By the candle, in London
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
The tea auction began as an East India Company practice: public sales of imported tea in London, held at the Company’s premises and conducted under an old convention in which an inch of candle was lit and bidding on a lot continued until it burned out and went out, the last bid standing. The Company’s monopoly made these sales the single point at which the price of tea for the British market was determined, which gave them enormous leverage and made the timing and size of each sale a matter of national commercial interest. When the monopoly ended in the 1830s the auction did not; it simply became a broader trade institution, and it moved to the street that became synonymous with it. The candle convention is worth more than an anecdote. An auction that ends at an unpredictable moment forces bidders to commit rather than to wait, which suppresses the tactical last-second bidding that a fixed close invites, and variants of it were used for other commodities and for shipping. It is an early and rather elegant piece of market design, and it survived in the tea trade long after most other goods had moved to conventional methods.
Mincing Lane
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
For well over a century, world tea pricing had an address in the City of London. Brokers, warehouses, tasting rooms and the trade press clustered in and around Mincing Lane, and a great deal of the vocabulary a modern buyer meets — the grade abbreviations, the broker’s valuation, the practice of circulating samples in advance — was formalised there. The London auction survived the end of empire by decades and finally closed at the end of the 1990s — 1998 is the date conventionally given — long after the actual pricing power had moved to the countries that grew the tea. Its closure is a good marker for the end of a particular commercial order rather than a change in how tea is sold, because the mechanism it used had already been replicated everywhere else. The physical infrastructure around it was substantial. Bonded warehouses along the river held enormous stocks; brokers maintained tasting rooms with standardised pots, bowls and lighting; and a trade press reported prices and volumes weekly. That apparatus is why so much nineteenth- and twentieth-century tea data exists at all, and it is also why the vocabulary of tea grading is English rather than Chinese, Hindi or Sinhala — the people who named the categories were the people who bought them.
The mechanism, which has barely changed
Made tea is delivered to a warehouse and divided into lots by invoice. A licensed broker draws and tastes samples, assigns a grade, and publishes a catalogue with a valuation against each lot; samples go out to buyers days in advance so their own tasters can assess them. On sale day lots are called in catalogue order and bought by blenders, packers and exporters. The tasting involved is a quality-control operation against a blend specification, not an appreciation — a buyer is asking whether this lot will do the job of the lot they bought last month at the price they need. Most auctions have moved to electronic bidding, which changed the room and not the structure. Timing is part of the mechanism rather than an incidental detail. Sales run on a weekly or fixed cycle, catalogues close on a set day, and a producer’s cash flow depends on getting made tea into the right sale. Unsold lots are withdrawn and re-offered, usually at a lower valuation, and the unsold percentage is watched across the trade as an indicator of market direction. A garden’s production calendar has to be organised around that rhythm, which is one of the ways a sales mechanism reaches back into agriculture.
The auctions that matter now
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
Calcutta is generally said to have held its first auction in the 1860s and Colombo in the 1880s, both well before the independence of the countries they sit in, which tells you they were built to serve the same buyers as London rather than to give producers leverage. Today the significant floors include Mombasa, Colombo, Kolkata, Guwahati, Cochin, Coimbatore, Chittagong, Jakarta and Limbe. Mombasa is generally described as the largest tea auction in the world, handling Kenyan tea along with much of the rest of eastern and southern Africa’s output, and its weekly averages are the closest thing the industry has to a headline price. Anyone who wants to know what is actually happening in tea should read those averages rather than any amount of consumer-facing copy. The move to electronic bidding removed one thing that mattered: the physical room in which a small number of large buyers could see one another, which was an obvious venue for coordination and had long been suspected of being one. Whether it changed outcomes is genuinely debated within the trade. What it plainly did not change is the underlying asymmetry — many sellers of a perishable commodity, a handful of very large buyers, and a fixed weekly deadline that only one side of the transaction has to meet.
What the system does to information
The auction is efficient at moving volume and reasonably good at pricing quality, and it is designed to remove identity. Lots are graded, bought by blenders and combined with dozens of others to hit a house profile that must taste the same next year. By the time the tea reaches a shelf, the garden, the country and frequently the continent have gone from the packet. That is not a failure — blend consistency is exactly what the mass market wants and is genuinely difficult to achieve — but it explains why single-origin, garden-named, harvest-dated tea reads as a speciality product commanding a premium. It is the information the auction system exists to discard. There is a consequence for this catalogue specifically that is worth being honest about. Because the auction discards origin, the entities TeaHQ is organised around — the garden, the flush, the cultivar, the harvest date — simply do not exist as recorded data for the overwhelming majority of the world’s tea. That is not a gap better research would close; it is an absence in the trade itself. Any catalogue that appears to cover all tea at garden level is in fact describing the small corner where such records happen to be kept.
And what it does to the people growing it
A smallholder or an estate selling into an auction is a price-taker in a commodity market, with returns determined by grade, season and demand from a relatively small number of large buyers. Sustained low auction prices for CTC grades have been a documented source of distress in producing regions, and have prompted interventions including reserve pricing and periodic proposals to reform or bypass the auction entirely. Direct sales, private treaty and speciality channels do bypass it, and they are a small share of volume. TeaHQ states the structure and does not quantify it, because the numbers move every season and none has been verified here — but the asymmetry between one seller of a perishable commodity and a handful of buyers is not a subtle feature of the system. Smallholders sit at the far end of the chain: they sell green leaf to a factory by weight, the factory sells made tea at auction, and what reaches the farm is whatever remains after processing and marketing costs several steps upstream of anything they control. Payment is often split between an initial rate and a later bonus tied to the year’s results, which makes income variable and hard to plan around. That structure is the reason auction averages and rural household income in producing regions move together.