The Monsoon Ledger: Choosing a Pepper Route in 1230
In 1230 a merchant house in Aden keeps a ledger for two pepper routes. Route A: Malabar to Aden directly, sailing with the northeast monsoon in December–March; freight 4 dinars per bahar; transit 21 days; pirate losses averaged 1 cargo in 12 over the last decade. Route B: Malabar to Cambay, then overland by camel to Aden; freight 6.5 dinars per bahar; transit 55 days; caravan losses 1 in 40. The firm's capital is 30,000 dinars, and it must return cash to investors within one sailing season. Last year a war between two Malabar kingdoms closed Route A's ports for six weeks, and the firm lost a shipment. The senior partner argues for Route B on safety; the junior partner argues for Route A on cost. Both cite the same ten years of records.
1. Assume a bahar of pepper sells in Aden for 14 dinars. Convert both routes into an expected cost per bahar delivered, and state which route the arithmetic favors. Then explain whether the arithmetic should settle the decision.
Read the worked answer
Route A: 4 dinars freight + (1/12 × 14 dinars of lost cargo) ≈ 4 + 1.17 = 5.17 dinars, leaving a margin near 8.83.
Route B: 6.5 + (1/40 × 14) = 6.5 + 0.35 = 6.85 dinars, leaving 7.15.
On expected value Route A wins by about 1.7 dinars per bahar, a margin roughly 19% better, and it also turns capital around in 21 days rather than 55 - decisive given the one-season repayment rule.
The number should not settle the decision. The 1-in-12 loss rate averages ten years of what were presumably independent accidents: one ship, one storm, one pirate squadron. The six-week closure was not that kind of event. It was correlated - it hit every consignment simultaneously - and it can repeat. Pricing a systemic risk as though it were a coin flip systematically understates the danger to a firm whose entire capital must return in a single season. The arithmetic favors A; the risk structure may not.
2. The senior partner says the war was a one-time event. Using the idea that some risks are correlated rather than independent, and the constraint that capital must return within one sailing season, argue what a prudent firm should actually do.
Read the worked answer
The claim to test is 'one-time.' What produced the closure? A war between two Malabar kingdoms over port revenue - a political conflict, and political conflicts recur on a scale of years, not decades. So Route A's loss history is not a stable probability, and the firm cannot treat last year as an outlier drawn from the same distribution as storms.
With 30,000 dinars due back on a fixed date, the firm's real exposure is ruin, not a slightly thinner average margin. A prudent response is to stop treating the choice as binary:
- Diversify across routes, sending most capital by Route A in the early monsoon window and a smaller share overland, so a closure damages a portion rather than the season.
- Write the risk into contracts - a commenda-style partnership spreading loss across several investors, or a freight contract paying the captain only on delivery.
- Hold a cash reserve sized to survive one lost consignment.
The judgment being tested is that contract design and diversification, not a better average, are the tools for correlated risk.
3. A historian calls the Indian Ocean 'one integrated market.' Give one element of the ledger that supports the claim and one that challenges it, then explain what evidence would actually settle the question.
Read the worked answer
Support: the ledger prices the same commodity toward the same destination along two different paths, and freight differentials (4 vs 6.5 dinars) are legible in both. That is a price gradient connecting Malabar, Cambay, and Aden - a market transmitting signals along a route is at least loosely integrated, and merchants arbitraging between the legs are the mechanism.
Challenge: the two routes are not equally available at all times. The war shows that political authority, not price, decides whether the sea leg exists at all. The overland alternative is not a substitute at equal cost; it is a rationed fallback. Integration is therefore seasonal and conditional, and a single firm's costs cannot distinguish 'one market' from 'several markets connected intermittently.'
What would settle it: weekly price series from multiple ports, not one firm's books. If Aden and Cambay pepper prices move together within a sailing season and diverge sharply when a strait or port closes, the market is integrated in the economist's sense. One ledger can suggest integration; it cannot demonstrate it.