Shareholders in the Dutch Eighteenth-Century Atlantic Trade

This article provides the first quantitative evidence of the indirect benefits of shareholders of the Middelburgse Commercie Compagnie (MCC), a Dutch Atlantic trading company and the biggest slave trader in the eighteenth-century Dutch Republic. MCC shares provided its owners with a preferred status as suppliers and customers of the company. This article focuses on two years, 1725 and 1770


Introduction
This article presents new insights into the economic impact and organization of the eighteenth-century Dutch Atlantic commodity and slave trade with an analysis of the suppliers and customers of the Middelburgse Commercie Compagnie (MCC), a Dutch Atlantic trading company and the biggest slave trader in the eighteenth-century Dutch Republic. In the literature, the company has often been described as unprofitable. 1 As an explanation for this seeming peculiarity, historians have argued that a significant benefit to the MCC directors and shareholders was their preferred status in supplying goods and services to the company and in purchasing returning commodities. 2 Though the argument is persuasive, the empirical evidence for this claim remains fragmented and limited. Corrie Reinders Folmer-Van Prooijen comments on this issue: 'It is not possible to quantify the gains made in this way but they must have been sufficient to make [the directors'] often considerable efforts worthwhile.' 3 While a complete assessment of the benefits may not be possible, this paper makes a significant step in quantifying the benefits to directors and shareholders as preferred suppliers and customers of the MCC using a sample of voyage record transactions from 1725 and 1770. The year 1725 serves as a benchmark for the company's early commodity trade, and 1770 serves as a sample year for the slave-trading period. The time-intensive nature of transcribing and analyzing the data currently prevents the addition of more years for analysis. This paper investigates to what extent shareholders and directors benefited from their position as preferred suppliers and customers of the MCC.

Historiographical relevance
The MCC was established in 1720 with the explicit goal to reinvigo rate commerce and the local economy of the city of Middelburg on the island of Walcheren in Zeeland. In the preceding decades, many commercial activities relocated to Holland -at least, that is what the MCC founders claimed. To avoid further decline, the municipal government and a group of capitalists previously invested in privateering formulated their plan to revitalize trade and shipping in Zeeland and raised capital for a new joint-stock company, the MCC. 4 Investors were not only enticed with potential dividends, but also with the privilege to reap benefits from these renewed trading activities by directly participating in them as suppliers. Article 24 of the MCC statutes prescribed that goods should be purchased publicly at the lowest price. Still, if shareholders offered goods at a competitive quality and price, they would receive a preferential status as suppliers. 5 The MCC's trading activities initially varied; they included many European, West African, and circum-Caribbean destinations. Reinders Folmer-Van Prooijen provides a detailed account of the MCC's early commodity trade and its shift towards the slave trade. However, the literature on the MCC's early period is relatively limited. 6 In total, the company organized 300 voyages between 1720 and 1802, an annual average of approximately 3.7 voyages. 7 The trade in enslaved Africans became the main activity of the company in the second half of the eighteenth century. 8 This gruesome trade was one of the few commercial activities that was not controlled by competing Amsterdam merchants. 9 Zeeland accounted for 78 percent of the Dutch slave voyages after 1734, and the MCC, as the biggest participant, organized 113 slave voyages between 1732 and 1802. 10 The well-preserved records of the MCC, therefore, play an essential role in the historiography of the Dutch slave trade. In 1961, Willem Sybrand Unger opened the debate about the economic impact of the slave trade with his assessment of 101 MCC slave voyages. While occasionally voyages could lead to large profits, he found the average profit to be very limited. 11 Building on Unger's estimates, Johannes Postma claims that 'the profits of Dutch free traders were amazingly low' , with no more than an average of 2 percent annual return on investment. 12 Likewise, Roger Anstey commented 'that the Dutch deemed it worthwhile to continue in a trade whose return [. . .], at an annual rate, was 1.43 percent.' 13 While these claims hold in the narrow definition of MCC accounting profits, Matthias van Rossum and Karwan Fatah-Black explain that these firm-centred 'profit calculations' do not sufficiently capture the economic significance of the slave trade. They propose a gross margin calculation instead and find that the total impact of the Dutch transatlantic slave trade of the seventeenth and eighteenth centuries lies between 63 and 79 million guilders. 14 Expressed in the eighteenth-century official bullion content of guilders, this would be between 605,430 and 759,190 kilograms of silver. 15 Such broader studies have been critical in shifting the attention away from the narrow focus on the profitability of the slave trade to a more complete assessment of its economic impact on the Dutch Republic.
In the historiography of the British slave trade, Eric Williams's Capitalism & Slavery initiated the debate about the role of the slave trade and the system of Atlantic slavery in connection to the Industrial Revolution in Britain. 16 According to Gerhard de Kok, Williams assumed a high profitability in the slave trade, which is partially responsible for the subsequent focus on profit calculations. 17 Recent estimates of profits in the eighteenth-century British slave trade have converged on an average of at most 10 percent, lower than previously assumed. 18 Still, Kenneth Pomeranz and Joseph E. Inikori emphasize that the Atlantic system, with its slave plantation complex and American bullion necessary for European trade in Asia, provided the resources and market conditions for the Industrial Revolution in Britain. The slave plantation complex supplied cheap commodities, especially sugar and tobacco, to Britain, while the American and Caribbean colonies became a growing export market for manufacturing industries in the core. Furthermore, American bullion was crucial for the imports of Asian commodities, especially textiles, which were also often re-exported to the Atlantic, including West Africa. 19 Economic development on the scale of the Industrial Revolution in Britain did not occur in the late eighteenth-century Dutch Republic. Still, many sectors benefited significantly from their connection to the Atlantic commodity and slave trade. Victor Enthoven estimates that the total value of Dutch Atlantic commerce in the eighteenth century exceeded the value of Dutch trade with Asia, while historians had pre-15 Based on 9.61 grams of silver per guilder in Jan Luiten Van Zanden, 'The prices of the most important consumer goods, and indices of wages and the cost of living in the western part of the viously assumed the Verenigde Oost-Indische Compagnie (VOC) trade with Asia was far more significant than Atlantic commerce. 20 Pepijn Brandon and Ulbe Bosma find that the total economic activity related to Atlantic slavery -including import, export, and processing of slave plantation commodities -made up 5.2 percent of the gross domestic product of the Dutch Republic in 1770. 21 In Zeeland, apart from providing wages to many sailors, sectors such as ship construction and gunpowder production benefited from the Atlantic slave trade. 22 Cacao processing in Middelburg thrived because of its trade connections to plantations in Suriname. 23 Moreover, a large and varied group of Dutch investors capitalized on slave ship insurance. 24 De Kok focuses on the effects of the slave trade on the urban economies of Walcheren and approximates that, in 1770, 10 percent of the income of Middelburg was directly connected to the slave trade. 25 In De Kok's more recent Walcherse ketens. De trans-Atlantische slavenhandel en de economie van Walcheren, he estimates that, between 1755 and 1780, 5 to 6 percent of income in Middelburg was linked to the slave trade. 26 The financial benefits of the MCC operation to its shareholders and directors were indeed significant, but the profits were often not directly derived from dividends or an increasing share price. De Kok explains that: 'many investors in the slave trade may have been less interested in its direct profitability and more in obtaining a beneficial position as preferred supplier to slaving companies and partnerships.' 27 Reinders Folmer-Van Prooijen explains that, in the first half of the eighteenth century, directors, such as Joost van Huijen, owned local gunpowder mills, which supplied gunpowder to the MCC. Not only is it assumed that shareholders of the MCC supplied much of their goods, but there is also evidence suggesting that there was a high representation of directors among the customers of returning goods. 28 MCC directors were still involved as suppliers of the company in the second half of the eighteenth century, when, for instance, directors frequently insured MCC ships. 29 In 1770, shareholders in the slave trade also still owned many of the local gunpowder mills near Middelburg to supply slavers, including the MCC. Additionally, De Kok finds that smaller suppliers, such as bakers supplying bread to the MCC, were also shareholders. 30 However, the magnitude and implications of these indirect benefits are yet to be fully understood.
The peculiar organization of the MCC can also be understood in terms of a typical challenge facing early modern trading companies: the principal-agent problem. The standard case concerns opportunism by agents operating far away from the firm's headquarters, most commonly through private trade by captains and other personnel, which worked against the firm's objectives of maximizing profits for shareholders. 31 For the Hudson's Bay Company, another eighteenth-century trading company, Ann M. Carlos and Stephen Nicholas argue that the company overcame its agency problems through contracts, control systems, and a social structure by the end of the eighteenth century. 32 However, in this specific and atypical case of the MCC, the central question is whether the directors and the active shareholders (i.e., shareholders who also acted as suppliers or customers of the company) themselves posed a risk to the overall profitability and objectives of the company, as they acted both as principals and as agents, with private interests as suppliers and customers of the company. If, as suppliers and customers, shareholders received significantly better prices than non-shareholders, it could compromise the profitability and even continuation of the company. Still, the MCC continued operating into the nineteenth century, so this must have been a manageable situation.
To understand how much shareholders and directors benefited from their position as preferred suppliers and customers of the MCC, this paper aims to answer four research questions. Firstly, to what extent were shareholders and directors represented in the transactions of the MCC in terms of total value? Second, what proportion of all shareholders and directors was directly involved in the daily operations of the MCC? Third, was the MCC paying and receiving the market price when dealing with shareholders and directors? Fourth, to what degree was this organization stable over time? While there are limited years to investigate, any clear differences between 1725 and 1770 should become evident.

Sources and method
The online MCC records form the basis of this study, and the primary analysis of this paper involves matching the names in the shareholder register with the names in a sample of transactions, thus providing the critical data to answer questions 1 and 2. The voyage records (scheepsboeken) offer the most suitable transaction data and organized information with the names of suppliers/customers and details on the goods and services. The data are divided into key categories: outfitting, cargo, and returning commodities. 33 The annual MCC balance includes a shareholder record. As the shareholder register could change continuously, the selected transactions need to be within a narrow range of the date of the sample register.
A comparison of a year in the early period with a year in the later period can help us understand how continuous the company's organization was over time (question 4). The year 1725 serves as a benchmark for the early commodity-trading period of the MCC, and 1770 provides a sample year for the later slave-trading period of the company. By 1725, the company was firmly established after the initially chaotic subscription process in 1720. 34 1770 is a suitable year to study the later slave-trading period of the MCC, as it is uninterrupted by wars, like the Seven Years' War (1756-1763) and the American War of Independence (1775-1783). It is also commonly used as a benchmark in other studies of the Dutch slave trade. 35 The decision to focus on two years that are relatively far apart means that it is not possible to comment on the durability of relationship of suppliers and customers with the MCC, as almost none of the individuals active in 1725 would be alive in 1770.
The duration of a complete voyage could be up to two years, but only the data within the relevant periods (1725 and 1770) are analyzed. Ideally, only data from six months before and after the recorded shareholder register are included, but there are three voyages that fall slightly outside this range. For both 1725 and 1770, records for five outgoing ships (outfitting and cargo data) and five incoming ships (returning commodity data) have been transcribed and analyzed. Figure 1 provides a table with details on the ships used for this study. For 1725, the shareholder register was recorded at the end of August 1725. This was the ending date of the fiscal year in the first few years.  To find out whether the MCC was paying and receiving the market price when dealing with shareholders (question 3) can be challenging though not impossible. The MCC records provide data on prices paid by shareholders and individuals without shares, or 'non-shareholders' , when they are selling or buying an identical product. The MCC records often provide insufficient detail on the quantity and quality of goods, such as textiles, to draw strong conclusions on their relative price levels, but this is not always the case. For instance, insurance policies are relatively transparent and can be assumed to be interchangeable, so the MCC would be free to choose the cheapest insurance premium.

Challenges and limitations
The expenses and revenue related to voyages accounted for at least 80 percent of the average value in the MCC's profit and loss accounts between 1720 and 1755. 36 However, the MCC also included wharf and ropery operations, as well as financial activities (borrowing and lending). This poses a challenge when answering question 2 (what proportion of all shareholders and directors was directly involved in the daily operations?). The main groups which are potentially excluded from the sample are employees in the offices in Middelburg, creditors and debtors, and shareholders solely involved in the activities of the wharf or ropery.
The voyage records include many internal transactions, which poses a more serious challenge than anticipated. For instance, there are transactions with the wharf, for which the original external supplier cannot easily be identified, or there are movements from and to stockpiles, so that no final supplier or customer is recorded. For the outfitting transactions, internal transactions accounted for 50 to 60 percent of the total value of the combined transactions in both 1725 and 1770. For the cargo and returning goods records, this figure was less significant, though also up to 30 percent of the cargo value was related to internal transactions in 1770.
For clarity, the results exclude internal transactions or transactions that do not clearly relate to a supplier or customer (e.g., wages of groups of workers, or taxes in the harbour). The categories which are included in the figures are the shareholders, directors, and non-shareholders. The shareholder category includes firms of shareholders and widows of shareholders. The only other category remaining is 'unknown' (suppliers/customers), because of missing or illegible pages in the voyage records in 1725. These entries are likely related to external entities, but they cannot be identified.
The number of creditors and office employees was limited, so it is possible to identify most of them from other records or secondary literature. However, there is no perfect solution for the possible exclusion of active shareholders who do not appear in the sample, so an answer to question 2 remains tentative. In addition, it is possible that the involvement of shareholders fluctuated significantly in the years between 1725 and 1770, and with the method employed here, these shifts remain largely unknown. One final challenge concerns the potential misidentification of shareholders from inconsistencies in the spelling or handwriting of their names by MCC bookkeepers. More questionable 'matches' have not been considered as shareholders. It is important to consider that the estimates provided in this paper are probably relatively conservative, because of the aforementioned factors.

Investors
In August 1725, there were 385 registered shareholders of the MCC: 41 women, 343 men, and one firm. The majority of shareholders, 210, owned a full share with the par value of £500, granting them voting rights. 37 The shareholders who held a full share, or hoofdparticipanten, elected a number of representatives (commissarissen) among themselves who acted as a check on directors and performed audits. 38 The financial incentives to become a hoofdparticipant differed, but some were probably interested in the chance of becoming a director, a highly coveted position among the hoofdparticipanten and Middelburg elites in general. To be eligible for the director's position, shareholders needed to own at least a full share. 39 Out of 385 shareholders at least 131, or one third, appeared in the transaction records. Individuals with less than a full share were more involved (40 percent involvement) than the hoofdparticipanten (29 percent involvement). Indeed, the lower the investment, the higher the proportion of involved individuals. The 70 shareholders with the lowest fraction of a share (1/6 of a share with a par value of £83:6:8) were involved 56 percent of the time. Still, the hoofdparticipanten account-37 Currencies and percentages: monetary value in this paper is presented in pound Flemish (pond Vlaams), which was the official currency in Zeeland, for which the symbol was £. The pound Flemish was subdivided into schellingen and groten (1 pound Flemish = 20 schellingen = 240 groten). One pound Flemish was officially exchanged into six guilders, the currency used in Holland and most of the Dutch Republic. See Reinders Folmer-Van Prooijen, Van  ed for much more value in the transactions than those with less than a complete share.
In 1725, apart from the shareholders who acted as suppliers and customers, the MCC was also connected to shareholders through other arrangements, such as formal labour and credit agreements. The wages for vrijluiden or vrijmannen, who were responsible for the logistics and administration of goods within the city and docks, are sometimes registered as wages for workers, but they were in fact also shareholders. 40 Another group which is excluded from the voyage records are the creditors of the MCC. Contrary to Paesie's finding that only shareholders could become creditors, some appear to be non-shareholders in 1725, like Arend van Sonneveld, for instance, who lent the MCC £1000. 41 Still, in other cases the MCC owed large sums to shareholders. For example, the MCC owed £25,433:6:8 to Maria van de Claver, the widow of former MCC director Pieter de la Rue. 42 By the start of 1770, there were 424 shareholders: 46 women, 372 men, three organizations, one firm, and two groups of children who inherited shares from a parent. Among the shareholders there were 198 hoofdparticipanten, in addition to 225 smaller shareholders (less than a full share). This excludes the MCC, which purchased its own shares in previous years. Again, out of all 424 shareholders, at least 144, or one third, acted as a supplier or customer in the MCC slave voyages of 1770. There are two cases in which the ownership of a firm is ambiguous, as there are several family members in the shareholder register with a name corresponding to one firm. The biggest shareholder from the family has been selected as the active shareholder. Finally, there is a group of insurance brokers (shareholders), who are usually recorded together, and whose individual involvement is hard to measure. The MCC recorded few creditors in 1770, so this is largely inconsequential to the analysis.
In 1770, comparable to 1725, the smaller shareholders were more involved than the hoofdparticipanten. drop in transaction value connected to shareholders in 1770, which can partially be attributed to the decreased capital of the MCC through losses in the early years. However, it also appears that the directors were trading less with the company in 1770 than in 1725, which reduced the overall involvement of the hoofdparticipanten.

Outfitting
In 1725, 73 percent of the recorded outfitting value was linked to shareholders, and at most 27 percent was supplied by non-shareholders, so shareholders supplied more than twice the value of goods and services connected to non-shareholders. Surprisingly, the directors were almost absent from the outfitting costs, accounting for less than one percent of the outfitting costs.
A relatively high expense concerned the sails and ropes usually supplied by Abraham Claudoré Jr., bookkeeper of the MCC ropery, the Swarte Cabel. This transaction likely took place internally within the MCC ropery, which did not directly benefit Claudoré Jr. 43 Otherwise, there was no obvious difference between the goods supplied by shareholders and non-shareholders, except that insurance was typically supplied by shareholders. Presumably -particularly for individuals operating in highly competitive markets, like the insurance market -holding a share would be useful in becoming a supplier.
There is a seemingly substantial decrease in shareholder involvement in 1770 compared to 1725 for the outfitting costs. In 1770, the outfitting costs connected to shareholders accounted for £3,728:5:3, while non-shareholders accounted for £5,946:5:3 of the costs. There are a few reasons for this apparent increase in the involvement of non-shareholders. First, many of the employees of the MCC no longer owned shares in 1770. The bookkeeper for the MCC ropery, Anthonij d' Aillij, supplied £1,852:1:11 worth of goods, often described as 'rope work' (touwerk). 44 In 1725, the bookkeeper of the ropery, Claudoré Jr., was a shareholder, whereas, in 1770, d' Aillij was not a shareholder. Again, it is perhaps better to understand these purchases as internal MCC transactions, as the bookkeepers likely did not directly benefit from these transactions. Second, other employees, such as the draymen and the vrijmannen were no longer shareholders of the company in 1770. In total, the MCC spent £968:0:9 on fees and other logistical expenses, which exclusively went to non-shareholders in 1770. Finally, the ship Welmeenende was purchased from a non-shareholder for £525:10:0. Adding up these expenses (the ropery, fees, logistics, and the Welmeenende) results in a total of £3,345:12:8, more than half (56 percent) of the outfitting costs related to non-shareholders, which explains the higher involvement  26 In 1770, shareholders supplied most of the cargo for the five outgoing slave ships. However, there are still some significant differences with the 1725 sample. Surprisingly, the MCC directors were completely absent as cargo suppliers. A significant difference with the earlier period was the increased number of firms supplying the MCC. The biggest supplier among these firms was Snouck Hurgronje & A: Louijssen, which supplied textiles for the five outgoing ships, in total £4,702:10:5. The firm, owned by Steven Matthijs Snouck Hurgronje and Abraham Louijssen, was a slave-trading company based in Vlissingen. Snouck Hurgronje was an MCC hoofdparticipant. Abraham Louijssen was not registered as an MCC shareholder, but it was probably sufficient to have one partner with MCC shares to become a supplier.
The MCC also separately recorded expenses related to the cargo, such as cargo insurance and wages for logistical work. In 1725, the cargo expenses amounted to £2,787:15:5, most of which was related to insurance. £2,561:3:5 can be directly linked to shareholders. In 1770, these expenses totalled £5,805:16:3. The MCC paid £1,832:0:2 to shareholders for various products and services; the insurance premiums accounted for £719:13:2. In 1770, half of the expenses were related to the fee (lastgelt or recognitie) to the Dutch West Indian Company (WIC). After losing its monopoly in 1730, the WIC had retained certain privileges on the coast of West Africa, so the MCC paid a significant fee, based on  the size of the ships, for the legal right to trade there. 46 While a comparison is not as useful here, it is clear that shareholders played a dominant role as suppliers in both years.

Expenses for incoming ships
The MCC recorded two more separate categories of costs for returning ships: the expenses upon return (Thuiskomende Ongelden), and the costs related to the returning commodities (Onkosten op de Retouren). Conceptually, the former related mostly to the ship and its crew similar to the outfitting costs of outgoing ships, while the latter related to the returning cargo similar to cargo expenses for outgoing ships.
In 1725, the expenses upon return primarily consisted of the wages for officers and sailors, in total £7,454:14:2, or 91 percent of the total costs. A breakdown of the involvement of shareholders and non-shareholders is less meaningful here, because of the irregular expenses and transactions related to the half-profit goods. In 1770, the Thuiskomende Ongelden consisted primarily of the wages but also included some payments £110:5:5 to bakers and coopers with shares. The Onkosten op de Retouren mostly included fees to the officers and an MCC agent, Adriaan Gootenaar, related to the sale of enslaved Africans in the Caribbean. Shareholders were paid for expenses related to the auction of African products (mostly ivory) returning on the ship Zanggodin and the insurance for shipping this cargo from Suriname to Middelburg, in total amounting to £228:17:8.

Sale of returning goods
In 1725, as a single category, the returning goods accounted for most of the value in the voyage records, £48,077:0:6. A significant share of the returning goods was 'retained' by the MCC because the Beurs van Middelburg returned with £20,463:13:1 worth of cargo. This has been excluded from the analysis. Reinders Folmer-Van Prooijen explains that the MCC was unable to sell most of its goods on this specific voyage. 47 Shareholders bought 43 percent of the remaining returning goods, and the directors purchased almost 16 percent. Non-shareholders accounted for a minority of the purchases, 39 percent. Unknown or unidentifiable individuals bought 2 to 3 percent of the returning commodities.
In 1770, slave plantation owners who bought enslaved Africans from the MCC in Suriname or Essequibo usually paid with bills of exchange, a type of credit instrument. Many MCC ships returned in ballast, as planters would often directly ship plantation produce to their creditors in Holland to pay off their loans. The MCC typically only returned with bills of exchange which could be cashed on the due date at the acceptors in Holland, but it was also common to discount these bills to other parties in Middelburg. 48 The registered 'customers' of bills of exchange could thus either be the final acceptor or the entity to whom the bill was discounted in Zeeland. 49 79 percent of the total value (£81,505:18:0) of returning 'goods' in Middelburg consisted of bills of exchange. In addition, 10 percent of the transaction value related to protested bills of exchange (8 percent), pretenties (2 percent), and an assignatie (less than a percent). A pretentie was probably a type of financial claim, but further details are missing. The assignatie was a type of bill of exchange. 50 Because of the irregularities in accounting for financial instruments it is complicated to provide an unambiguous breakdown of customers similar to 1725. However, some general trends regarding these bills can be addressed.
Shareholders played a limited role in discounting bills. Approximately 5 percent of the value of the bills of exchange was discounted to shareholders or firms owned by shareholders, primarily because   of MCC director Mounier (total transaction value of £2,681:11:0). The VOC was the most active discounter of bills of exchange, buying 37 percent of the total value of bills of exchange (29 percent of the total value of returning goods). The only ship returning to Middelburg with commodities was the Zanggodin, which returned to Middelburg with a substantial amount of ivory and other products from Guinea. Shareholders and their firms accounted for slightly more than half of the revenue from this incoming ship. However, the Zanggodin is an exception to the rule by returning with commodities instead of bills of exchange, and the benefits to shareholders as customers of returning commodities were not as structural as for the earlier commodity trade of the MCC. The insurance payments for 'general average' ('Avarij Grosse') -in this specific case a payment for the murder of African captives during an uprising on the Zanggodin -also reveal some of the underwriters behind the insurance brokers of MCC slave ships. 51 The six underwriters linked to broker Fredrik Dibbetz & Son were most likely all from Amsterdam and did not own any MCC shares. However, four of the other seven underwriters, which were connected to the group of insurance brokers in Middelburg, were MCC shareholders. Furthermore, one of the remaining three underwriters was the Middelburg Insurance Company (Assurantie Compagnie Middelburg), which held a special status as a preferred supplier of the MCC in its statutes. 52

Price-setting
The small number of suitable examples for 1725 precludes any generalization about the differences in prices paid to shareholders compared to non-shareholders, but there are some minor indications that shareholders may have received better deals. Insurance policies potentially provide some insight here because of their transparency and interchangeability as a product. For example, the ship Cornelia travelling to Suriname had two insurers: shareholder Anthonij Baert from Amsterdam, and non-shareholder Pieter Kops. Both provided insurance for the 'hull' of the vessel, in addition to separate insurance for ransom in case the captain was captured by Barbary pirates. Shareholder Anthonij Baert received a 9 percent premium for his insurance, while the MCC only paid an 8.5 percent premium to non-shareholder and insurance broker Pieter Kops. However, both insurers received the same 3 percent insurance premium for the (cheaper) piracy insurance. While this situation is insufficient to draw strong conclusions, the directors also behaved in ways which appear to favour private interests over company interests.
Based on the 1770 sample, there is no apparent difference in the insurance premiums paid to shareholders and non-shareholders. These records suggest a type of 'hybrid' system in which the MCC engaged with one insurance broker from Holland, in addition to a group of insurance brokers in Middelburg, for each voyage. The insurance for the slave ships involved a relatively large group of suppliers. Suppliers in other product groups were often also numerous -for instance, there were 26 bread suppliers (20 of them were shareholders), but none of them supplied the bulk of the bread for a voyage. However, as De Kok explains, by 1780, only one baker acted as a supplier for each voyage, so there was not a limited capacity to supply bread by individual bakers. Capital for insurance was limited in Middelburg, though, so wealthy insurers from Holland were probably hard to substitute. 53 It seems that in many cases the MCC was balancing the company's needs with the legal right of shareholders to act as suppliers.
There are at least several other instances in which the MCC employed numerous interchangeable suppliers, such as coopers or gunpowder suppliers, without favouring any of them in a significant way. This is also the case for the bakers supplying bread, and perhaps to some extent, the insurance brokers. For example, for the ship Nieuwe Hoop, there were four gunpowder suppliers, two shareholders and two non-shareholders, who all received £189:17:2 for 104 powder kegs. 54 One possible explanation for this practice would be that by engaging with both shareholder and non-shareholder as suppliers of goods, like gunpowder, the price of these goods could effectively be driven down to a local market level by non-shareholders wanting to supply a share of the required goods. This practice would allow the MCC to pay the market price for products while upholding its commitment to treat shareholders as privileged suppliers. If price competition could only occur between shareholders, it could enable anticompetitive prac tices. The intentions of the MCC directors are unclear from these financial records, so further evidence of such policy from qualitative sources would be necessary. This explanation is also not universally applicable; for instance, all chest makers were shareholders.

Principal-agent problem
It is not completely clear if active shareholders profited at the expense of inactive shareholders or whether the company supported the interests of both groups. In effect, however, only directors had enough formal decision-making power to advance their own private enterprises through the MCC, and the objectives of the directors and the company appear to be better aligned in 1770 than in 1725. There are three points of evidence for considerinng the agency problems over time.
First, the qualitative evidence shows that, in the early years, the behaviour of the MCC directors was openly questioned. The shareholder representatives (commissarissen) were wary of excessive opportunism by agents, including the directors. The suspicion towards directors was not unfounded. For instance, in 1730, Boursse & Grijmalla (owned by director Hendrick Boursse) purchased returning cacao before an official public auction, which clearly violated the official regulations of the MCC. For unknown reasons, after 1733, there were no further public confrontations between the directors and shareholder representatives about this issue. 55 The directors may have changed their behaviour, or the problem simply persisted but was no longer officially documented.
Second, private trade was potentially much more lucrative in 1725 than in 1770. In 1725, the transaction value from private trade with the MCC linked to the directors, either directly or through their firms, was £18,139:3:1 compared to £2,885:6:0 in 1770, most of which (£2,681:11:0) was linked to director Mounier, who purchased bills of exchange from the MCC. However, the financial benefits from this trade in bills of exchange are uncertain, as the bills were increasingly difficult to liquidate, as a result of recurring credit crises in Suriname in this period. 56 Judging merely from these transaction values, the problem of the MCC directors' private trade was potentially greater to the MCC in 1725 than in 1770. One minor qualification would be that the company's capital had diminished significantly after the losses in the early period, but this condition cannot sufficiently explain the decrease in private trade by 1770.
Third, maximizing company profits was more lucrative to directors in 1770 than in 1725. The MCC statutes provided incentives to the directors to run a profitable trade and pay dividends to shareholders. In addition to the requirement to hold two MCC shares, article 20 of the statutes declared that directors would receive 10 percent of the profits or dividend payments. 57 This appears to have changed to 5 percent in the reorganization of 1729. 58 Regardless of the absence of salaries, the directors had incentives to work towards a profitable operation. The capital needed to buy two shares was much higher before the financial disasters in 1725 than in 1770. 59 Assuming the 1725 directors paid the original subscription price, they had invested £11,625:0:0, with significant losses in the following years. The few early dividend payments could not offset the large drop in share value. In 1770, the combined par value of the directors' shares was £14,583:6:8, but the market val-ue, assuming 20 percent of the par value, was £2,916:13:4. 60 The lower share price meant that the dividend yield on their shares was substantially higher than in 1725. The combined total dividends for directors at 3 percent of the par value of their shares and the extra 5 percent of the total dividends would amount to £781:4:4 in 1770, 61 certainly more than they could make in profits from their private trade with the MCC (probably a few percent of the transaction value, £2,885:6:0). Conversely, with £18,139:3:1 in private trade in 1725, it was probably more lucrative for directors to focus on this income stream than to solely aim for maximizing company profits.

Differences in share retention between active and inactive shareholders
A final point to consider is the difference between active and inactive shareholders in the retention of shares. In 1770, active shareholders held their shares for longer periods on average than inactive shareholders, possibly indicating the perceived value of MCC shares for these individuals. Figure 7 shows that this difference is especially clear after two years, in 1772, when the percentage of remaining active shareholders from 1770 was approximately 93 percent, while only approximately 84 percent of inactive shareholders from 1770 still owned shares. 62 Intuitively, active shareholders had an incentive to retain their special status as suppliers or customers. In contrast, inactive shareholders presumably were more sensitive to changes in the share price and individual demand for liquidity. Over time, the difference persists, but the two groups converged slightly, most likely because of external factors, such as death.
The position and motivations of inactive shareholders compared to active shareholders have not yet been considered. While it is not possible to fully assess the situation of inactive shareholders within the scope of this paper, it is possible to give a general impression of their position throughout the century. Inactive investors could receive a high return on investment in specific periods, for instance, in the 1760s. However, for an average long-term shareholder between 1730 and 1780, the annual return would have been moderate depending on when they bought and sold the shares. In other periods, such as the 1720s and the post-1780 period, the return would most likely be negative. 63  was not exceptional in this regard in the eighteenth-century Dutch Republic; VOC shareholders received relatively low returns if they invested at the beginning of the eighteenth century and held these shares until its liquidation. 64

Conclusion
In recent years, new studies of Dutch Atlantic commodity and slave trade have offered a better understanding of their economic impact on the eighteenth-century Dutch Republic. 65 Van Rossum and Fatah-Black also rightly argue that previous firm-centred profit calculations provide a limited understanding of the economic impact of the slave trade. 66 Nonetheless, firm-centred analyses, with a broader focus beyond profitability, remain vital to understanding the organization and wealth distributing mechanisms among stakeholders of these companies. In the case of the MCC, many benefits to shareholders did not come through regular profits, but because they could act as preferred suppliers and customers of the company. The findings presented in this study build on existing evidence and assumptions of the MCC's organization, but this research presents the first systematic study and quantification of such benefits.
In both 1725 and 1770, at least one third of the MCC shareholders profited from their shares as preferred suppliers and customers of the company. For outgoing ships, shareholders directly or indirectly supplied the majority of the goods and services, except for the outfitting of the ships in 1770. Shareholders, by far, supplied most of the cargo on the departing ships in both 1725 and 1770. For returning vessels, most of the returning commodities were purchased by shareholders in 1725. In 1770, the returning 'cargo' consisted primarily of bills of exchange, which were usually not discounted to shareholders; the VOC was the main customer of these bills. The MCC itself accounted for a large part of the transactions in the voyage records, but it is unlikely that these are all internal exchanges, so other records -related to the wharf, for instance -could be used to investigate further connections to shareholders. It also remains notoriously difficult to extrapolate a profit margin from these transactions. Without empirical data on the profit margins of the supplied and purchased products, the precise profit from these transactions continues to be speculative. This paper emphasizes the necessity for more intricate analyses of company structures and implicit beneficiaries to assess the economic impact of the Dutch Atlantic trade in enslaved people and slave-produced goods, while the existing literature often remains focused on more conventional statistics or accounting figures such as company profits. The results presented here underline that financial gains were most likely accumulated by individuals or firms that were adjacent and connected to the trading firms, yet less so by the long-distance trading companies themselves. Still, with the current state of evidence, there are some important qualifications to this conclusion. First, without clear profit margins on products or services, it cannot be quantified how much money was gained by specific shareholders or directors. Second, it is unclear at this time if similar trading companies in this period also provided such indirect benefits to shareholders, or if the MCC was