Who Bore the Risk of Japan's Sea Trade?

A document circulating inside Tokio Marine in August 1892 warned about losses at the company's Liverpool agency. Sailing ships were producing an especially troubling relationship between receipts and claims. The proposed response was to reduce the amount the company would insure on their cargoes.^[1]

The men considering the proposal did not need to be aboard a ship for its loss to damage their business. They had accepted obligations that brought distant accidents into the company's accounts. An institution selling protection against the sea had developed a problem with its own exposure.

That difficulty gives a longer history of Japanese maritime trade an unexpected direction. The important change was not simply from dangerous sailing to safer transport. Merchants, lenders, carriers, and insurers developed different arrangements for deciding who would absorb a loss. Following those arrangements reveals how people could enter sea trade without going to sea—and how getting a ship safely home might still leave money owing.

Before his death in 1615, the Hakata merchant Shimai Sōshitsu advised his heir Nobuyoshi against fitting out a vessel and traveling overseas himself. He recommended smaller advances spread among several ships. Nobuyoshi and his son Masanori subsequently lent to merchants conducting overseas trade rather than dispatching ships of their own.^[2]

Their business included nagegane, maritime loans secured against ships or cargoes. If the vessel was wrecked, repayment was waived. A completed voyage brought repayment of principal with substantial interest.^[2]

This joined two services in one transaction. The borrower received money before the voyage, when purchases and preparations required it. The lender also accepted a specified danger: the money might never become repayable. The return on a successful voyage compensated for more than waiting.

The placement of the condition mattered. Under an ordinary debt, the destruction of goods bought with borrowed money does not, by itself, make the debt disappear. Here the bargain deliberately linked repayment to the maritime outcome. A lender could have a claim upon a ship and yet lose that claim when the vessel was lost. Security and exposure existed inside the same agreement.

For the borrower, the arrangement could prevent a maritime catastrophe from being followed by an unchanged repayment demand. It did not make the voyage costless. Money had been committed, goods might be gone, and everyone aboard remained physically exposed. What changed was the distribution of a particular financial consequence between the parties.

Sōshitsu's advice addressed the lender's side. Several smaller commitments could prevent the loss of one ship from consuming the whole amount advanced. This was a way of limiting concentration, not evidence that all voyages had independent fortunes. Ships could encounter the same conflict or disruption; several borrowers could depend on the same market. Dividing advances altered the size of a single exposure without abolishing the wider conditions of trade.

It also separated participation from command. The Shimai could have money at stake without choosing every action aboard the vessel. Their business therefore required judgments about people and voyages they would not personally supervise. Remaining ashore reduced one kind of danger while preserving another: their return still depended on events carried back from elsewhere.

A Portuguese-language document in the Shimai papers, dated 16 October 1633 at Nagasaki, concerns a failure of a different kind. Francisco Carvalho acknowledged an outstanding debt and his inability to repay within the original period. The agreement extended the undertaking and brought in his son-in-law, Sebastião de Almeida, who would repay if Carvalho could not. Three guarantors added their signatures.^[3]

The surviving statement does not identify a shipwreck as the reason for the unpaid debt. That distinction is essential. A maritime loan's release condition did not turn every commercial disappointment into an entitlement to walk away. Insolvency, delay, and a qualifying loss at sea raised different questions.

The new agreement answered inability to pay by changing the arrangements for payment. More time might give a debtor a chance to obtain funds. Another person's commitment could supply a further source of repayment. The people named around the debt mattered because the lender's claim now depended on more than the original borrower's resources.

Yet a guarantor was another promise, not a recovered cargo. Adding a name could improve the creditor's position only insofar as the person could and would answer the obligation. The document records the commitment; it does not certify the later arrival of the money. Its usefulness lies in showing the work required after the original expectation had failed.

Kinship entered that work through Almeida's position as son-in-law. It supplied a relationship through which responsibility could be extended, but the writing made the undertaking explicit. A relative's existence and a relative's agreement to answer a debt were not interchangeable. The commercial arrangement required the second.

The archive itself preserves the consequences of troubled transactions. Fukuoka City Museum explains that the surviving group of Shimai maritime-loan instruments remained with the family because repayment had fallen behind.^[2] They are unusually revealing evidence about difficulty, but a poor basis for calculating how often every loan succeeded. The documents returned by a satisfied lender, or otherwise lost after completion, do not necessarily survive beside them.

What these records establish is a repertoire rather than a success rate. One agreement could remove repayment after a maritime disaster; another could preserve the debt, extend the time, and add people expected to meet it. A merchant's ability to keep participating depended partly on which explanation of nonpayment the parties accepted.

The allocation changed when the provider of transport also undertook to compensate the cargo owner. In the domestic coastal trade of the Tokugawa period, arrangements called kaijō ukeoi joined carriage to a promise concerning cargo losses. Tokio Marine's historical account describes shipowners and shipping wholesalers providing this protection in relation to freight charges, in the trade associated with the higaki and taru cargo vessels.^[4]

The customer was buying a service rather than borrowing the funds needed for a voyage. Protection entered through the contract to carry the goods. The party undertaking to move the cargo also accepted a financial responsibility if the promised service went wrong in a covered way.

That difference changed the direction of the relationship. In a nagegane loan, money moved from the lender toward the trading venture before departure; a specified catastrophe could extinguish repayment. In the carriage arrangement, the cargo owner paid for transport and received a claim against its provider. There was no need for the carrier first to lend the value of the goods to the customer.

Combining carriage and compensation had practical attractions. A customer already needed to choose someone to receive, load, and deliver the cargo. A loss undertaking could become another term of that choice. It brought the protection into an existing commercial relationship rather than requiring a separate search for someone willing to bear the danger.

The combination also left the promise exposed to the carrier's own resources. The same disaster could damage its ship and generate demands concerning other people's cargo. Being responsible for an accident did not ensure that a business retained enough money to compensate every affected customer. The obligation and the ability to fulfill it remained separate problems.

The surviving historical description supplies the form of this service, not uniform conditions for every coastal route. Its importance is the institutional alternative it reveals. Sea risk could be allocated inside a loan or inside a freight agreement. A later specialist insurer would offer a third possibility: buying an undertaking from a business whose principal role in the transaction was bearing the specified financial risk.

Tokio Marine began cargo-insurance business in 1879 and added hull insurance in 1884.^[4] Separating those objects was consequential. The goods and the vessel did not necessarily belong to the same person. A cargo owner could seek protection for merchandise without buying the ship; a shipowner could insure the vessel that carried other people's property.

The insurance premium paid for an undertaking to indemnify a covered loss. The insurer did not have to supply the trading capital or perform the carriage itself. This made it possible to place protection alongside financial and transport arrangements already made elsewhere.

The separation enlarged the possible business, but it also increased what the insurer needed to know. A lending merchant might know a borrower through repeated dealings. A carrier controlled part of the work performed aboard its vessel. An insurer accepting a distant risk through an agent could possess neither relationship in the same form. It had to decide which information would justify which commitment.

The company's overseas growth exposed that problem. Its own historical account dates direct underwriting abroad to 1880 and a further British agency expansion to 1890. Overseas business became a major part of its income; high-risk underwriting contributed to a financial crisis in 1894.^[5]

The 1892 Liverpool warning shows the concern before that crisis. The internal document compared the losses associated with sailing vessels with the receipts from insuring them and proposed tighter limits. Business historian Kamiya Hisaaki reproduces it in his study of competition in marine insurance. He could not establish whether that particular proposal was put into effect.^[1]

The comparison represented a demanding form of institutional learning. A premium received today looked like income, but it accompanied a promise that could become expensive later. Increasing the number of policies enlarged both. To judge whether the business was improving, the company had to connect receipts with the risks accepted and the claims those risks subsequently generated.

That task was different from merely observing that more ships meant more opportunities. A large volume of badly priced commitments could weaken the business providing protection. Conversely, refusing a dangerous contract could reduce immediate receipts while preserving the means to pay other claims. The amount accepted on each vessel became part of the company's own safety arrangements.

Competition complicated the calculation. Kamiya's research identifies intense rivalry in the 1890s and an agreement among four insurers in 1897 to coordinate rates. The arrangement did not permanently stop undercutting.^[6] Insurers were trying to maintain prices for an obligation that customers naturally preferred to acquire more cheaply.

From the cargo owner's side, a lower premium made protection easier to purchase. From the insurer's side, prices too low for the commitments accepted threatened the resources behind the promise. Maritime commerce had gained a specialist risk-bearer, but keeping that specialist dependable became another commercial problem. Distance had moved from the ship's itinerary into the company's organization, its agents, and its accounts.

The Commercial Code promulgated in 1899 addressed an allocation that can initially seem less intuitive than insurance. A person whose property survived could be required to help pay for property deliberately sacrificed to save it.

Article 641 defined general average: losses and expenses resulting from a captain's measures to save ship and cargo from a common danger. Article 642 then distributed the burden among the interested parties according to specified values.^[7]

Consider the logic of jettison—the deliberate disposal of cargo to lighten a vessel. This is an illustration of the rule, not a reconstruction of one unrecorded voyage. The goods thrown overboard belong to someone. The goods saved may belong to other people. Leaving the loss entirely with the first owner would make that person's property pay for a benefit shared by the rest.

General average turned the common rescue into a shared financial obligation. The critical distinction was between an accidental loss suffered independently and a sacrifice or expense incurred for the common safety. Not every item washed away in a storm belonged in the same account as cargo intentionally surrendered to preserve the vessel and its remaining load.

The code made the accounting specific. Its contribution rules referred to the values of the saved ship and cargo, half the freight, and the amount admitted as general-average loss. Subsequent articles specified where and when values were to be assessed and limited liability by the values remaining at arrival or delivery.^[7] The system could not operate simply by announcing that everyone should be fair. It needed a basis for comparing the interests that had benefited and the property that had been given up.

It also treated some interests differently. Article 645 excluded crew wages, provisions, and specified clothing from the contribution calculation while allowing damage to the listed property to be shared by the other interests. The contents of a ship were not one undifferentiated collection of commercial assets.^[8]

Documentation could create a more severe asymmetry. Under Article 646, cargo loaded without bills of lading or other sufficient evidence for valuation could be excluded from receiving a contribution for its loss, while its interested parties remained liable to contribute toward other general-average losses.^[8] The code attached a consequence to the difficulty of establishing a claim's amount. A person could owe into the arrangement without qualifying for the same treatment when making a claim upon it.

Insurance could then sit over this allocation. Article 655 made the insurer responsible for the insured party's general-average contribution, subject to the rule for partial insurance.^[9] Thus a merchant could have intact goods, owe a contribution because someone else's property had helped save them, and seek reimbursement under an insurance contract. Three distinct steps connected the physical rescue to the eventual payment.

The code's provisions do not establish the outcome of every adjustment or lawsuit. They do show how much had to be decided after survival: which action qualified, which values entered, who contributed, and whether a further contract shifted the contribution elsewhere.

The same 1899 code defined marine insurance around losses arising from accidents connected with a voyage. Its general obligation was expressly subject to statutory and contractual qualifications.^[9] A policy was not a promise to make its holder whole after every unfavorable commercial result.

Some boundaries concerned the object and duration of cover. The code distinguished insurance of a vessel from insurance of cargo, prescribed information identifying the ship and ports, and specified when responsibility began and ended. A change of voyage or a substantial change in danger could affect the insurer's responsibility, with exceptions for the circumstances stated in the law.^[10]

These requirements connected the insurer's calculation before the voyage with what subsequently occurred. A commitment made for one undertaking could become quite different if its destination or conditions changed. Equally, a compelled deviation could not always be treated as though the insured had freely selected a new venture. The exceptions were part of defining the bargain, rather than evidence that the bargain had ceased to matter.

Other boundaries concerned the cause of loss. Article 667 excluded specified damage arising from an object's inherent nature or defects, ordinary deterioration, and intentional misconduct or gross negligence by relevant parties. It also excluded ordinary voyage expenses such as pilotage, port, lighthouse, and quarantine charges.^[10] Spending money on a voyage and suffering an insured accident were different grounds for demanding payment.

That distinction returns us to the overdue loan of 1633, though the institutions had changed profoundly. The earlier creditor had to distinguish a waived maritime debt from one still payable by a struggling merchant. The later insurer had to distinguish a covered accident from an excluded cause or an ordinary expense. In both settings, explaining the loss helped determine who would bear it. Neither arrangement relieved every participant of every consequence.

The change across these centuries lay in how that responsibility was organized. A lender could accept the possibility of losing principal. A carrier could include compensation within the freight bargain. A specialist company could undertake defined losses for a premium, while general-average rules distributed a deliberate sacrifice among those whose interests it preserved. Several arrangements could apply to the same voyage.

They also placed different demands on the person expected to pay. A guarantor needed resources when the debtor failed. A carrier needed resources after damage that might affect its own vessel. An insurer needed to preserve funds against commitments accumulated across many ships. The Liverpool warning exposed the last difficulty from inside the institution: the business assuming other people's risks had to limit and understand the risks it had made its own.

Japan's sea trade therefore depended on more than people willing to face danger. It depended on people prepared to specify what they would lose, under what conditions, and on whose account. Those agreements could make another voyage possible. They could also carry the consequences of a distant wreck into an office whose occupants had never seen the ship.