Methods

I conducted ethnographic research over four-month periods in Fiji (August–November 2000) and Ghana (December 2000–March 2001), collecting a variety of data on the structure and operation of the markets. During these periods, I engaged in participant-observation at the stock exchanges, brokerage firms, and regulatory agencies every business day, observing all trading sessions and front- and back-office tasks during nonmarket hours. I also completed thirty in-depth, semi-structured interviews with stock exchange staff, regulatory officials, stockbrokers, and other market participants.Footnote ¹ Interviews ranged from thirty minutes to three hours. I drew upon specific instances from my field research during the interviews to formulate questions and topics of conversation. Finally, I analyzed documents at the regulatory agencies, stock exchanges, and brokerage firms that articulate or explain the regulatory process.

Underlying the analysis in this article is a textual comparison of the legislation in both countries that created securities regulatory bodies, with particular focus on regulatory functions and powers. Analytically, I also draw upon field notes of the participant-observation research and the written materials to understand how individuals engage in their day-to-day work. I use the interviews with regulatory officials in Fiji and Ghana to gain insight as to how these officials define their tasks, gather information, and implement policies. Finally, the article draws upon the array of research I undertook during the market sessions at the exchanges and with brokerage firms to examine the effects of the regulatory actions on market participants' orientation and behavior.

Background of Research Sites

Both Fiji and Ghana are former British colonies, gaining independence in 1970 and 1957, respectively. In the 1980s and 1990s, the governments of Fiji and Ghana decided to emphasize capital market development as a component of economic policy, reflecting the increasing prevalence of these markets worldwide (Reference HelleinerHelleiner 1994). As part of this emphasis, Fiji and Ghana each created a regulatory agency responsible for promoting and regulating the development and operation of the capital market through acts of Parliament. In 1993, Ghana passed the Securities Industry Law, Provisional National Defence Council (PNDCL) Act 333 (SIL), which created the Securities Regulatory Commission (SRC).Footnote ² In 1996, Fiji passed the Capital Markets Development Authority Act, number 13 (CMD Act), creating the Capital Markets Development Authority (CMDA). The passage of each piece of legislation occurred in proximity to substantial changes in the operation of the capital markets in each country, particularly the establishment of more formal stock exchanges.

The Ghana Stock Exchange (GSE) was created in 1989. During the mid-1990s, Ghana was undergoing structural adjustment under arrangements with the International Monetary Fund. As part of the adjustment program, Ghana undertook to liberalize the foreign exchange control regime and to privatize some large state-owned enterprises. The organization of the GSE, including most basic trading, listing, and membership regulations with the exception of trading by nonresident foreigners, did not change during this period (other than by growth). By the end of the 1990s, the GSE wished to change to an electronic trading system, but project funding through a World Bank grant was cancelled. The GSE still changed the trading rules to what they would have been if the electronic system had been implemented but continued to trade manually.

Formally, Fiji's Suva Stock Exchange (SSE)Footnote ³ had existed since 1979, but it initially operated only as a trading post. Investors who wished to purchase or sell shares would send their request to the exchange, where one of the two staff members would enter the order into a ledger, matching orders once each week. As part of the effort to spur capital market development in Fiji, it was decided that the SSE should become a more active market. Thus, in 1996, the SSE began to operate a call market, in which brokers would bring client orders to the exchange in an on-floor auction process.

Both Fiji's SSE and the GSE are small by international standards. At the end of 2000, the SSE listed ten companies with a total market value of US$126 million, while the GSE had twenty-two listed companies with a total market value of US$518 million.Footnote ⁴ During 2000, both exchanges had small total trading volumes, with SSE total turnover at US$3.7 million and GSE turnover at US$7.2 million. The number of brokerage firms roughly corresponds to the size of each of the markets. In Fiji during 2000, there were four stock brokerage firms, each employing two or three registered brokers; there were 12 brokerage firms in Ghana, employing between one and four registered brokers. Finally, the number of staff at the regulatory agencies is also proportional to the market sizes, with five staff members in Fiji and nineteen in Ghana at year-end 2000. Total regulatory body expenditures in 2000 were US$172,000 in Fiji and US$344,000 in Ghana, or 4.7% and 4.8% of total market trading value, respectively.

In addition to comparable sizes, the two stock markets are similar in relation to their domestic economies and global financial markets. Both exchanges list a substantial proportion of domestic subsidiary companies of transnational corporations (one-third to two-thirds), with the balance being domestic companies. While foreign investors own a larger aggregate percentage of shares on the GSE, most of these investors are not active in buying and selling shares, and all rely on Ghanaian brokerages. While brokers and the exchanges in each country follow events in overseas capital markets, neither stock market moves in reaction to these overseas markets (Reference LarsonLarson 2003). Both exchanges trade three times a week, during mornings only. Neither is central to the provision of capital to the domestic market, as bank finance dominates for most enterprises in each country, and neither exchange is the source of substantial trading in government, quasi-government, or corporate debt. Each country also has a large number of companies with equity finance that are not listed on the respective exchange.

Market Surveillance Techniques

The fundamental difference in regulatory approach between the CMDA and the SRC can be seen through an analysis of practices of market surveillance. While both agencies have staff positions whose titles and job descriptions include “market surveillance,” the tasks involved are substantially different.

For Fiji's CMDA, the job of market surveillance is process-oriented, with primary attention focused on the details of specific transactions. The market surveillance officer attends all stock exchange trading sessions and compiles a market surveillance form. If any trading is “irregular” or “does not make sense,” the officer notes the situation. The market surveillance officer draws on additional CMDA staff members to check the trading practice against the rules and regulations in order to “look at the situation and determine if there is a case to pursue.” The tentative conclusion reached by the CMDA appears in a written market surveillance report, which may require further investigation. An official explained, “If there is something that happens [such] that we need to review certain documents … we conduct a special inspection [to] review the things that we see that we think are related to a certain event.” After reviewing this information, including the response of the party that is suspected of breaching the rules, the staff of the CMDA reaches a conclusion. If a violation occurred, it forwards this to the members (directors) of the CMDA, who determine the appropriate regulatory action.

The officer's surveillance of the market relies on having a “feeling” for the market. The officer explained that during the year,

This example illustrates several aspects of CMDA's market surveillance. First, the surveillance is of the process of trading. The surveillance officer pays attention to how and when orders are placed (and matched), rather than merely paying attention to the outcomes of orders (trades and prices). Second, the attention to process highlights the importance of having market knowledge that stretches across time. It is important to know what orders were placed or removed in previous sessions. Finally, the surveillance officer has developed a tacit sense of what seems correct or incorrect. Thus, something that “looks funny” triggers a deeper investigation.

In contrast, market surveillance for Ghana's SRC consists of annual inspections of each brokerage firm and focuses on brokers' adherence to standards of business conduct and financial adequacy. The SRC and GSE jointly select issues to focus on during the inspection process and use a set of topics and questions, along with a sample of records, to assess the brokerages' practices. The practices that are assessed, in the words of the field officer, are “the issues fundamental to the protection of clients, because our main jobs are protecting the client and developing the market.” One member of the market surveillance department cooperates with an official of the GSE to examine the accounts and records of the firm. After a field visit, the market surveillance officer writes a full report that is used internally in the SRC to draft a letter to the management of the brokerage. The management letters often raise the same issue with a number of firms, suggesting corrective actions to the brokerage firms. For example, in all of the letters to management in 2000, the SRC noted that the brokerage firms needed to appoint compliance officers.

The inspections process is quite similar to an audit. The SRC market surveillance field officer explained,

As in Fiji, inspectors in Ghana must rely on “instinct”:

Pursuing “hunches” may uncover an item that raises suspicion, leading to more detailed investigation, as the field officer recounted:

The field officer explained that they began “investigating this with a lot of other brokerages” as part of the annual inspections. This explanation encapsulates the surveillance process used by the SRC. The records of the firms are investigated at a summary level. If something seems out of the ordinary, it is investigated further, often through questions and answers. If an offense is discovered, it becomes a part of the inspection routine at other brokerages.

To bring the comparison with Fiji's CMDA into sharper relief, the surveillance by Ghana's SRC is not of the process of trading, but of the front- and back-office practices of brokerage firms. Second, the SRC surveillance in Ghana is based on an official applying an observation to additional locations, in essence developing standards for all market participants. In contrast, the CMDA's market surveillance requires an official to observe across time in order to understand the process.Footnote ¹¹ Finally, surveillance in both markets requires tacit knowledge to trigger more detailed investigation, yet this knowledge is of a different nature in each market. The trigger for the CMDA is the behavioral practice of a market participant. For the SRC, it is the accounting of firm behavior.

Examination of how each agency completes the task the other calls “market surveillance” further highlights the different approaches to regulation. At the SRC, the Corporate Finance and Investment department is responsible for tracking trading on the stock exchange. The department uses the Official List—a daily release from the stock exchange summarizing volumes traded, price changes, and high bid and low offer prices—to insure proper reporting of this information and to prepare monthly reports about aggregate trading activity and market trends. In contrast to the market surveillance official at the CMDA, the SRC official who tracks the market relies upon information supplied by third parties, rather than firsthand knowledge, and is concerned with the outcomes, rather than the process, of trading.

The CMDA conducts regular and special inspections of brokers. These begin with an interview that includes “specific questions on rule compliance.” While not drawing too sharp of a distinction, since the SRC uses some of the same technique of posing questions to brokerage firms and since the CMDA examines financial records, there is some contrast between the choice of the entry point into the inspection of the brokerage firm. A CMDA official's explanation of inspections illustrates the distinction between how and which records are examined:

Here, a sample of records is examined as part of the process of bringing the client orders to market. The CMDA official is concerned with the “flow” of orders and records. While the CMDA is also concerned with the ongoing ability of the firm to conduct business, the use of order entry inspection as a “building block” highlights the unique importance of specific details of order processing, in contrast to the SRC.

It is important to note that despite the fundamental differences in the regulatory approach to market surveillance and broker inspection,Footnote ¹² some aspects of regulatory activities in Ghana and Fiji are the same. Officials in both agencies rely on inference based on individual experience to select cases for further investigation. The CMDA official uses the knowledge of market dynamics, while the SRC official relies on the “auditor's instinct” to sense that something is wrong. Both regulators stretch their observations beyond their present time-space location: The CMDA looks at patterns of trading over time, while the SRC uses findings from one firm to set the general agenda for inspection for all firms. This extension of observation is a key manner in which the regulators build tacit knowledge, gaining the experience of knowing how the regulated parties behave. Finally, both the CMDA and SRC operate with the same ends in mind: client protection and market development.

Yet their techniques of inspection and regulation differ. The CMDA, in both market surveillance and regulatory inspections, examines the manners in which orders to trade securities are processed. In this regard, regulation is constructed as a component of the transactional process (although not wholly reduced to it). The SRC, in contrast, constructs market surveillance as an auditing function. The entry point is not through specific transactions, but an accounting view of the brokerage firm. This applies equally to trading on the stock exchange, as the SRC is concerned about the broad outcomes of trading (volumes, price movements) rather than the mechanics that were involved in producing the outcomes. From a regulatory perspective, the variation in approaches demonstrates the trade-off between the values promoted by a particular technique and the risks associated with such an approach. A closely involved regulator gains responsiveness and an inside perspective but risks capture; while regulation from a distance provides a clearer delimitation of regulator independence, it risks formalism.Footnote ¹³

Trading Outside the Rules

To gain more insight into how CMDA and SRC regulatory practices differ, it is instructive to compare how two similar instances were handled. Violation of stock exchange trading rules regarding the matching of orders occurred in both locations. While rare, these instances are analytically useful to demonstrate how participants deal with “trouble” cases and how the practices of regulation play out in each location.

In Fiji, the process of matching particular orders became a matter of contention during one trading session. The issue of what constituted a “matching order” arose when a broker placed an order to buy shares at a price higher than the price of an existing selling order. During trading of Talanoa Corporation,Footnote ¹⁴ Broker A placed an order to sell at $3.01 per share that remained unfulfilled on the market. Subsequently, Broker B listed orders to sell Talanoa shares at $3.03. Broker B then took out a calculator to determine the total quantity of shares on the orders just listed at $3.03. Broker B then placed an order to buy this quantity of shares at $3.03. The SSE executed all the orders at $3.03, matching Broker B's buying order with Broker B's selling orders and bypassing Broker A's selling order at $3.01. Broker A protested that the order at $3.01 had priority, since it was at a better price. An SSE staff member replied that there was no match between the orders to sell at $3.01 and buy at $3.03, since $3.01 does not equal $3.03. After this, Broker B again listed an order to sell (this time at $3.05) and an order to buy (at $3.05), bypassing another order of Broker A's to sell at $3.04. The SSE again matched Broker B's orders. The head of trading at Bula Brokerage (Broker A's firm) retrieved a copy of the SSE Business Rules and showed them to the SSE staff member responsible for making decisions in the matter. Despite this protest, the SSE stuck by the earlier decision. During this entire time, the CMDA market surveillance officer recorded all of the orders and trades, verifying them with a market official from the stock exchange.

Upon returning to the CMDA office, the market surveillance officer re-created the call market session trading by writing the orders and trades on a whiteboard in chronological order. The surveillance officer then discussed the trading with another CMDA staff member, testing various interpretations of what had happened. One member hypothesized, “Broker B made a mistake. The buy order is higher than the sell order. That does not provide the best execution. Also, Broker A should have been more aggressive [in protesting the SSE's decision].” After identifying the relevant trading rules, the staff members reworked the chronology by listing the orders in sequence, noting when and how trades were matched. Then they repeated the process a second time by scrutinizing each order in light of the trading rules. This reconstruction of the trading identified eleven separate steps of the trading process, with specific notation of each breach of the trading rules.

After this initial reconstruction was complete, a third staff member was brought into the conversation, with the discussion focusing in detail on the ordering of the breaches of the trading rules. The CMDA staff members asked questions of one another, attempting to pose challenges to the tentative conclusions about the situation.Footnote ¹⁵ The staff determined that the situation warranted further action in directing the stock exchange. The entire process began with the CMDA's market surveillance, which is based on the surveillance officer's knowledge of the basics of how the market should work and was enabled by the officer's daily recording of information. This information was used to examine the processing of each relevant transaction, in line with the construction of regulation as part of the market process.

The CMDA includes the activities of the stock exchange within its regulatory purview. It examines whether activities breach the rules and essentially re-creates the stock exchange as it should operate, relying on stock exchange rules and the actual “inputs” to the exchange—the flow of orders. CMDA officials locate, read, analyze, and then test the application of the rules relevant to the situation under consideration. When staff members think they have interpreted the rules and applied them to the situation correctly, the entire process is presented to and discussed with a supervisor, including as many possible responses as could be encountered. Finally, the CMDA decides whether a situation warrants their direct involvement, rather than waiting for a broker to issue a complaint letter. This proactive involvement in market operations sharply contrasts with the SRC activities in Ghana.

During my research, the GSE was in a period of transition, changing the process by which orders were matched. Under the new system, order-matching procedures were in price-time priority with instant execution. This contrasts sharply with the previous trading regime, in which any broker could match another's price on a bid or offer and, if the order was large enough, share proportionally in any trade. The shift in trading rules, thus, posed a problem for brokers: If one broker was the first to place an order to sell at a particular price, all of his or her shares would have priority over those of a broker who placed an order to sell at the same price moments later, meaning that trades would often go to the first broker, rather than be shared among a number of firms.

One brokerage firm noted a provision in the rules: Brokers could enter “market orders,” designating the order as “MKT” rather than putting a specific price on the order. Under the rules, a market order has priority over an order with a specific price. A broker could decide to use a trading tactic to list an order to buy or sell at MKT and immediately place the corresponding order to complete the transaction; this could circumvent the priority rules, getting both the buying and selling sides of the transaction. At one of the earliest “live” sessions of continuous auction trading, the firm found itself in a situation in which it wished to complete a transaction where it had both a buying and selling order for the shares of a company, but it could not fill both of the orders because another firm already had an offer to sell the shares at that price. The former firm decided to use this tactic, listing the selling order at MKT and less than a minute later listing the buying order, claiming the trade. At this point, a very heated argument broke out on the floor of the stock exchange, with the broker justifying the tactic by referring to the provisions for market orders in the trading rules. After studying the rulebook, a GSE official grudgingly upheld the trade. As was standard, no officials from the regulatory agency were present. Within one week, other brokers—including the broker whose order was bypassed in this first instance—started to learn and use the tactic.Footnote ¹⁶

As other brokers started to use the tactic, they lacked understanding of the trading rules concerning the tactic and made errors. As a result, some brokers claimed tradesFootnote ¹⁷ that they should not have: The broker from Cantonments Stock Brokers (CSB) placed an order to buy Volta River Bank (VRB) at market and an order to sell the same number of shares at 2300, claiming the transaction. When the broker placed the orders, however, there were existing orders on the board to buy VRB at 2300 and to sell VRB at 2305. Thus, if the CSB broker had first entered the order to buy at market, CSB should have bought the VRB at 2305 (from the member with the existing selling order at 2305). If the broker had entered the selling order first, CSB should have sold to the buyer with the bid at 2300. In spite of the fact that a number of other brokers were watching the transaction, no broker caught the error.

I witnessed other, similar incidents in the following weeks. Conversations with and among the brokers indicated that these were honest mistakes. Brokers thought they were using the rules correctly. In the instances that somebody notified a stock exchange official about a mistake, the typical pattern of response would be first to verify that the mistake had been made, followed by a check of the order tickets for the time that the orders were placed. If the GSE official concluded that a mistake had occurred, the official would typically go to the errant broker and explain the problem. The broker would usually apologize and claim that he or she did not realize the error. In the end, the GSE would issue a warning— usually a simple verbal warning. This action was designed to be preventive, as the transgression of the rules was seen primarily as a temporary, developmental issue. GSE officials generally saw a mistake as a flaw in the rules that would be corrected. At the same time, GSE officials viewed each broker as deserving one chance to make a mistake, since they were often trying to use these tactics to keep up with their rivals.

Unlike the CMDA, the SRC was notably absent in the instances of trading violations at the stock exchange. As no formal disciplinary action was taken in these instances, the SRC received no notification of specific violations of trading rules. The SRC was, however, indirectly involved in the attempts of the GSE to come up with a correction to the trading rules. This effort was part of a wider range of work by a committee charged to oversee the phased implementation of the new trading arrangements. In the committee meetings, the SRC representative was often a minor player. The representative would participate, asking questions and offering suggestions, but did not take an active leadership role, nor would the representative be consulted about how the SRC staff would look upon possible rule changes. It seemed, at times, that the representative was there as much to be informed about what the GSE would be doing as to engage in any sort of oversight or regulation.Footnote ¹⁸

This analysis touches on the more fundamental issue of the general relationship between the regulatory agency and the stock exchange. In Ghana, the relationship is more or less one of a partnership, with the SRC delegating authority to the exchange and jointly inspecting brokerage firms with the exchange. As an official at the SRC noted, “The stock exchange itself would have seen [any instance of market malpractice]. As a matter of course, they have to give us that report. We even ask them to investigate and report to us.” The SRC is detached from the daily operations of the stock exchange but maintains regulatory power over the exchange. As an SRC official explained,

Formally, these are the same powers held by the CMDA, but in practice the method of implementing regulation constructs the trading activities of the stock exchange in a different manner. The stock exchange's daily operations in Fiji are an object of regulation by the CMDA, in which primary importance is attached to having a process that follows the rules as prescribed. In contrast, the SRC leaves daily operations as an internal matter for the GSE, which, according to an SRC official, “makes sure trading is taking place in an orderly manner.” The SRC engages the GSE trading oversight as part of a broader partnership between exchange and regulatory agency. The GSE becomes a source of information (through reports initiated by the GSE or through requests for investigation and reports from the SRC) for regulation. In this regard, the stock exchange's daily operations in Ghana are not an object of regulation, but a source of SRC regulation.

How Brokers Dispute

As discussed above, the first time a broker at the GSE used an order designated as MKT to claim a trade, an argument erupted on the trading floor. The second time the tactic was used by the same broker, a few days after the stock exchange had upheld the practice, an even more heated argument ensued, with nearly every broker shouting and arguing against the broker who initiated the trade. Other than the broker initiating the MKT trade, no brokers referred to the trading rules in their arguments. Instead, brokers invoked themes of fairness, arguing that others “could not trade like that.” Fellow brokers were the sole audience of this vocal outburst. The following week, the broker who “lost” the trade the first time placed an order at market to get around another broker's order, stating, “They made a fool of me, but now I will use their trick.” Now there was no argument about the legitimacy of the tactic. When I asked a broker about this, I was told “it was accepted already.”

In Fiji, the most explicit trading dispute was the instance of matching orders for the Talanoa Corporation at different prices as discussed earlier. In this case, the head of trading of one of the firms whose order was bypassed obtained a copy of the trading rules and presented an argument to the stock exchange staff about the rules. The broker could have presented an argument either to or about the other brokerage firm, arguing that the firm had a duty to get its client the best price. Instead, the broker's argument centered on the exchange's application of its trading rules. When this was unsuccessful, the head of trading called the CMDA later that day to complain about the stock exchange's decision.

These two examples highlight important differences between broker orientations in the two exchanges. Both disputes concerned the interpretation of rules in cases of ambiguity: In Fiji, the debate concerned what constituted a “match,” while in Ghana, the status and processing of the new category of orders designated as MKT was contested. First, the initial audience and framing of the arguments are important. In Ghana, broker-centered arguments about fairness concentrated the dispute among the brokers despite the fact that the stock exchange had final authority. Brokers could have argued to a stock market official that the interpretation put forth by their competitor was based on a faulty understanding of the purpose of MKT orders under the trading rules.Footnote ¹⁹ In Fiji, the broker invoked specific trading rules rather than principles of fairness and transferred the dispute to the exchange. Both framing and audience shaped the potential outcomes, as the regulatory agency can levy formal sanctions, while brokers have no such authority. Second, despite the fact that a broker in each location cited the trading rules, the broker in Ghana used the rule as a defense when challenged after having planned the trades. In contrast, the broker in Fiji had no idea that any reference to the rule would be needed. The broker's use of the rule as an unplanned, immediate response in Fiji suggests that rule-consciousness is much more a part of the repertoire of brokers on the SSE.

Finally, there is the difference in the finality of the ruling of the stock exchange. The broker in Fiji appealed to the regulatory body after the ruling of the exchange, while no such appeal occurred in Ghana despite the fact that most brokers did not agree with the ruling. In Ghana, the brokers continued to dispute the practice the second time it was used even though the stock exchange had held that it was within the rules, suggesting an attempt to establish a floor norm against the tactic. Indeed, the practice only became recognized as accepted after a second broker used it, signaling its acceptance as a norm of trading, rather than simply an allowed practice under the rules. This second broker, by announcing “I will use their trick,” appealed to a norm of the fairness of turnabout. Instead of appealing to the rule or even simply trading the order, the broker's statement justified the behavior against the initial user of the MKT tactic as fair. These differences illustrate an underlying difference in legal consciousness as it relates to disputing: Brokers in Ghana conceived of and expressed disputes in terms of norms and fairness, while brokers in Fiji emphasized the formal principles of rules.

The difference in orientation toward the rules between Fiji and Ghana is further underscored by the strong adherence of brokers in Ghana to norms of fairness even when the trading rules would have allowed other practices, illustrated by an incident in Ghana in which the norms of fairness were explicitly invoked to prevent a broker from taking advantage of orders on the market. During one pre-opening session, a broker had listed an order to sell shares of a company as MKT. Since trading had not yet opened, this trade would happen at the price of the highest order to buy at the opening of trading. One broker noticed this and said to two other brokers, “I will wait to drop the price,” meaning that the broker was going to enter an order to buy at a substantially lower price just prior to the beginning of trading. The other brokers responded, “It is not fair,” counseling the buying broker to not drop the price (and, just to make sure, these brokers also told the selling broker what the other broker might do). While aware of what the rules allowed, the buying broker did not simply “drop the price” to take advantage of the situation, but announced an intention to other brokers. In making this statement, the broker initiated a conversation about the potential trade, opening the opportunity to hear what others thought of the trade, rather than simply taking advantage of the rules. The other brokers' response—that it “is not fair”—expresses the importance of a normative order outside that of the trading rules on the floor.

How Brokers Handle Rule Violations

In both Ghana and Fiji, brokers are likely to protest a competitor's rule violation if the violation has a direct impact on their orders. In the case of rule violations that do not have a direct impact, brokers in Ghana and Fiji respond differently. Brokers in Ghana rely on internal status and practice norms, while brokers in Fiji call upon the resources of the regulatory agency.

After initial disputes about the MKT tactic were settled, more brokers on the GSE used it over the next few weeks. Some brokers, however, did not understand the technical details of how to use the tactic properly, which depended upon the broker time-stamping and listing the buying and selling orders in the proper sequence.Footnote ²⁰ On one occasion, a broker listed orders in the wrong sequence. Another broker, who had no affected orders, checked the tickets and pointed out the error. This second broker stated, “Brokers should be watching each other. We are each other's keepers. When I use the MKT order, I do it right. I want to make sure others do it right.” The broker is not expressing a sense of competition—that the broker wants to make sure others “do it right” so that he can make sure to not lose any trades—but is enforcing a norm and educating other brokers in the proper use of the technique. This sense of norms and tutelage also highlights an element of status involved in proper execution, witnessed by a conversation between two brokers:

Broker B, who had made an error using the technique in the previous trading session, was rebuilding status by demonstrating a more complete understanding of how and why the sequencing of orders matters. Broker A's response indicates that competence and status, rather than proving that the technique and the trade were proper, were the main concerns. Neither the broker who “wants to make sure others do it right” nor Broker B were directly affected by the potentially offending orders, yet they commented on them. In this regard, while there is a concern that the procedure used is correct within the rules, the consciousness of these rules is strongly mediated through status and practice norms, as they reflect on the “kind of broker” others think one is.

In contrast, brokers in Fiji used technical rule violations as a more direct part of their competitive toolkit by alerting regulatory officials at the CMDA. Thus, when a brokerage firm learned that a competitor had accidentally sold a larger quantity of shares than the competitor had, the firm tipped the CMDA. These actions can cause some short-term problems for the firm that is investigated and are also a means by which to “spoil the identity” (Reference GoffmanGoffman 1959) of the competing broker, perhaps bringing a greater degree of scrutiny in the future. Thus, allegations occur even when there may be no clear rule violation. When one firm was upset about the manner in which a competing broker sold shares to raise the cash position for a client, the firm complained to the CMDA that the broker was engaged in market manipulation. While the CMDA investigates complaints that are backed with evidence, an official noted that “too often, brokers talk to [a CMDA official] on the floor [of the SSE]. That's competition and rivalry.” Investigations, even absent any finding of wrongdoing, can have a significant impact on a firm.

The differences between how status is implicated, how sanctions are applied, and how arguments are framed relate to the more fundamental difference between internal, floor-based norms in Ghana and the external rule orientation in Fiji. In both instances, there is an attempt to have a sanction applied to the violator. In the case of the brokers in Ghana, this attempt is through impugning the competence and status of the brokers vis-à-vis the other brokers. Compliance, therefore, is not necessary for the sake of following law, but for individuals to uphold their character within this situated field. In Fiji, however, the attempt to sanction invokes the more formal process of a regulatory agency investigation. Such efforts are not concerned with the status of brokers amongst brokers but may be attempted to harm the status of the competing firm with the regulatory agency. Thus, while brokers in Fiji may want to illustrate compliance to other brokers, the rationale behind this presentation is to avoid unwanted regulatory attention, rather than to uphold one's character in the eyes of one's peers.

How Brokers Express Trading Norms

In both markets, brokers have substantial influence on the prices of clients' orders and, as a result, subsequent trades. Brokers have interests in presenting the market as a stable place to invest to a public that is not accustomed to financial volatility and in maintaining their share of the market. Within this environment, certain norms of how to trade have formed, particularly in relation to price declines (Reference LarsonLarson 2001).

On the GSE, brokers used to be expected to “restore the price” of a company's shares if they bought or sold these shares at a price lower than the previous closing price, particularly in the case of large, block trades. Brokers would restore the price, creating a trade by listing orders to buy and sell a small number of shares (typically one hundred, the minimum size of a trade to change a price) to bring the closing price in line with the previous day's closing price. Under the new trading system, the norm of restoring the price was violated when a broker needed to match a large buying order with a large selling order at a price 2% below the previous close. To do this, the broker first had to supply shares to an open order to buy shares at the previous closing price. When other brokers saw the trading, they became upset. The reactions were swift: “Now they [the shares] have come down. You have to take the price back up” and “Push the price up.” When these simple commands did not work, the brokers shifted tactics. One broker stated, “We are learning. You should have reserved one hundred [shares] out of that and closed it at [the previous closing price].” Another broker reinforced this lesson by stating, “If they had bought one [hundred shares fewer] and the rest at [the previous close], that would be fine.” When even these comments did not achieve the desired result, one broker announced to all the other brokers, “They are going to go home and they are not willing to do anything about it. … They don't care.” No broker, other than the broker who engaged in the trade, expressed an opinion that differed from the general negative assessment. Away from the floor, two brokers expressed that they had no problem with the trading. The strength of the floor norms is demonstrated in that one of these brokers was one who was complaining on the floor—such an expression was possible only in a private interview (whether or not it was sincere).

When the broker in Ghana violated the norm of restoring the price, the reaction of the other brokers was immediate and vocal. The matter-of-fact commands from the other brokers to bring the price up demonstrate that the norm was unproblematic in their eyes, despite explicit prohibitions against false trading and market manipulation in both the trading rules and the SIL.Footnote ²¹ When the initial attempt to alert the broker to the duty to restore the price was unsuccessful, the other brokers undertook it as their duty to educate the broker as to how the trade really should have happened, offering an opportunity to repair the reputation of the trading broker, who could align the action as a simple mistake. When even this was not successful, the other brokers resorted to a shaming tactic to sanction the broker (“they are not willing to do anything about it … they don't care”). While the “offending” broker could have sought protection by claiming a defense that the trading rules established that the trade had to happen in this manner (which would have been a true statement), this was not the case. In response to the other brokers' comments, the trading broker responded, “We had to amend the order so many times. You should accept that you do not understand the situation.” Instead of using the rules that supported the actions in this situation, the broker appealed to other brokers to accept the trading as excusable due to the circumstances. Further, the broker's claim that others did not understand was not concerning the trading rules, but the particular situation of the clients for whom the broker was trading.

In Fiji, there was a similar norm to maintain stable prices in certain shares. In one instance, however, a broker had a client who needed a sum of money in a short time frame and gave the broker instructions to sell shares as low as 10% below the previous closing price. During trading, the broker ended up selling some of the shares 5% below the previous close, which is where the issue closed that day. After trading, the broker explained to others the situation of the client, giving an excuse for the violation of the norm. More telling, however, was the reaction of another, newer broker who said in response, “Don't worry. I'll make sure the price goes up on Monday.” After the inexperienced broker was out of earshot, one bystander (an official of the regulatory agency) said, “That was a very stupid thing to say.” In recounting the story later, the initial broker stated that the new broker's comments showed a lack of skill, since a “good” broker may make sure the price goes up but not say anything about it publicly.

The importance of rule orientation among the brokers in Fiji is also evidenced by the explicit focus on the trading rules that would be expressed in the brokerage offices—particularly when brokers were violating the rules. While I was with a brokerage firm, a broker had prepared orders for that day's trading under the assumption that this broker would gain recognition on the trading floor prior to competing brokers and be able to transact a number of buying and selling orders at one price; however, at the call market, a competing broker was recognized first and listed an order to buy (that was not filled) at a different price, preventing the broker from listing the prepared orders. Rather than place the orders and trade against the competing broker, the broker simply held the orders for the next session. Knowing that the CMDA could check specific order forms and understanding the applicable rules (orders must be placed at the next call market after receipt), the broker said,

The broker knew that not placing the orders is a violation and that the CMDA may well discover it. To minimize the likelihood of discovery, the broker must have sufficient awareness of the rules and, thus, know that the date on an order form should be the same date (or one day earlier) as the order is placed on the board.

The difference in rule consciousness between the two sets of brokers is quite apparent. Knowledge of these rules is part of the cognitive frame used by the brokers in Fiji in preparing and processing orders and in trading on the exchange, particularly when the brokers use formal rituals only to comply. Such compliance, however, requires formal silence on the trading floor about the norms that violate the rules. Certainly, this rule consciousness is related to the more constant presence of the regulatory authority in Fiji as compared to Ghana; however, brokers themselves reproduce this orientation toward the formal rules in their practices of competition and measures of status. The broker who openly expressed the price norms in Fiji was seen by colleagues (and regulators) in a manner similar to the broker in Ghana who violated the norm of restoring the price. When trading norms are violated in Ghana, the attempt is made to educate or shame, keeping the norms within the trader-based group. In Fiji, this role of education and shame is transferred to the regulatory body.