / 221
Transportation Companies (Equities) / 305
Manufacturing, Miscellaneous, and Utility Companies (Equities) / 358
Corporate Debt (Bonds) / 289
U.S. Debt (National bonds) / 135
State Debt (State bonds) / 386
Municipal Debt (City and county bonds) / 176
Total / 2,170

The total number of different issues (2,170) is somewhat inflated by the fact that the project members erred on the side of splitting the same security under separate “key codes” rather than risk lumping disparate securities together under a single key code. For example, we discovered quotations for New Creek Coal, New Creek Iron, and New Creek Water. Given the similarity of the names, there is a possibility that this was a single firm, the full name of which might have been New Creek Coal, Iron, and Water Company. (Contemporary newspapers simply may have shortened the name in three different ways.) Or, there might have been three different companies. Since we were not certain, we kept them separate, assigning them three different key codes -- M-2056, M-2060, and M-2066, respectively. Only further research into the issues in question can reveal whether the prices represent one, two, or three different issues. Similar examples abound.

Electronic Format of the Data Set

We provide the data in two formats, Microsoft Excel and Microsoft Access. Instructions on how to use those programs, or to convert the files to other data formats, cannot be provided here. Users should consult their program documentation or software consultants for operational details.

The Access database is designed for researchers who wish to search for particular types of securities, e.g. state debt instruments, or years, e.g. 1818-1822, across all markets. Database assistant Namsuk Kim has included some basic queries to get researchers started, but researchers are not restricted to his efforts. They can create other queries, or port the data to another database, to meet their particular needs. Again, however, they must consult Access’s documentation and/or the IT departments of their home institutions for further instructions and technical guidance.

The Excel files are designed for researchers interested in a particular market or security. The index file, SEC-index.xls, contains the key codes and names of the securities, grouped by type: banks and trusts (key code F); insurance (I); transportation (T); manufacturing, miscellaneous, and utility (M); corporate debt (B); U.S. debt (US); state debt (S); and municipal debt (C). Each key code is composed of both the appropriate letter identifier (e.g. S for a state bond) and a dash followed by a unique four-digit number. For instance, F-0560 is the key code of the Chesapeake Bank, C-0506 of Jersey City 6 percent bonds.

The data are grouped in files by market. For example, all securities prices for the Richmond market are in the file Richmond.xls. Securities quotations are arranged by type within each file/market. Within each file, the worksheets always follow the same order: the equities of banks and trusts; insurance; transportation; manufacturing, miscellaneous, and utility; then all varieties of corporate debt; followed by U.S. debt; state debt; and finally municipal debt. The number of separate worksheets varies depending on the number of securities. One of the smaller markets, New Orleans, has all of its securities on one worksheet or “tab.” In the other files/markets, all securities of the same type are located on their own individual tab or worksheet. For instance, in the file Charleston.xls, all the companies on the first worksheet are “Banks,” all the issuers on the second worksheet are “Insurance” companies, etc., while all the debt instruments are on the fifth and final worksheet. In the New York file, by contrast, Banks are so numerous (Excel can only handle about 230 columns per worksheet) that they are broken into two worksheets, and each type of debt is included on at least one worksheet of its own.

Researchers should use the index file to identify the key codes of the security or securities in which they are interested. If an issuer is not included in the index file, no securities quotations for it were found. If a key code cannot be found in a specific Excel file, that particular security did not trade in that market. In fact, most of the securities in the data set traded in only one of the markets. Major exceptions to this rule are the Bank of the United States (first and second) and U.S. bonds, each of which traded in most (if not all) of the markets. Researchers interested in tracking the prices of those instruments over all markets will have to download each of the Excel files or use the Access database.

Researchers should search for an issuer of interest under alternate names. A researcher interested in the Albany Bank, for instance, should also look under Bank of Albany. Similarly, researchers should search for their issuer under multiple categories. This caveat especially holds true for railroads, manufacturers, and miscellaneous companies, quite a few of which engaged in multiple enterprises, e.g. transportation and iron manufacturing, and hence might be listed under either or both the Transportation and Manufacturing categories. Some industrial or utility companies had banking powers and hence might be listed under banks. Water utility the Manhattan Company (F-1390) is a prime example, but not the only case. Finally, the line between banks, trusts, and insurance companies, especially life insurance companies, was often a hazy one. So researchers should look for financial companies of interest under both Banks (key code F-****) and Insurance (key code I-****).

Researchers should also bear in mind that a company might have issued both equity (stocks) and debt (bonds) and even multiple forms of each. For example, the Baltimore and Ohio Railroad (B&O R.R.) sold two forms of common stock (T-0080 and T-0100) and at least 7 varieties of bonds (B-0040 to B-0100). Moreover, researchers should note that preferred stocks, hybrids between equities and bonds, might appear with other equity issues or in the corporate bond list. So researchers should also check the Corporate Debt list for bonds of interest.

Data Limitations

Project members took pains to ensure that the data represented in the Excel files and Access database are as accurate as possible. The data was closely checked for transcription errors. The quotations, particularly fractions, were at times illegible. Where a price clearly indicated a fraction, but the numerator or denominator could not be ascertained, we assumed a fraction of ½ (transcribed as .5). Prices that were completely illegible were ignored.

Due to the inherent limitations of the source material, however, the data set is not entirely “clean.” There are several important limitations to keep in mind. The data set contains:

a)  gaps where no quotations are available. This occurred, especially early in the period under study or in the smaller markets, where no newspapers carried securities prices. Gaps also appear where some issues/dates of newspapers known to carry securities prices are no longer extant.

b)  minor inconsistencies in dates both intra- and inter-market. For instance, the data set might have prices of securities trading in Boston on 1 March of a given year, New York on 2 March, Philadelphia on 7 March, Baltimore on 24 February, and no quotations from Charleston. Moreover, the precise nature of the dates was usually not clear, so undoubtedly inter-market inconsistencies exist as well. Specifically, dates sometimes represent the date of the quotation and sometimes the date of the newspaper.

c)  both transaction data from actual sales and broker-quoted data that may be an average of recent transactions or even the mere estimation of a broker.

d)  bid-ask prices at times but only one price at others. The single price quotations may be bids, may be asks, may be midpoints, or may be actual transaction prices. Moreover, the actual transaction prices may have been brokered prices or auction prices. Where dealers were involved, generally the bid and ask prices were published.

e)  quotations of equities that are sometimes ambiguous as they may refer to price ($) or percentage of par. Thus, a quotation of 50 for a particular stock may mean $50 or 50 percent of par. In some instances, we have changed percentage quotations to dollars, rendering, say, a quotation for Bank of North America (F-0250) stock of 110 as 440 because that bank’s par value we know was $400 for much of its early history. Similarly, a quotation of 110 for the Insurance Company of North America (I-0810) was rendered as 11 because the par value of that company’s stock was $10. In many cases, the par value of stock was $100, making the distinction between dollar quotations and percent of par quotations irrelevant. Similarly, all bonds had face values of $100 for quotation purposes, so a quotation of, say, 95, unequivocally means $95 for a bond with a face of $100. The par value of equities, however, was not always ascertained and sometimes changed. In those cases, we have simply transcribed the reported figure. As a general rule, equities in New York were quoted as percentage of par throughout the period of study. No such hard and fast rules apply to the other markets, with quotations changing from dollars to percentage of par depending on the issue and the reporter. Researchers should note, therefore, that abrupt shifts in the quoted prices may be due to actual changes in the stock’s market price, or may simply reflect changes in the stock’s par value or changes in the basis of quotation. For example, an immediate drop from 100 to 30 may indicate a company in serious distress or simply a $30 par stock being quoted in dollars instead of percent par. The drop might also indicate a devaluation of the par value of the stock. The securities price lists from which the data set was created are almost always entirely mute on the causes of such shifts; only exhaustive firm-specific research can reveal its cause. Finally, in the London market, most equities prices are in pounds sterling but some, most notably the Bank of the United States (F-0310), were given in U.S. dollars.