Abstract
The Mughal exchange economy can be visualised as comprising twin circles of cash and credit. In the first, transactions were conducted in currency money (metallic and non-metallic) consisting of fresh imports and pre-existing stocks. In the second, payments were deferred to pre-fixed dates using credit instruments. The circles of cash and credit grew in the sixteenth and seventeenth centuries, possibly more in the latter. This paper is about a unique professional group, the ṣarrāfs, whose members engaged themselves in both these circles. They were essentially assayers and moneychangers who operated in the market or worked for a client (state, village community, members of the ruling class, merchants), or carried out both functions at the same time. The possession and handling of cash enabled them to diversify their operations, the most important being banking. As bankers, the ṣarrāfs accepted deposits on interest, gave commercial and consumption loans, and transferred money from one place or person to another through bills of exchange (hundī) and book entries (giro). The ṣarrāfs facilitated the movement of money and merchandise also by covering risks on payment of premium (inland, marine and credit insurance).
Ṣarrāfs as Assayers and Moneychangers
As their Arabic name indicated, the principal function of ṣarrāfs in Mughal India was to test (assay) and change money. The Mughal monetary system was multi-metallic and the exchange of one type of coin for another was a popular practice since customers had to pay for goods in the market in the same coin in which the price was quoted. Shopkeepers were not generally expected to have enough expertise to change money and this was the professional prerogative of the ṣarrāfs. The ṣarrāfs assessed the weight and value of the coins for a small commission. A qualitative system of assaying coins was common among the ṣarrāfs in India based on the use of needles and touchstone. 1
An early seventeenth century Dutch account furnishes similar information about the profession of the ṣarrāfs of Gujarat as money-testers. 2 It highlights the trust the market placed in the expertise of the ṣarrāfs in certifying standard weight and fineness. If a coin was not found to be genuine or of the same value as certified by them, the ṣarrāfs compensated the customer. Tavernier tells us that the ṣarrāfs of Golconda sealed bags of coins after ascertaining their value and the ‘bags circulated as sums of money without being opened throughout the year’. 3 Paradoxically, the ṣarrāfs both added to the transaction cost and reduced it.
Copper coins were the money of everyday transaction in Mughal India. Prices and wages were paid in copper coins, and if anyone had a silver coin, it had to be exchanged for small purchases quoted in copper. 4 Such exchanges were undertaken usually by the ṣarrāfs. 5 According to a Mughal official report from Surat, the most important port-city of Gujarat, dated 1661, there were moneychangers known for dealing in small change (şarrāfān i khūrda farosh). The ṣarrāfs purchased freshly minted copper coins (dāms) from merchants controlling the supply of Japanese copper to the mint. The copper coins were sold to those who needed small change, and the silver rupees obtained in the sale were invested in fresh cycles of copper–silver exchange. The commission charged on the exchange constituted the ṣarrāf’s profit. 6 The customers of the ṣarrāfs included those Mughal officials at Surat who supervised the repair of the imperial ship Ganj i Sawāī. The officials needed regular supplies of copper dāms from the ṣarrāfs to pay the wages of carpenters and other workers employed in the repair of the ship. A crisis of copper in Gujarat in c. 1660 disrupted the cycle of bi-metallic exchange when the copper merchants diverted the supply of dāms from the mint to the areas where the coins fetched a higher price in silver. As a result, the exchange rate of dāms rose against rupees adding extra cost to the repair of the ship. The state intervened to restore the supply of dāms to the local economy. 7 In Ahmadabad, capital of the province of Gujarat, the ṣarrāfs dealt with the shortage of copper by giving currency to a fiduciary coin of iron. The Mughal administration, on its part, incentivised copper minting by abolishing seigniorage. 8 In Surat itself, the local administration prohibited the resale of Japanese copper by a proclamation with drum roll throughout the city in order to reserve the supply of the metal for the mint and the moneychangers. 9
The Mughal state used the services of the ṣarrāfs in two ways. First as assayers in the mint of the raw material supplied for minting as well as freshly minted coins. 10 Second, the ṣarrāfs assayed and changed money at the point of revenue collection so that the state received its share without incurring loss. 11
Ṣarrāfs as Mint Suppliers
Beyond internal exchanges, recoinage and remonetisation, the ṣarrāfs handled foreign import of monetary metals meant for the mint. Due to India’s favourable balance of trade, most merchants who exported commodities brought in return foreign coins and raw metals (gold and silver bullion and copper). All this had to be converted into Mughal coins for fresh purchases. In order to save time and avoid dealing with mint officials, merchant-importers of foreign coins and metals sold their consignments to ṣarrāfs. 12 A passage in the commercial report of a French factor gives us a sense of how the ṣarrāfs controlled the supply of silver bullion to the mint after fixing the exchange rates of Spanish dollars imported from the New World and Europe: 13 Earlier, an English factor at Agra, made a similar complaint about the ‘loss’ incurred on the sale of the reales of eight (dollars) on account of the high fineness of the Mughal rupee and the commission charged by the ṣarrāfs. 14 Yet, merchants carried a wide variety of foreign currencies and bullion to India because they knew that goods could be readily bought with cash obtained from the mint or the money market. Around 1659, Francois Bernier sketched the complex route taken by world bullion in search of merchandise with Mughal India at the centre. 15
State Monopoly, Mint Supplies and the Ṣarrāfs
With regular profits assured from assaying, money changing and bullion sales, the ṣarrāf was the master of the money market. However, from the middle of the seventeenth century, Mughal imperial and provincial officials could also claim a share in profits from the buoyant bullion business in the port cities by creating monopolies in what might otherwise have been a free market. 16
The first phase of the monopoly began towards the end of Shahjahan’s reign (c. 1658) with the sale of exclusive rights to coin money at the mint. According to one report, at least 300 ṣarrāfs bought these rights by 1660. 17 As one would expect, this was followed by new market prices for bullion and coins which the ‘mint’ ṣarrāfs were able to enforce with the help of the administration. 18 Accordingly, even at a fixed mint price, the market value of the bullion fell with an increase in ṣarrāfs’ profits. The monopoly holders (‘mint ṣarrāfs) had little time to rejoice in the new set up when they realised that the state was committed to creating its own monopoly in the bullion market at their expense. In 1661, the port administrator (mutasaddi) of Surat, Mustafa Khan, convinced the imperial administration that there would be an additional gain to the exchequer of 100,000 rupees if annual bullion sales to the mint were to be handled exclusively by the state. 19 The new policy was put into effect almost immediately in all parts of the empire and spelled disaster both for the ṣarrāfs and merchant-suppliers. 20 Ironically, with the virtual elimination of the middlemen, this would have meant the beginning of a real open-coinage system, though at a cost the merchants were not prepared to bear. When it came to the sale of their bullion, the English merchants of the Company summed up the restrictive nature of the monopoly and its impact on the ṣarrāfs in a letter sent from Surat in the very first year: 21 The raison d’être of the policy and its immediate impact on the trade of the region were revealed in another letter written two years later: 22
Such a restrictive practice had little chance of success in an open market for two reasons: the strong commercial presence of the ṣarrāfs in all major cities, and the lack of expertise on the part of the state officials in handling the complex business of exchange. The initial protest of the ṣarrāfs to boycott the mint was less effective. But once the merchants joined the fray, causing a disruption in the caravan traffic from the hinterland and a fall in the customs revenue of Gujarat, 23 the authorities were compelled to yield. The ṣarrāfs were back in business. 24 Only this time, they (‘taunksal shroff’) had to face their own kinsmen (private shroff) who opted to operate in the open market against all odds. 25 Even though there was a total prohibition on sales of bullion to anyone except those working with the mint, the free market ṣarrāfs managed to circumvent the monopoly by offering competitive prices and using all possible means to deliver the proceeds to the merchants. 26 This provides the context for the governor (subadar) of Gujarat ordering the mint officials in 1689 AD to restrict the sale of uncoined gold and silver (tilā wa nuqra ghair maskūk) solely to the mint. 27 Subsequently, as a result of a complaint lodged by Mīr Bāqar, the superintendent of the Ahmadabad mint, an imperial order was issued (1698) that, to avoid loss of revenue, no one be allowed to refine (lit. melt) gold and silver anywhere other than the mint. 28
The state control of the supply of bullion to the mint, through a bunch of listed ṣarrāfs, continued as long as the state was strong enough to regulate the market through its various apparatuses. With the decline in state power in the early parts of the eighteenth century, the ṣarrāfs began to retake control of their business as in the pre-monopoly years. Now, the mint was apparently thrown open to all the ṣarrāfs and bullion merchants who could bring supplies as and when they wanted. The mint administration still had a regular supplier of their own—mahājan i muqarrarī— but he had to compete with the rest even though he enjoyed concessions in the payment of seigniorage dues. 29
Ṣarrāfs as Bankers and Dealers of Bills of Exchange (hundis)
Banking in Mughal India involved receiving deposits and lending money by ṣarrāfs on interest. 30 The loans were generally given by the ṣarrāfs to commodity merchants but consumption loans too were advanced to creditworthy members of society. The English and Dutch factories of the two companies in India took loans on interest from ṣarrāfs to tide over the shortage of cash during high seasons. 31 Roques gives us a long description of the network of loan transactions involving commodity merchants, ṣarrāfs and brokers. 32
Roques’ account highlights, among other aspects of commercial usury, the practice of charging compound interest. Short-term loans allowed quick decisions to be taken on the employment of reserve capital for maximum gain. 33 Extensions were granted on the basis of renewed contracts for the lenders to compound interest. 34 A shorter loan contract also gave the merchant the option of paying his debts just after the sale of goods. 35 If a loan had to be contracted for a longer duration, the interest charged on it was slightly higher than the monthly rate. 36
The ṣarrāfs, as bankers, accepted deposits from individuals, merchants and state officials who had cash to put out on interest. A classic description of the practice of bankers accepting cash deposits from customers comes from the letter of an English factory sent from Agra in 1645, a time when the city was losing its status as the capital of the Mughal empire to an emerging Shahjahanabad (Delhi):
Those that are great monied men in the towne, and live only upon interest receive from the sheroffs [ṣarrāfs] no more than 5/8 per cent per month. The sheroffs dispose it of to others from 1 to 2.5 per cent…Now when the sheroff (for lucre) hath disposed of great sommes to persons of qualities at great rates, not suddenly to be called in to serve his occasions, then begin his creditours (as in other parts of the world) like sheepe one to runn over the neck of another, and quite stifle his reputacion. Thus…hath two famous sheroffs bynn served within a month, one of which failing for above three lack of rupees, diverse men have lost great somes and others totally undone thereby; which hath caused men of late to bee very timerous of putting their monies into sheroffs hands.
37
This interesting example is of deposits payable on demand held by the Agra ṣarrāfs from the wealthy people of the city. 38 The amount was then lent out at higher rates of interest to creditworthy clients. The difference between rates paid on deposits (0.625 per cent per month) and loans (1 to 2.5 per cent) made up the profit of the banker. Since the borrowers had taken up money for a fixed duration not to be returned at once, when the depositors decided to demand their money back, the ṣarrāfs had no choice but to declare bankruptcy. The depositors lost their money and the ṣarrāfs their reputation. 39
In addition to individual capital, the cash reserves of the state were also sometimes put on deposit with the ṣarrāfs. In 1623, the men of prince Shahjahan ordered the ṣarrāfs of Ahmadabad, to remit a sum of money from the provincial treasury (khazāna) of Gujarat to Mandu, where the prince was camping. The order caused a scarcity of money in Ahmadabad, which suggests that money was bodily transported to Mandu. 40 But when hundis were used for transfer, the result could be different. Just a couple of years earlier, the governor of Patna, Muqarrab Khan, delivered cash to the ṣarrāfs to be transferred by hundis to Agra; and money then became plentiful in Patna. 41
The ṣarrāfs accepted cash deposits against bills of exchange (hundi) given to those who wished to transfer money from one place or person to another. 42 Merchants, travellers and money-users, in order to avoid the risks of transportation, deposited cash with ṣarrāfs in exchange for a hundi. 43 Often, the hundi was used for remittance with money changing. In the business of buying foreign species from merchant-importers, such as dollars, ducats, laris, abbasis and kobans, the ṣarrāfs made it attractive for the seller to accept payment in hundis drawn upon the places where goods had to be bought. The agents and correspondents of ṣarrāfs redeemed the hundis in Mughal currencies. The inland remittance of state treasury from the port of Surat, which had a customs house accepting foreign currencies, also involved conversion of foreign and regional monies into Mughal rupees. 44
The hundi brought cash into the hands of the ṣarrāfs with commission on the transaction. Although large sums of money of the Mughal state were usually transferred physically under heavy escorts, as in the case of the prince’s ‘treasure’ sent from Ahmadabad to Mandu, small funds were remitted using ṣarrāf’s hundis. 45 At the district (pargana) level, the revenue collector (karori) was expected to remit money to the treasury by hundi. 46
As bankers, the ṣarrāfs effected book transfer (giro) by maintaining accounts of their clients in which credit and debit entries were made. Depositors often drew upon their bankers to make payments. 47 A petty Mughal official, harried by a dismissed Afghan servant to settle his outstanding salary claims, wrote a note of payment (sanad) drawn on a ṣarrāf. 48 Ledgers thus gave ṣarrāfs considerable leverage in converting cash into credit. In order to avoid a short fall in their cash holdings, the ṣarrāfs tended to delay the conversion of bills into cash. This they did by opening a credit account in the name of the bearer or by making an entry in the ledger against his name if he already had an account with them. This was contrary to the usual practice of converting bill money into cash called anth. 49 The ṣarrāf also created a system of book credit which dispensed with cash. If any one decided to break the network of floating credit by demanding cash, he was paid in coins of inferior value, called kachcha anth, even though his account was kept in newly minted coins of higher value, called pakka anth. 50 The difference in the exchange values of old and new coins determined the rate of anth (Persian ṣīgha i ānṭh. Hindi ankara) and it fluctuated with the availability of money in the market. The bankers often raised the rate of anth, while encashing hundis, when there was a shortage of money (kamī i zar i naqd). In 1714, the ṣarrāfs of Ahmadabad increased the rate of anth deduction to a high percentage, action that drew huge protests from the merchants and brought the market to the brink of a crisis. 51
Since all deferred payments were ultimately settled in cash, the limit of credit could not be stretched beyond a point. The ṣarrāfs with a wider network of exchange had to move cash between places by encouraging people to deposit where it was scarce and withdraw where it was in plenty. 52 The people who deposited cash were those who purchased the ṣarrāfs’ hundis. The people who borrowed money to finance their business issued their own hundis purchased (discounted) by the ṣarrāfs at prices which included interest for the duration of payment. By altering exchange rates of the two types of hundi—bill of remittance and bill of credit—the ṣarrāf could induce merchants to give or take money and improve his cash balance position. If this did not work, for a variety of reasons, the ṣarrāfs transported money physically to their agencies through couriers if the money happened to be in gold, or by cart if large quantities of silver were required to be delivered. 53
Mercantile Credit and Risk Sharing: Ṣarrāfs as Insurers
The ṣarrāfs organised mercantile credit to finance commodity trade and transfer funds. Yet another important business practice of the ṣarrāfs was to provide protection for goods in transit and underwrite risks of losses on capital advanced in loans. Merchants living in medieval times knew only too well the dangers of long distance trade and travel. In its most elementary form, risks could be shared through in-partnership arrangements where two or more persons shared the profits and risks in agreed proportions. 54 In the Mughal empire, partnership (Hindi sājha; Pers. shirkat) was a recognised form of financing and expanding a business venture and the best means to share risks and profits. 55 Similarly, it was not unusual for merchants to acquire cash and goods from their principals to commence trading under commenda, and gains from commenda (muzarbat) were common enough in the seventeenth century to be considered legal and taxable by the Mughal state. 56
In most partnerships, the management of risk was internal to the process of trading. In other words, it was the merchant who had to bear the losses resulting from failures in transit or business. Similarly, the policy of entrusting capital to diverse factors travelling in different ships or caravans was a further step in dividing risks, though it could never secure the total investment of a financier and often enhanced the transaction costs due to increased supervision and accounting. 57 Merchants’ concern for their capital and goods in the face of limited state protection, particularly at sea, required a system of maximum security against risks and uncertainties. The rise of insurance in Florence, Genoa and Venice in disguise in the thirteenth century and then in its true form (third party covering risks on acceptance of premium) towards the end of the fourteenth century provided an effective solution to this problem. 58 Insurance as a viable method of securing trading interests was firmly established in Bruges in the fifteenth century and in Antwerp, Amsterdam and London in the sixteenth century. 59
Inland and Marine Insurance
We get firm evidence of insurance in the Mughal empire only in the seventeenth century but it is not clear whether this was a parallel development or inspired by the trading methods of the Portuguese (earlier Genoese and Venetian) merchants or of the two East India Companies. 60 There is no reference to marine insurance in Indian language records but a description of inland insurance appears in a Mughal history written in Punjab in 1695. In this account, the excellence of India is adduced, among other things, by the fact that merchants and travellers could undertake their journey without fear of thieves and robbers. 61 The author Sujān Rāi described yet another marvellous phenomenon with reference to the work of the ṣarrāfs. This was the practice of conveying ‘goods and merchandise and other property and baggage’ safely (ba salamat) by the ṣarrāfs on payment of a small fee. This practice was called bīma. 62 A reference to a similar practice, called ‘bemah’, and defined as ‘ensurance’, occurs in the correspondence of the English factories at Surat in 1661: Here, the ṣarrāfs not only agreed to transport specific varieties of textiles from Agra, for timely despatch to Europe, but also to finance the whole venture for a fee. 63 A still earlier reference to bīma for insurance occurs in Prince Aurangzeb’s correspondence during 1653–54 in which receipt is acknowledged of 109 pearls sent from Ahmadabad to Aurangabad ‘by way of bīma’, obviously having been insured for their high value. 64
These descriptions point to a system where insurance proper was joined to carriage and, in at least one case, to mercantile credit as well. The ṣarrāfs undertook the actual conveyance of the merchant’s consignment and did not simply bear the risks of loss of consignment. The charges, indeed shown with the cost of carriage, suggested that the insurers were often the transporters themselves. 65 In such arrangements, insurance might be given not only against risks but also against incidence of taxes and transport costs.
In picking up business of this kind, the ṣarrāfs either made use of or encroached upon the territories of a class of people who specialised in undertaking payment of customs dues as well as cartage in return for a consolidated sum. These men, called adhvaya (Gujarati for carter), were employed by merchants to overcome the twin problem of inland trade: availability of carriage and payment of transit dues (rāhdārī), particularly when the goods traversed the territories of local chieftains. 66 The adhvayas were essentially carriers (probably banjaras) who made extra profit out of savings from tax payments while accepting a lump sum from the merchants. 67
Besides this specifically Indian form, there existed a more usual form of insurance, which was devoid of any obligation on the part of the ṣarrāfs to convey the goods themselves. This included insurance of goods and money in transit over land and sea. 68 From the moderate rates quoted for such transactions, it appears that the charges could not possibly have covered the payment of customs dues and cartage.
Given the medieval conditions of overland trade, insurance generally served to protect trade against robbery. 69 The Mughal state, like most medieval regimes, considered maintenance of security on reviles a part of good government and held officials responsible for any crimes committed on routes in their jurisdiction and making them pay for the losses suffered by the victims. 70 The appointment orders of administrative officials mentioned as one of their duties that of securing highways (muhafazat i shahrah) and, in the event of a theft, that of producing either the culprits or the stolen goods. 71 Reflecting on the apparent decline of law and order in the last decades of the seventeenth century, Roques recollected the official measures of protection in the past with some nostalgia. 72
The insurers too employed all possible means to minimise the risks to which the insured values were exposed. One rather unorthodox method was to seek protection from the potential plunderers themselves and, although viewed as unethical, seems to have worked to the benefit of all parties: 73
The rates of insurance premium responded more to the conditions of security than the time and distance required to transport goods from one place to another. This is borne out by the rates quoted at Surat for Broach and Baroda, being 0.25 and 0.62 per cent of the value insured respectively for 1682. Even though Broach was situated half way between Surat and Baroda, the rate for Baroda was more than double as the journey was more dangerous and merchants were constantly reminded of taking special precautions on this route. 74 For Baroda too insurance was given only when merchants agreed to take the longer route through Rajpipla. This was because the insurers knew that the chieftain of this somewhat independent principality always granted a safe passage on payment of transit money. 75
A breakdown in the law and order of various regions in the empire, generally preceded by political instability or a natural calamity, brought the business of insurance to a standstill or induced substantial rise in the premium.
76
In 1694, when the widespread effect of drought, famine and plague led to food riots and high way robberies, the situation posed a problem for both the merchants and insurers:
In Agra etc. the rabble in tumultous manner take corn by force where they can find it without respect to govt. In Guzzarat near Broach have been the like uproars and the Governor of Curkhurred (!) near Ahmadavad was killed on the same account. The roads are so infested with robbers that we have altered our resolution and ordered factoryes to insure all the Comp’s goods, and insurance too is risen cent per cent.
77
If inland insurance was a tricky affair, marine insurance was vastly more complex. Here, too, the ṣarrāfs were major players and provided financial protection against total loss or partial damage of sea going ships and cargoes. The evidence makes it difficult to assess the participation in marine insurance of other mercantile groups, Indian merchants and ship-owners. The references traced so far relate only to the insurance of English or Dutch ships or their cargoes aboard English or Indian vessels. 78 A further limitation is that European sources do not contain the terms of insurance to determine the nature of eventualities under which claims were made and premiums fixed. It appears, however, that piracy, wreckage and other unseen risks of long sea voyages were covered. In one case, the delay in the arrival from the Red Sea of a bullion ship (with a cargo worth one million reales of eight) forced the premium to rise from 3 to 30 per cent. 79 In another instance, the reason ascribed to a 10 per cent premium charged on a Dutch ship was the long voyage it was poised to undertake. 80 The insurers, mostly ṣarrāfs, believed in the risks at sea increasing with the length of the sailing distance.
It is not clear whether war risks were also covered while contracting marine insurance, though it appears from a lengthy dispute between the English and the Surat insurers that this may have been the case. During the first Anglo-Dutch war of 1640–52, the English insured the cargo aboard the ship Supply for 7,500 rupees. After its seizure by the Dutch, the English advanced their claims but the insurers refused to pay on the grounds that the English ship sailed in times of peace between the two companies and was voluntarily surrendered to the Dutch. 81 The ambiguity over the assessment of risks and subsequent claims prolonged the dispute though apparently the insurers did not compensate the English for the loss of their ship—a complex story of a Surat ṣarrāf insuring European treasure rather illegally transported from Bombay to Surat in 1694 can be followed in English records. 82
Credit Insurance
The risk of loss through financial transactions was seldom insured. An exception to this was credit insurance, whereby the insurers covered the creditors for their losses.
83
In 1648, the English factors remitted money from Gujarat to Sind and insured their hundis. The low premium paid by them indicates that commercial papers were exposed less to the hazards of loss than goods or bullion.
84
If such cases were widespread, it would mean that the movement of capital through credit instruments was secured by insurance. The evidence that lenders of risk-sharing loans (called awak or respondentia) too cushioned themselves against loss comes from the description of a risk-sharing commercial practice designated as awak-vyaju:
a transaction in which a person who had made a respondentia advance enters into an engagement with some third person, who for a bonus or stipulated interest, insures him against loss.
85
Insurance became a profitable business for the ṣarrāfs particularly when they were able to spread the risks over various contracts or through reinsurance. 86
A Mughal moneychanger (s.arra-f) plying his trade. Coins of several denominations are on display for exchange including rupees. Bibliotheque Nationale France, Paris, reproduced in Fernand Braudel, The Wheels of Commerce, tr. Sian Reynolds (London, 1982), p. 126.
