In the early eleventh century, a merchant in Chengdu could carry a fortune in his sleeve, as long as the fortune was made of paper. Everywhere else in the world, money meant metal. In Song-dynasty Sichuan it had started to mean a printed sheet with red and black seals, a register number and a promise. These notes were called jiaozi (交子), and most historians of money regard them as the world’s first true paper currency. Heavy iron coins and busy Chengdu markets created the need for them. They began as receipts from private deposit shops, a business later run by a group of sixteen merchant houses. Within a generation the state had taken them over and, in 1023–24, set up the Jiaozi Bureau to print them. It capped how many could be issued, held a reserve against them and gave each issue a fixed life.
This article covers how that happened, why it happened in Sichuan rather than in the capital, and what followed over the next four centuries. That includes the empire-wide notes of the Southern Song and the inflation that eroded them, the paper money of the Jin and Kublai Khan’s chao, which impressed Marco Polo, and the Ming note whose collapse pushed China on to silver. The sources are thinner and more contested than popular retellings suggest, and the disagreements are noted where they arise.
Why Sichuan? A rich province with the wrong kind of money
An iron-currency zone
The short answer is iron. Most of Song China used bronze coins, round with a square hole and strung in hundreds and thousands. After the Song conquered the Later Shu kingdom in 965, it kept Sichuan as a separate monetary zone whose everyday currency was iron. Iron was cheap, so each coin was worth little and large sums meant huge quantities of coin.
The Song dynastic history (Songshi) records an official, Zhao Anyi, reporting in 991 that in Sichuan a single bolt of silk gauze cost 20,000 iron coins. The later reformers Xue Tian and Zhang Ruogu put it in terms of weight. In a memorial preserved by the Song writer Li You, they said ten strings of small iron coins weighed 65 jin and one string of the large coins 12 jin, so a street purchase of three to five strings was already too heavy to carry. A Song jin was a little over 600 grams, which puts ten strings of small coins at roughly 40 kilograms. Peng Xinwei’s Monetary History of China gives similar figures: about 25 jin per thousand large iron coins and 13 for the medium size.
Chengdu’s markets
Heavy money mattered because Sichuan was rich. The Chengdu plain, watered by the Dujiangyan irrigation system, produced grain, silk and tea. A local record attributed to the official Zhao Bian lists a fair for every month in Chengdu, from lanterns in the first month and silkworms in the third to brocade in the fourth and peach-wood charms in the twelfth. Merchants came from outside the province to buy silk and tea in bulk, and they needed a way to hold and move large sums without hauling cartloads of iron. Chinese historians such as Li Ting and Qi Xia trace the jiaozi to this mismatch between a commercial economy and a clumsy currency, and to the credit relationships that bulk trade created.
There was also a precedent. Under the Tang, merchants deposited money in the capital and received certificates to redeem elsewhere, known as “flying cash” (feiqian). The Songshi says the Song notes were probably modelled on it. Flying cash, though, was a bill of exchange between two fixed points, not something passed from hand to hand in the market. The jiaozi would be.
The deposit shops and the sixteen houses
What a private jiaozi looked like
The first notes were private. Li You’s Song chao shishi gives the fullest description. The wealthy houses of Chengdu met at set times to print notes together on a single, uniform paper. The printed design showed buildings, trees and figures. Each shop added its own signature mark and secret codes, printed in alternating red and black as private checks. The amount was written in by hand, with no upper limit. Anyone who deposited coin received a note that was accepted near and far. When a holder cashed a note, the shop kept 30 cash from every string of 1,000, a 3 per cent fee.
Li You also says the houses printed extra batches as the silk and grain harvests approached and used the proceeds to buy shops, land and valuables. That is where the trouble started. The shops were investing money that their note holders could demand at any moment. Li You describes fraud, lawsuits and crowds outside shops whose owners had shut their doors. When officials stepped in, holders recovered only 700 or 800 cash for each string.
When did it start?
The first private jiaozi cannot be dated precisely. The Songshi credits the system to Zhang Yong, who governed Chengdu from 994 and again from 1003. It says he introduced notes worth one string each, renewed every three years, and that sixteen wealthy households ran them. Richard von Glahn, the leading Western historian of Chinese money, dates the first notes issued by Sichuan firms to the end of the tenth century. Some Chinese scholars place the deposit shops earlier still. On their reading, Zhang Yong licensed and consolidated an existing practice, probably in the Jingde reign (1004–1007). Historians also disagree about what to call these notes. Some argue that a receipt redeemable at one firm is really a promissory note, and that paper money proper began only when the state issued it. Others count the private notes, because they already circulated from hand to hand.
The state takes over: the Yizhou Jiaozi Bureau
According to the Yuan-period writer Fei Zhu, Xue Tian first proposed an official note office in about 1016 and received no reply. A later governor of Chengdu, Kou Jian, wanted the notes abolished. After Kou left and Xue Tian took over, he and the fiscal commissioner Zhang Ruogu were told to report on the matter. They argued that abolition would cripple trade and proposed a state monopoly with private issue banned. The court agreed.
1023 or 1024?
This is why books give two dates. Fei Zhu places the founding of the Yizhou Jiaozi Bureau in the eleventh month of the first year of the Tiansheng reign, which fell at the end of 1023. The first official issue ran from the second month of the following year, so the first state notes circulated in early 1024. Both dates are correct because they mark different steps. Documents quoted by Li You show the bureau stamping each note with official seals. Every note had a matching entry in a counterfoil register, so it could be checked and cancelled when redeemed. The state kept 30 cash per string, the same fee the private shops had charged.
Rules of issue: limits, terms and reserves
The official jiaozi looks modern because the state tied its own hands.
- An issue limit. Each issue, called a jie, was capped at 1,256,340 strings. The Songshi and Fei Zhu give the same figure.
- A reserve. According to the Songshi, each issue was backed by a reserve of 360,000 strings of coin, a little under 29 per cent of face value. This was partial backing, and the system depended on holders not all demanding coin at once.
- A fixed life. At the end of each term, old notes were exchanged for coin or new notes. This removed worn and forged notes from circulation and kept the quantity under control.
The length of the term is disputed. The Songshi says three years, with 22 issues in 65 years, while many modern accounts say two. The usual explanation is inclusive counting: a term that ran into a third calendar year was “three years” on paper but a little over two in practice.
Denominations changed too. Fei Zhu says the first notes ranged from one to ten strings. In 1039 each issue was fixed at 80 per cent ten-string notes and 20 per cent five-string notes. In 1068 the balance swung to 60 per cent one-string notes and 40 per cent 500-cash notes, which were far more useful for everyday trade. At about the same time the official Dai Meng won two changes. From 1072, two issues could circulate at once, and the bureau got its own paper workshop.
Fighting the forgers
The Song fought forgery with law, materials and printing. From the start of the Xining reign in 1068, forging jiaozi was punished like forging official sealed documents. When notes spread beyond Sichuan, the Songshi says, knowingly passing forgeries was punished, and so was failing to report a neighbour who did. Privately making note paper earned penal servitude.
Paper, colour and seals
Fei Zhu’s description of the Sichuan notes after 1107 lists six printed impressions, all with decorative borders. Four were in black: an imperial-sanction mark, the series, the term, and a back stamp showing a historical scene. One was printed in blue and a red roundel in vermilion. The bureau employed 81 printers, six engravers and six casters, which suggests cast metal plates. It is one of the earliest detailed records of multicolour printing used for security.
Paper mattered as much as ink. A Southern Song memorial in the Songshi says the early empire-wide notes were printed on fine Sichuan paper that forgers struggled to copy. When later issues switched to a local paper, forgery rose.
The disputed bronze plates
No jiaozi is known to survive. What survives from the Song, or is claimed to, is a handful of bronze printing plates, and these are among the most disputed objects in Chinese numismatics. The National Museum of China holds a plate inscribed “Xingzai huizi ku”, naming the note treasury at the Southern Song capital, Hangzhou. It measures 18.4 by 12.4 centimetres, summarises the law against forgery and has movable type for the denomination and series. The museum calls it the world’s oldest surviving banknote plate. A second plate, known as the “Qian si cang” plate, was long popularly labelled a jiaozi plate. Its text restricts use to places outside Sichuan, which rules that out. Scholars have dated it to the early twelfth century or to the late Southern Song. The historian Gao Congming argues that both plates are twentieth-century fakes from the Beijing antiques trade, while Yang Jun and Zhang Anhao argued in 2023 that both are genuine Southern Song plates. The question is not settled.
From jiaozi to qianyin and huizi: paper money spreads and inflates
The Northern Song overreaches
The Tiansheng discipline did not last. From the 1090s, the Songshi says, issues grew to pay for grain and troops on the north-western frontier. In 1105 other regions were ordered to use a new note, the qianyin (錢引). Sichuan kept its old system until 1107, when its bureau was renamed. That is why both 1105 and 1107 appear as the “end” of the jiaozi. By then issues had reached more than twenty times the Tiansheng quota. At the changeover, one new note was worth four old ones. The Songshi says that in the Daguan years (1107–1110) no reserve was kept and printing went on without limit, and that once the old rules were restored, the notes recovered their value.
The Southern Song and the huizi
After the Jurchen conquest of the north in 1127, the Southern Song made paper a national currency. In 1160 Qian Duanli was ordered to issue huizi (會子) from Hangzhou. From 1162, forging them was punishable by death. In 1168 the government set a three-year term and a limit of 10 million strings per issue. Von Glahn argues that the system worked because officials tried to “maintain equilibrium” (chengti). They collected part of the taxes in paper, used silver to buy up notes when their value fell, and limited issues.
The wars with the Jin and then the Mongols broke those limits. According to the Songshi, the limit rose to 30 million strings per issue in 1195. By 1209 two old notes were exchanged for one new note, and by 1232 two issues together exceeded 329 million strings. In 1247 the fixed term was abandoned. In Sichuan, around 1162, 41.47 million strings of qianyin were outstanding against iron reserves of just 700,000. Yet von Glahn concludes that depreciation did not displace paper. By the Mongol conquest, he writes, it had become the monetary standard of the Chinese economy.
Jin and Yuan: paper money under conquest dynasties
The Jurchen Jin
The Jin dynasty in the north began issuing notes, called jiaochao (交鈔), in 1154, before it cast coins of its own. The Jinshi (History of the Jin) says these notes had a seven-year term and explicitly followed Zhang Yong’s Sichuan jiaozi, adopted because copper was scarce. It dates the Jin’s monetary troubles to the later abolition of that term, after which many more notes were issued than came back. Jin notes warned that forgers would be beheaded and promised informers 300 strings.
Kublai Khan’s chao
According to the Yuanshi (History of the Yuan), Kublai issued a silk-based note in 1260, followed that autumn by the Zhongtong note, two strings of which equalled one tael of silver. In 1276 bronze printing plates replaced wooden blocks. In 1287 a new Zhiyuan note was issued, worth five of the old. The Yuanshi gives annual issues of 73,352 ding (a unit of 50 strings) in 1260 and 1,419,665 in 1276, the year the Southern Song capital fell. Von Glahn notes that the Mongols tried to ban bronze coins and silver from trade and rely on paper alone. A 2024 study by Hanhui Guan, Nuno Palma and Meng Wu traces a path from silver convertibility to nominal convertibility and finally to unbacked currency, with military pressure driving overissue until the dynasty’s final notes of the 1350s.
What Marco Polo saw
Europeans first read about Chinese paper money in Marco Polo’s book. In Henry Yule’s translation, revised by Henri Cordier (1903), Polo says the Great Khan’s mint worked in such a way “that you might say he hath the Secret of Alchemy in perfection, and you would be right!” He describes paper made from mulberry bark, “something resembling sheets of paper, but black”. Officials sealed each note in vermilion, and “Any one forging it would be punished with death.” He was struck by how light the notes were: “ten bezants’ worth does not weigh one golden bezant”.
“Their buying and selling is carried on exclusively by means of pieces of paper, each of the size of the palm of the hand, and stamped with the sultan’s seal.” (Ibn Battuta, translated by H. A. R. Gibb, 1929)
Polo needs careful reading. Yule noted that the tree was probably the paper mulberry, not the silkworm’s mulberry. The sources also conflict on detail: Polo says worn notes were exchanged for a 3 per cent fee, while Ibn Battuta says it was free. Frances Wood’s Did Marco Polo Go to China? (1995) questioned whether he went at all. Hans Ulrich Vogel’s Marco Polo Was in China (2013) answered that Polo’s details about Yuan money, salt and revenue are too accurate to be hearsay.
The Ming baochao and the turn to silver
The Hongwu emperor launched the Da Ming tongxing baochao in 1375. A one-string note in the British Museum is 34.1 by 22.2 centimetres and printed in black on grey paper, with a picture of ten stacks of a hundred coins and three red seals. According to the Museum’s publication Ming China: Courts and Contacts 1400–1450, the text threatens forgers with execution and promises informants 250 taels of silver. The same study found that the paper mixed mulberry, bamboo, rice straw and other fibres. It also points out that the picture of coins was misleading, because the notes could not be converted.
The Ming notes had no reserve, no term and no issue limit. Von Glahn identifies the state’s failure to collect taxes in its own paper as the fatal mistake. The British Museum authors describe the notes as more or less abandoned by the 1430s. Trade, and gradually taxation, moved to uncoined silver weighed by the tael. Traditional accounts date the fiscal shift to 1436, when part of the southern grain tax was commuted into “gold floral silver” (jinhua yin). Noa Grass (Ming Studies, 2017) treats that date as partly an official narrative, and the historian Li Yuan notes that the term’s earliest surviving definition comes from 1582. What is not in dispute is that after the baochao failed, China went centuries without a working paper currency.
Why the jiaozi matters in the history of money
Europe’s first banknotes came in 1661 from Stockholms Banco, founded by Johan Palmstruch. According to Sweden’s central bank, the Riksbank, the background was familiar: heavy copper coinage, then a run on the bank after lighter coins were minted in 1660. In the Riksbank’s words, “The new thing about Palmstruch’s banknotes was that they were not linked to any deposit.” The notes were popular, too many were printed and their value fell. In 1664 the government ordered them withdrawn. The move from heavy base-metal coin to paper to overissue took a few years in Stockholm and about a century in Sichuan.
Von Glahn places Chinese paper money eight hundred years ahead of Europe’s first viable paper currency, the Bank of England notes of the early eighteenth century. He credits Chinese paper and printing, and the nature of Chinese coinage itself: “From the outset, then, Chinese money was fiduciary money whose value was based on trust.” William Goetzmann, author of Money Changes Everything (2016), has described the Song as producing “very clever financial innovations, such as paper money”, while stressing that the state also took over lucrative financial business from merchants. The jiaozi shows both sides. Merchants invented the notes. The state gave them rules and then broke those rules under the pressure of war.
Conclusion
The jiaozi began as a local fix for a prosperous province stuck with a currency too heavy to carry. When private issuers failed, officials chose regulation over prohibition. The Tiansheng rules of 1023–24 set out many of the tools of later monetary management: one authorised issuer, a quota, a partial reserve, a redemption fee, registers against forgery and a fixed life for each issue. The next three centuries show war finance wearing those tools down in the Northern Song, the Southern Song, the Jin, the Yuan and, from the start, the Ming.
No jiaozi survives, the printing plates are disputed, and the sources disagree about dates and terms. The outline, though, is well documented. Merchants and officials in eleventh-century Sichuan worked out how to make people trust a printed promise, and later governments showed how quickly that trust could be spent.
Sources and further reading
- Songshi, Jinshi (juan 48) and Yuanshi (juan 93), treatises on money; Li You, Song chao shishi, juan 15; Fei Zhu’s account of Sichuan notes, in Cao Xuequan’s Shuzhong fangwu ji.
- Richard von Glahn, Fountain of Fortune: Money and Monetary Policy in China, 1000–1700 (1996); “The Origins of Paper Money in China”, in The Origins of Value (2005); “The Rise and Demise of Paper Money in Imperial China” (Hoover Institution).
- Peng Xinwei, A Monetary History of China, trans. Edward H. Kaplan (1994).
- William N. Goetzmann, Money Changes Everything (2016), and his 2016 interview with The Atlantic.
- Helen Wang and Caroline Cartwright, “Paper Money of the Ming Dynasty”, in Ming China: Courts and Contacts 1400–1450 (British Museum, 2016).
- National Museum of China, “Xingzai huizi ku” bronze note plate.
- Guan, Palma and Wu, “The rise and fall of paper money in Yuan China, 1260–1368”, Economic History Review (2024).
- Marco Polo, trans. Yule and Cordier (1903); Ibn Battuta, trans. Gibb (1929); Vogel, Marco Polo Was in China (2013); Wood, Did Marco Polo Go to China? (1995).
- Noa Grass, “Imperial Silver Laundering”, Ming Studies 76 (2017); Sveriges Riksbank, “1661 – First banknotes in Europe”.
