The Carolus Premium: When Silver Wasn't Worth Its Weight

On 4 May 1877, the North China Daily News printed its daily commercial quote for two foreign silver dollars circulating in Shanghai. Both coins contained essentially identical silver. One — the Mexican dollar still in active production — was quoted at 76.25 taels per hundred. The other — the Spanish Carolus dollar, not minted since 1810 and gradually disappearing from circulation — was quoted at 83.00. A 14% premium for a coin no foreign state was still issuing, paid willingly by Chinese merchants who knew exactly what silver was worth by weight. At peak, two decades earlier, the premium had reached roughly 30% over silver content in the Yangzi Delta. This is the story of why a coin that should have been worth its weight wasn't — and what that tells us about how money actually emerges.

What the Carolus dollar was

The Carolus dollar is a Spanish silver 8 reales — the piece of eight — minted in colonial mints across the Spanish American empire under Kings Charles III (Carolus III, reigned 1759–1788) and Charles IV (Carolus IV, reigned 1788–1808). The bust-design coins date from roughly 1772 onward — the year a major recoinage standardized the new portrait — and the last Carolus IV issues went to mint in 1810. Physically, each coin weighs roughly 27 grams, contains about 0.78 troy ounces (~24.4 g) of pure silver at a fineness of approximately 0.903, and bears the Spanish king's portrait on the obverse and the Spanish royal arms on the reverse.

Hundreds of millions of Carolus dollars were produced. Many crossed the Pacific via the Manila Galleon (Essay #2). By the early 19th century, they were circulating throughout China — and especially in Jiangnan, the lower Yangtze region centered on Shanghai. In Jiangnan they didn't function as ordinary silver. They functioned as money in a fiduciary sense: accepted by face value among insiders, traded at substantial premium over their bullion content, recognized as a kind of money distinct from other silver. They became, effectively, an unofficial Chinese currency.

The premium

What the data actually shows

The most carefully documented record of the Carolus premium is the daily price quotes published in the North China Daily News, a British-run Shanghai paper that ran continuous commercial market data from the 1860s onward. Warren Bailey and Bin Zhao's 2014 paper Explaining Premiums for Spanish and Mexican Silver Coins in Qing and Republican China (Financial History Review, Cambridge) assembled the NCDN data into the standard time-series for the period:

Date	Carolus premium over bullion silver content	Context
c. 1830	~5–10%	Early phase — premium has emerged but is not yet large.
c. 1840	~18–20%	Premium growing as supply declines and Jiangnan commerce matures.
1855 (peak)	~30%+ in the Yangzi Delta	Period accounts also report ~50% for Carolus over a Mexican Cap-and-Rays peso — but that figure compares two coins, not coin-vs-silver-content.
1860s	~20–25%	Stabilizing, slowly declining.
May 4, 1877	+14.1%	NCDN data: 83.00 taels per 100 coins vs. ~72.74 taels of pure silver content.
1890s	~5–15%	Premium declining as scarcity becomes terminal and Mexican / national-trade-dollar coins take share.
c. 1910	Premium effectively gone	Republic-era currency reforms approaching.

Two different "peak premium" figures circulate in the secondary literature and are easy to conflate. Bailey & Zhao's 1855 reading of the period sources puts the Carolus premium over its own silver content at roughly 30% in the Yangzi Delta. The often-quoted "50% premium" figure also dates to this peak period, but it measures Carolus over Mexican Cap-and-Rays — relative coin preference, not premium over bullion. Both numbers are real; they measure different things.

Why a premium for one specific coin?

Multiple reinforcing reasons converged to make the Carolus dollar a fiduciary anchor in 19th-century Jiangnan:

1. Familiarity and accumulated trust in the design.

By the 1830s, Carolus dollars had been circulating in China for decades. Chinese merchants — especially in Jiangnan — knew exactly what a real one looked like. They knew the weight, the color, the specific way the silver crystallized at the rim, the typical wear patterns. A counterfeit Carolus was easier to spot than a counterfeit of any other coin because the merchants had handled so many real ones.

2. Specific design features that made counterfeiting hard.

The portrait of the Spanish king required substantial die-cutting skill to fake convincingly. The lettering was complex. The edge milling was distinctive. A counterfeiter producing a fake Carolus had to clear a higher bar than one producing a generic silver coin.

3. The 'Four Work' nickname — IIII as gong.

The Roman numeral IIII in the king's name CAROLVS IIII looked, to Chinese eyes, like the Chinese character gong (工, meaning "work" or "craftsmanship"). The coin acquired the nickname Four Work — and this visual mnemonic made the Carolus instantly identifiable to merchants who didn't read Latin script. Cap-and-Rays Mexican pesos, struck after Mexican independence, lacked the visual hook. The nickname is documented in numismatic literature including the Chopmarks.com archive of W. F. Forster's writing on Old Carolus Dollars and is a recurring touchstone of the Chinese collector tradition.

4. Fixed supply, gradually declining.

Spain stopped minting Carolus dollars in 1810. The supply was finite and gradually shrinking — through wear, through occasional melting into sycee, through coins disappearing into long-term storage. Once a fiduciary preference for the coin had emerged, the declining supply fed the premium directly. By the 1850s, the supply was scarce enough that the premium reached its peak of roughly 30% over silver content in the Yangzi Delta.

5. Network effects in Jiangnan's commercial economy.

Jiangnan ran a fiduciary monetary regime — coin identity mattered, and specific coins functioned as money beyond their bullion content. Once enough Jiangnan merchants accepted Carolus dollars at a premium, accepting them at a premium became the rational thing for any new merchant to do. The network locked in. This is the same dynamic that makes any fiat currency stable: it's worth more than its underlying material because everyone agrees it is.

The peak years and the Taiping Rebellion

The Carolus premium hit its peak — roughly 30% over silver — between 1853 and 1856. That window overlaps the most intense early years of the Taiping Rebellion (1850–1864), one of the deadliest civil wars in human history, which devastated the lower Yangtze region — the heart of Jiangnan's commercial economy.

The correlation looks causal — though the connection is inference from the timing rather than a claim spelled out in those terms by Bailey & Zhao. The Taiping disruption broke ordinary commercial settlement networks across Jiangnan. Trust in counterparty arrangements collapsed; standard credit instruments became unreliable. In that context, a coin that everyone in the region knew, that couldn't easily be counterfeited, and that had a fixed and gradually shrinking supply became disproportionately valuable as a payment medium of last resort. The premium peaked just as the war did.

The premium fell off as the Qing government regained control after 1864. By the 1870s, when the NCDN daily-quote series begins in earnest, the premium had stabilized at roughly 14% — high enough to matter, much lower than the wartime peak.

The 1856 paymaster anecdote, examined

The often-told version of this story — including in some popular numismatic accounts — is that US Navy paymasters in Shanghai had to buy Carolus dollars at substantial losses because the local economy required them. The version is dramatic and gets repeated; the contemporary record is more interesting and more inverted.

What the period sources actually show is closer to this: in Shanghai, the local merchant economy did pay a 50% premium for Carolus over Mexican pesos at peak. US Navy pursers, paying salaried officers and crew, refused to extend that premium when paying out — period observers recorded that "Carolus dollars count not more than a Mexican or American with the pursers of the US navy" (documented in the LSE Working Paper 173/13, A Trojan Horse in Daoguang China?, 2013). A salaried Navy officer, paid the standard rate, would find that the dollar he received in pay was worth less in a Shanghai shop than a Carolus would have been. The institutional answer from inside the US Navy was to push back against the local fiduciary preference: pay in standard coin, don't fund the premium.

The contemporary observation — that Navy pursers wouldn't pay the premium even though the local economy did — is more revealing than the inverse. It illustrates the limits of the Carolus fiduciary system: it operated at the level of voluntary local merchant preference, not at the level of universal currency status. A foreign institution with its own monetary commitments could refuse to participate, and routinely did. The premium worked because the merchants chose to honor it; institutions outside that merchant community could decline.

This is the more careful claim than the dramatic version: a fiduciary money emerges from collective merchant preference, but it only operates inside the community that chooses to honor it. Step outside that community — paymaster, foreign government, naval logistics — and the coin becomes silver at bullion value again.

Why this didn't happen in Guangdong

Guangdong — the southern province where Hong Kong is — ran a fundamentally different silver economy: a pure commodity-money standard. Silver was valued by weight and assayed purity, with coin identity essentially irrelevant. Any silver of known weight and purity could enter the system. Heavily chopped, broken, foreign — all circulated together at bullion content. Richard von Glahn's 2007 International Journal of Asian Studies paper Foreign Silver Coins in the Market Culture of Nineteenth-Century China is the standard source for this Jiangnan/Guangdong split.

In Guangdong, a Carolus dollar had no fiduciary premium. It traded at exactly its silver content. The same coin that fetched 30%+ premium in the Yangzi Delta in 1855 traded at melt value in Canton.

This regional difference is one of the most important and least-discussed facts of the period. There was no single "Chinese" answer to what foreign silver was worth. The same coin had different values in different parts of the empire because the parts were running different monetary regimes — one fiduciary, one commodity. (This regional split is one of the recurring threads of the arc; Essay #1 raises it for chopmarks; Essay #5 returns to it as part of the comparative framework.)

Why the premium ended

Three forces converged in the late 19th and early 20th centuries to end the Carolus premium:

Supply collapsed below the level needed to sustain a premium-trading pool.

By the 1890s, the Carolus dollar was scarce enough that not every transaction could find one. Once a coin becomes too scarce to actually use, the premium based on using it has nothing to support. Mexican silver pesos and the new national trade dollars (US, Japanese, British) became the practical alternatives, and merchants gradually accepted them at smaller premiums.

China centralized its silver currency.

The 1933 currency reform under the Nanjing Government abolished sycee and the tael as legal monetary forms, retaining the silver dollar as the sole standard. The broader and more decisive reform came in November 1935, when the Nationalist government nationalized all silver dollars and replaced them with the fabi — paper currency unbacked by silver. By that point, the Carolus premium had effectively ended as a market phenomenon decades earlier; what 1933 and 1935 did was close out the legal context that made the premium possible.

The fiduciary preference itself shifted.

As Jiangnan's commercial economy modernized, the specific reasons that had made Carolus a preferred coin (familiarity, counterfeiting difficulty) lost relative importance. New tools — modern banking, paper notes, eventually a unified Chinese currency — provided fiduciary monetary functions in ways that didn't require betting on a specific historical coin design.

What this teaches

The Carolus premium is one of the cleanest documented historical examples of an emergent fiduciary money — money that's worth more than its underlying material because the economic community has chosen to treat it that way.

Three things stand out:

• Money can emerge without a mint. There was no Spanish state issuing the Carolus dollar by 1855 — Spain had stopped 45 years earlier. Mexico was independent. Nobody was officially backing the coin. The Chinese commercial community decided it was money, and that was enough — at least within the community that made the decision.
• Network effects can make small differences durable. Carolus dollars, Cap-and-Rays Mexican pesos, and US Trade Dollars were all silver of similar weight and fineness. Different premiums emerged for them entirely from network effects in merchant preference. Once a preference locked in, equivalent alternatives didn't displace it — they just slotted in below it in the local hierarchy.
• Fiduciary money has a community, not a universe. The premium worked inside the Jiangnan merchant community. Step outside it — into Guangdong, into the US Navy paymaster's office — and the coin reverted to silver at bullion content. Fiduciary monetary regimes are bounded. The boundaries aren't always political; they can be commercial, regional, or institutional.

The Carolus premium also rebuts a common assumption about commodity money — that silver always traded by weight in commodity-money economies. In some regions and at some times, specific silver coins traded as fiduciary instruments at significant premiums. Commodity money and fiduciary money aren't binary categories. The same metal, in different forms, could function as either depending on which regional regime it entered.