One of the goals of the US Stocks Database and the UK Stocks Database is to eliminate the survivorship bias and the exchange bias that all other stock databases suffer from. Survivorship bias occurs when only stocks currently listed are included in a database and delisted stocks are excluded. Exchange bias occurs when historical data for stocks that were previously listed on another exchange or traded over-the-counter are excluded, providing an incomplete picture of the history of the company’s stock price.
This can also create a sector bias. If a sector tends to have stocks listed on a particular exchange or over-the-counter and those stocks are excluded, whole sectors may be excluded from the database. For example, most financial stocks listed over-the-counter until the 1970s. Any U.S. database which excludes over-the-counter stocks before the 1970s excludes banks and insurance companies. Although individual banks and insurance companies were small, at any given point in time over 1000 banks and insurance companies listed over-the-counter in the United States.
For US Stocks, some databases have extensive histories for one exchange, such as the New York Stock Exchange, but ignore price data that occurred when the stock traded on the AMEX, a regional stock exchange, or over-the-counter. For a stock, such as Dr. Pepper or Xerox, for example, this may mean that decades of information about the company and its stock are not available.
Exchange Bias in The Course of the Exchange
In the United Kingdom, the exchange bias occurs because databases focus on trading on the London Stock Exchange, but ignore trading in Scotland, Ireland or on provincial exchanges. This bias means that whole sectors of the British economy, such as canal stocks, may be excluded or have their role minimized.
The London Stock Exchange was founded in 1801 and until then, no organized exchange existed in England. The Course of the Exchange recorded the prices of stocks and bonds traded in London in the 1700s, but included primarily the English Funds. Although canal stocks went through a bubble in the 1790s, they were completely ignored by The Course of the Exchange in the 1790s, leaving us with no record of the stocks’ behavior.
Although The Course of the Exchange was the official source for trades on the London Stock Exchange, it was surprisingly selective. Each issue was two pages and focused on stocks trading in London. The Irish Funds were almost completely excluded from it. Scottish stocks received a special section in 1827, but this was removed in 1844 to make way for railroad stocks. Canal stocks were the largest sector in The Course of the Exchange in 1825 with over 70 canals listed, but had shrunk to 4 by 1864 when The Investor’s Monthly Manual listed almost 20. British mining stocks didn’t receive a special section until1857 even though local newspapers had tracked their behavior since the 1830s.
Provincial shares were generally ignored unless the company grew large enough to be traded in London. Relying exclusively on The Course of the Exchange or The Times of London meant that many stocks traded in Britain, but not in London were often ignored. The United Kingdom suffered from exchange bias just as the United States suffered from exchange bias.
The primary reason I can think of for this bias was the decision to keep The Course of the Exchange to one sheet, or two pages, in the 1840s. When the number of railroads, preferred stocks and bonds increased dramatically in the 1840s, the editors had to either expand to four pages, or reduce the number of non-railroad stocks that were covered. Unfortunately, they chose the latter.
It wasn’t until 1864 when The Economist began publishing The Investor’s Monthly Manual that a single source covered all the shares that traded in London, the provinces, Scotland and Ireland. Before 1864, the only way to get a comprehensive overview of shares that traded in England was to go to provincial newspapers which provided data on the prices of locally traded stocks. A newspaper such as The Times of London was quite comprehensive in its coverage of stocks and bonds. Separate sections were provided for Railroads (British, Colonial and Foreign), Government Securities (English Funds, Colonial Securities, American Government Securities and Rails), Mines (British and Foreign), Joint Stock Banks, Docks and Miscellaneous Shares. Over 250 companies were listed in The Times or in The Course of the Exchange in 1864, but a closer analysis reveals how selective the choice of companies was.
If you look at the Joint-Stock Banks section of either resource, you would think that a list of over 50 banks would cover the full gamut of banking throughout the British Isles, but you would be wrong. The list of joint-stock banks in The Times includes only London-based English banks and foreign banks that were listed on the London Stock Exchange. Very few provincial banks, and none from Ireland or Scotland, are listed.
The Canal Bubble
Similarly, one of the most famous sectors of the British economy, canals, is entirely absent from The Times. The reason? Most canals were located in the Midlands of England and were relatively small in size. In the 1700s, the canals were traded locally and not in London where the English Funds dominated trading. Some canals had a very limited number of shares outstanding and were not liquid enough to trade in London. The Loughborough Canal, for example, only had 70 shares outstanding. Shareholders also rarely sold their canal shares. For the Leeds and Liverpool canal, built in 1789, 60% of the original shareholders still held their shares in 1795 and 46% in 1800. Some shares only became available because the owner died.
Since there were no provincial stock exchanges, and the shares were rarely traded in London, how were canal shares bought and sold in the 1790s? As a review of provincial newspapers reveals, in the1790s, a local broker or shareholder placed an advertisement in the local newspaper announcing that he had shares available for purchase or that an auction of shares would be held on the morrow. Many of the ads provided information on recent dividends paid by the canal to entice bidders to purchase them, but unfortunately, the auctioneer never followed up the next day with an ad detailing what price the shares had sold for.
Unfortunately, we may never be able to measure in detail the size of the canal bubble because there is almost no record of what price canal shares traded at in the 1790s. The table below shows the prices of shares of several canals from a Birmingham auction in 1792 and their price in 1812. Most of the shares were issued at £100, and by 1792 most were trading at a strong premium. However, what path the shares took between their issuance at £100, 1792 and 1812, we will probably never know.
|Melton & Mowbray||155||108.5|
|Trent & Mersey||450||1200|
Between 1791 and 1794 parliament allowed 84 canals to be built creating the canal mania of those years. Further booms in new companies occurred between 1807 and 1808 and between 1824 and 1825. The boom of 1824-1825 involved not only mining companies in South America, but local coal and gas companies.
We owe our knowledge of share prices in the early 1800s to enterprising brokers throughout England who took out advertisements in provincial newspapers and leading magazines to advertise trading prices of shares not part of the English Funds. Were it not for Mr. Scott at 28, New Bridge Street in London, who started providing monthly data on shares to The Gentlemen’s Magazine in 1806, we wouldn’t have the knowledge of share prices that we have. Mr. Scott had advertised in London papers, announcing auctions of shares at Garraway’s Coffee House in the 1790s and 1800s, but only in 1806 did he begin publishing share price. The European Magazine, The Farmer’s Magazine, The British Magazine, The London Magazine, The Circular to Bankers, The Economist and others provided share price data for individual companies covering all sectors of the British economy.
The Irish Funds
In addition to the English Funds, there was also the Irish Funds. Ireland was in personal union with England in the 1700s, recognizing the king as sovereign, but maintaining a large degree of autonomy over local matters. After the French Invasion of Ireland in 1798 and the Irish rebellion of 1798, Britain determined to make Ireland a political part of Great Britain.
In 1800, the Acts of Union created the United Kingdom of Great Britain and Ireland. Because the Acts of Union didn’t occur until 1800, Ireland issued its own debt, established the Bank of Ireland in 1783, the Grand Canal in 1756 and the Royal Canal in 1789. Consequently, we have a better picture the price behavior of Irish government debt, Bank of Ireland stock and Grand Canal shares than we do of shares issued in Scotland or in the provinces. Consequently, the Grand Canal and the Royal Canal are the only canal companies for which we have price data during the canal bubble. As the graph below shows, Grand Canal stock doubled in price in 1792, but declined in price thereafter.
Contrast this with the data we have on Scottish stocks. The Bank of Scotland was established in 1695, the Royal Bank of Scotland in 1727, and the British Linen Co. in 1746 but no data on their share prices exist until the 1820s. Scotland never issued debt separate from England so no Scottish Funds exists. Data on Scottish share prices before the 1820s is simply unavailable.
The Rise of the Provincial Exchanges
In 1824, The Scotsman began publishing Prices of Edinburgh Stocks in each weekly issue. What is interesting is that over 40 shares are listed each week showing that there already was an ongoing market for these shares, but until then the prices for the shares weren’t published. The Scotsman published this list in the midst of the stock bubble of 1825, but once the stock bubble was over in 1826, The Scotsman stopped publishing the list of stocks. The Course of the Exchange began listing Scottish stocks in 1827.
The timing and popularity of local share lists in provincial newspapers gives an idea of how interest in shares changed over time. When interest was high, local share lists appeared. When interest fell, the share lists disappeared.
The Scotsman was the only provincial newspaper to publish a local share list in the 1820s. Other provincial newspapers published stock prices, but only of The English Funds, Foreign Funds and foreign shares that traded on the London Stock Exchange.
It wasn’t until 1830 that local share price lists began to appear locally as interest in railroads and other shares picked up. The Liverpool Mercury first published local share prices on March 12, 1830, the Leicester Journal published a list in 1833, the Bolton Chronicle in 1835, The Scotsman in 1836, the Coventry Standard in 1836 and the Cornwall Royal Gazette in 1837. Most of these share lists were courtesy of local brokers who were trying to drum up business for themselves.
The introduction of local share lists in the newspapers coincided with the growth in provincial exchanges. The first exchanges established outside of London were in Liverpool and Manchester which were formed in 1836. Scotland got its first exchanges in Glasgow and Edinburgh in 1844 when Sheffield and Leeds also established exchanges. The railroad bubble reached its apex in 1845 when a dozen exchanges were founded. The exchanges in Bristol, Newcastle and Aberdeen continued to exist after the railroad bubble burst, but the exchanges in Hull, York Huddersfield, Nottingham, Halifax, Bradford and Leicester closed when interest in shares waned.
The amount of information the newspapers provided varied from one source to the other. Ideally, the newspaper provided information on the number of shares outstanding, the par value of shares, the paid value, the market price and the dividends paid in the last year. Unfortunately, most of the sources did not provide all that information.
When the railway bubble reached its apex in 1845, many newspapers only provided data on railroads shares, dropping all reference to canals, banks, insurance companies and other local companies. After the railroad bubble collapsed in 1845, many newspapers stopped publishing share information altogether. As interest in stocks grew again in the 1850s, provincial papers once again began to publish share data. By the time The Investors Monthly Manual was published in 1864, most local shares were covered by one paper or another. The IMM simply made it easy to get all this information in one place.
Eliminating the Exchange Bias
The United States and the United Kingdom provide a good contrast with one another. In the United States, regional exchanges flourished until the 1900s. Although New York was the center for share trading in the United States, just as London was in the United Kingdom, including only the shares traded in New York or London provides an incomplete picture of the share market in either country.
In particular, stocks traded in Ireland and Scotland were often ignored in London, and whole sectors, such as the canals, were minimized or ignored. Information on shares in Dock stocks is readily available because the docks were located in London, while canal stocks were often ignored because they lay in the provinces.
Were it not for the willingness of stock brokers in the Midlands to provide share lists in provincial newspapers to drum up business, it would be difficult to fully trace the gradual decline of canal stocks as railroads rose at their expense. Similarly, the history of the banking sector would be limited to London banks and would exclude many provincial banks were local newspapers ignored, and the behavior of Welsh mines in the 1840s and 1850s would be lost were it not for the invaluable information provided by newspapers in Exeter and Cornwall.
Global Financial Data has made a point of including all shares traded in both the financial capitals of New York and London as well as shares traded on regional and provincial exchanges in the US and in the UK. This enables GFD to avoid the exchange bias, the sector bias and the survivorship bias which focusing on a single exchange creates. Without this, any analysis of the behavior of shares in the past is not only incomplete, but inaccurate.