Market Integration during the South Sea Bubble:
Evidence from the Royal African Company

Ann M. Carlos*, Jonathan Hill, and Nathalie Moyen, University of Colorado
*Department of Economics
University of Colorado
Boulder CO 80309
Telephone: 303-492-5168 Fax: 303-492-8622

October 1998: Very Preliminary Summary


Much has recently been written on the question of speculative excesses in the American stock market. Soaring share prices, merger mania and rising property prices seem to indicate the presence of a bubble economy. More worrying is the question of when booms turn to busts. While many argue that the American economy had entered a new era, history suggests that what goes up can often come down, with market pundits citing as examples the Stock Market crash of 1929 and the South Sea Bubble of 1720 as examples of market bubbles.

In his excellent book, The Rise of Financial Capitalism, Larry Neal has documented the path of the South Sea Bubble as arising from attempts to accommodate the government's financial needs. War had created a costly debt structure for the government and an illiquid asset structure for the holders of that debt. To lower the cost to the government of this debt and to provide debt holders with liquid and easily transferable assets, the system needed a new financial instrument. Experimentation to arrive at that instrument led not just to the Bubble but resulted in the first perpetual annuity or the consol as a government debt instrument of choice(Neal, p. 90). Neal clearly documents how the Bubble affected the price of South Sea Company, East India Company and Bank of England stock. These were, however, the monied companies of the period which provided the government with the capital it needed to prosecute its wars. As such, Neal provides us with a description of one part, albeit the most very important part, of the capital market. But there were other companies operating in this market.

In the decades following the Glorious Revolution there was a dramatic increase in the number of joint-stock companies floated on the London market. What is much less clear is how the Bubble affected these firms. Using a unique data source, this paper examines the impact of the Bubble on the course of Royal African Company stock and share prices. In doing so, the paper asks two questions. The first is primarily descriptive. What happened to the price and volume of transactions in Royal African Company stock during the course of 1720? The second question is more analytical. It asks whether the bubble apparent in the share prices of Royal African Company stock from April to October of 1720 was, in fact, a rational bubble or whether it does describe some speculative excesses.

Royal African Company - Financial Structure

The Royal African Company received in 1672 a royal charter giving it a legal monopoly of the British slave trade between Africa and the West Indies. While the receiving of the charter has been much discussed in the literature (Davies), what has received much less attention is the fact that a charter was the only legal way to become a limited liability, joint stock company. At the time the Royal African Company received its charter, it was the second largest of the joint-stock companies after the East India Company. Profitable through the 1680s, the wars of the 1690s led to serious losses, as a result of which the company began to license traders resulting in a dramatic decline in market share. (Carlos and Kruse)

By 1712, the company was essentially bankrupt. The price of the company shares was 2 on a par of 100. Rather than winding up the business, the company underwent a major financial reconstruction, dramatically writing down the existing capital stock. Share prices rebounded for a number of years, only to decline to 16 by 1719. The Royal African Company was not part of the political/financial complex of monied companies. It was a trading company with a checkered history which makes it an interesting case study.

The data set used in this paper comes from the company stock ledgers which provides us with the names of buyers and sellers of company stock for 1720 and a direct measure of the volume of transactions. In addition, we use Neal's compilation of share prices from Castaing's Course of the Exchange over the same period. Thus, although we do not know exactly at what price each share was traded, we can look at the pattern of share prices and volumes.

What makes this examination much more interesting is the nature of the share data with which we are working. In April 1720,the Royal African company offered a large new stock issue to the value of 1,569,000 pounds. In all previous cases, the company had sold any new stock issue itself. In this case, the company used an used an underwriter to sell this stock. In an agreement dated 7 April 1720, Joseph Taylor paid 75,696 pounds for the whole issue. Thus Joseph Taylor bought the issue at 5% of its par value. The company did, of course, receive 75,000 pounds as a new cash infusion. Because all stock were sold in units of 100, there were now 15,690 new shares available.

In any capital market, the issue of new shares will depress the prices of the existing shares if there has been no change in the underlying value of the company. The existing shareholders of the company would therefore have to be compensated in order to have them to agree to the new issue. The agreement drawn up did, in fact, attempt to compensate the existing shareholder. They were promised a dividend of 10% one year hence in April 1721. Because of this dividend, the company had to record the stock transfers of the "engrafted" stock in a separate transfer book until the middle of 1721. The company records, therefore, provide us with insight into the ways in which the new shares entered the market and the transactions pattern of both old and new shares. Castaing's Course of the Exchange also listed separate share prices for these two all but identical shares.

Financial Information and Bubbles

The structure of the company records and of the Financial press allows us some very interesting insights in the operation of the capital market of the period. In this section of the paper, we look at a number of different issues which can only be summarized briefly here. We examine the extent to which the market correctly valued the old and new stock. Was the path of prices over the course of the bubble different for these shares? Did the prices of these shares converge over the second half of 1720 and in the aftermath of the bubble?

When we look at prices, we find that the price of the existing stock was 24 in January, 1720. By the beginning of March it had risen to 50, by April 60, where it stayed until the middle of May. By the beginning of June the share price of the old stock was 140. This was when Castaing begins to list the price of the engrafted stock. The first listing at the end of May was 95 and this rose to135 June 1. Both prices fluctuate through June, July and August, hitting levels of 170 and 140 respectively. Then in September, they begin to fall ending the year at 45 and 35. The correlation between the movement of both stock is 0.95 as one would expect. But we also examine whether the price differential between these two shares accurately captures the 10% dividend promised one year hence.

Mirroring the behavior of prices, we also examine the behavior of volume. Although we do not know at what price each individual transaction took place, we have a complete record of each transaction in the old and new stock. Thus we look at the degree of market integration in terms of buyers and sellers of the stocks. Preliminary examination of the share transfer data suggests that the volume of transactions in the new shares was significantly higher than in the old and had a somewhat different path over the course of the Bubble.

The total number of transactions in the old stock of the Royal African Company was roughly 1100 transactions. This is purely a measure of the number of interfaces between a seller and a buyer from December 1719 to December 1720. It is not a measure of the value of the stock transferred. The largest number of transfers took place in March 1720 with 297. In June, July and August, there were 186, 128 and 71 transactions. There was another 41 in September and only two more to the end of the year.

The pattern for the engrafted stock is quite different. There was a total of 4336 transactions in the period from May to December 1720. There were 872,834 and 514 transactions in the three months of June, July and August. Thus activity in the engrafted stock was very high. What is also interesting is that the majority of individuals buying or selling were selling very small amounts of stock and involved in only one to three transactions. The vast bulk of the transactions involved no more than 300 of stock or three shares. Eugene White (1990) has argued that some of the instability injected into the share market in the 1920s was the result of new and inexperienced buyers. These data suggest that for the engrafted stock, there may have been 'new' investors in the market. There were, relative to the other transfer books examined, buyers who were not able to sign their names (both male and female).

Using the names of buyers and sellers allows us to look at the social structure of the market and to examine the dispersal process of the new stock from Joseph Taylor to the larger market. Joseph disposed of the stock issue in very large block to what is clearly the aristocracy of the period. But from there it was very quickly dispersed into the mercantile community. As found in other work (Carlos, Key and Dupree), goldsmiths were the most active traders in the engrafted stock, acting once again as brokers.

The paper also examines the extent to which there was a rational bubble in Royal African company share prices. Using the tools developed by Donaldson and Kamstra, we try to ascertain whether this asset price can be defined as a rational bubble using an examination of the different pricing regimes over 1720. Although we think about the Bubble as a single period, for the market and for investors, there were a number of different regimes. This is particularly the case for the engrafted stock. Buyers of that stock did not have to pay the full price of the shares up front, but rather contracted to pay in a series of installments. Each call for payment, therefore, represents a different price at each of those points in time.

In summary, evidence from this case study, in conjunction with the work on the monied companies, gives us an interesting insight into the ways in which the market responded to the South Sea Bubble. Understanding how the market operated during this period may give us some insight into whether the behavior of stock market bubbles are similar or different in emerging markets relative to more developed markets.


Carlos, Ann M. and Jamie Brown Kruse (1996), "The Decline of the Royal African Company: Fringe Firms and the Role of the Charter," Economic History Review: Vol. 49, pp. 291-313.

Carlos, Ann M, Jennifer Key and Jill Dupree (1998), "Learning and the Creation of Stock Market Institutions: Evidence from the Royal African and Hudson's Bay Companies," Journal of Economic History: Vol. 58, pp. 318-344.

Carswell, John (1960), The South Sea Bubble, Stanford: Stanford University Press.

Davies, K.G. (1957), The Royal African Company, London: Longmans.

Donaldson, R. Glen and Mark Kamstra (1996), "A New Dividend Forecasting Procedure that Rejects Bubbles in Asset Prices: The Case of 1929 Stock Crash," The Review of Financial Studies, Vol. 9, no. 2, pp. 333-383.

Galenson, David W. (1986), Traders, Planters and Slaves: Market Behaviour in Early English America, Cambridge: Cambridge University Press.

Neal, Larry (1990), The Rise of Financial Capitalism: International Capital Markets in the Age of Reason, Cambridge: Cambridge University Press.

Scott, William Robert (1951), The Constitution and Finance of English, Scottish and Irish Joint-Stock Companies to 1720, three volumes, New York: Peter Smith.

White, Eugene (1990), "The Stock Market Boom and Crash of 1929 Revisited," Journal of Economic Perspectives, 4, pp. 67-84.