Prologue
On May 22, 2010, Laszlo Hanyecz exchanged 10,000 BTC for two pizzas. This day was later dubbed Bitcoin Pizza Day, commemorated annually by the crypto community in a specific way—multiplying the value of 10,000 BTC from that day by the current fiat price to produce a starkly contrasting figure, captioned "the most expensive pizza in history."
This article does not calculate that compound interest.
This article discusses something else: on May 22, 2010, Bitcoin (BTC) was, for the first time, used as a unit to price another commodity in a real transaction. This is the verifiable moment when BTC first functionally played the "unit of account" role among the three functions of money (medium of exchange / unit of account / store of value). This moment sits at the midpoint of a longer timeline of mechanisms—from cost-of-production anchoring (NewLibertyStandard), to P2P matching and discovery (Bitcoin Market), to centralized matching with continuous quotes (Mt.Gox). What Bitbase Research aims to reconstruct is this nine-month prehistory and its comparable forms with the evolutionary paths of price discovery mechanisms for various asset classes throughout history.
Chapter One · What is a "Price"—The Birth of a Unit of Account
1.1 The Three Functions of Money and the Conceptual Boundary of "Price"
The functions of money are standardly defined in financial economics textbooks. Frederic Mishkin, in Chapter 3 of the 13th edition of "The Economics of Money, Banking, and Financial Markets," divides monetary functions into three items [1]: medium of exchange, unit of account, store of value. John Hicks's judgment in the opening of "Critical Essays in Monetary Theory" (Oxford, 1967) is more concise—"money is what money does"; money is defined by its functions, not its physical form [2].
This distinction has specific methodological significance for this article. Something "having an exchange rate" is not the same as "having a price." An exchange rate can be unilaterally published—a third-party observer calculates the conversion ratio between two assets and then announces it; it does not require market participants to conclude transactions based on it, nor does it require others to reference this ratio to price their goods. Price is different. Price is the product of market matching; it requires at least two parties to agree on a conversion ratio for a specific transaction at a specific moment, and this ratio can be recognized, referenced, and reused by third parties.
By this distinction, before May 22, 2010, BTC was in a specific semantic state: it had an exchange rate, but it did not have a market price in the market sense.
1.2 The Functional Performance of "Unit of Account"
Among the three functions of money, medium of exchange and unit of account are often intuitively conflated, but they are not equivalent. The former describes "using this thing as an intermediary in a transaction"—as long as both parties are willing, anything can perform this function. The latter describes "using this thing as the unit for pricing other things." A commodity being used as a payment instrument in a transaction does not necessarily mean it simultaneously functions to price that commodity.
On May 18, 2010, Laszlo Hanyecz wrote in the initiating post of Bitcointalk topic=137: "I'll pay 10,000 bitcoins for a couple of pizzas," and explicitly allowed the other party to order Papa John's pizzas on his behalf [3]. No US Dollar (USD) amount appears in the entire initiating post—he did not say "pizzas worth how many dollars"; he said "10,000 bitcoins for two pizzas." When the completion of this transaction was reported at 19:17:26 UTC on May 22 [4], something happened at the functional level: two pizzas were priced in units of "10,000 BTC."
This is the verifiable moment when BTC was first functionally performed as a unit of account. "Functionally performing this role" and "fully satisfying this function in monetary definition" are two different things. The latter requires this unit to be consistently used as a pricing identifier across groups of traders, a condition only partially met after Mt.Gox's continuous quoting. May 22 is the first time this function appeared in an actual transaction.
Notably, the figure "about $25" does not appear in the initiating post of May 18. Hanyecz gave this fiat anchor for the first time in a subsequent reply in the same thread on June 12, 2010, at 20:14:44 UTC [5]. BTC first assumed the function of pricing pizzas in a transaction; the semantic act of "reverse identification between BTC and USD" occurred 21 days later. This temporal sequence has methodological significance: BTC's functional performance as a unit of account occurred before the bidirectional identification between BTC and USD was completed.
Chapter Two · Mechanism Timeline
This chapter reconstructs the three-tier mechanism evolution from October 2009 to July 2010 in chronological order. Each section provides firsthand timestamps, firsthand post IDs, and a description of the mechanism itself.
2.1 Tier One · Cost-of-Production Anchoring (From October 2009)
NewLibertyStandard (hereafter NLS) published a BTC-to-USD exchange rate formula on its webpage newlibertystandard.wetpaint.com in October 2009. The website is now offline. As of May 14, 2026, the earliest recorded archive.org snapshot is from December 29, 2009, 13:26:10 UTC [6]. The formula on the snapshot reads:
During 2009 my exchange rate was calculated by dividing $1.00 by the average amount of electricity required to run a computer with high CPU for a year, 1,331.5 kWh, multiplied by the average residential cost of electricity in the United States for the previous year, $0.1136, divided by 12 months divided by the number of bitcoins generated by my computer over the past 30 days.
The methodology of this formula is to anchor BTC's value to the "marginal production cost of running a miner"—electricity cost divided by BTC output per unit time. It is essentially an extension of the classical labor-energy theory of value to digital assets: BTC had no market; it was priced by the electric meter. This is a "unilaterally published exchange rate," not a market-matched price.
The first exchange rate figure publicized by NLS, 1 USD ≈ 1,309.03 BTC, has been widely reused in secondary narratives, but there is no original post by NLS himself on the Bitcointalk forum citing this number [7]. This study adopts the archived snapshot data and honestly notes: as of May 14, 2026, this specific number is only available from the first-hand source of the archive snapshot.
Separate from NLS, the earliest recorded BTC-to-USD transaction was conducted by Martti Malmi, an early Bitcoin core contributor—on October 12, 2009, Malmi transferred 5,050 BTC to NLS, received $5.02 via PayPal, and confirmed it in a tweet on January 15, 2014 [8].
2.2 Tier Two · P2P Matching and Discovery (From March 2010)
The second tier mechanism was initiated by dwdollar (real identity Dustin Dollar). On January 15, 2010, at 09:42:18 UTC, dwdollar initiated a motion in Bitcointalk topic=20 msg=100 [9]: "I'm in the process of building an exchange. ... It will be a real market where people will be able to buy and sell Bitcoins with each other."
The platform Bitcoin Market began service on March 17, 2010. dwdollar's update post that day read: "Looks like we had our first real trade around noon!"—this was the first BTC-to-USD transaction matched via a quasi-public order book.
The matching mechanism of Bitcoin Market needs to be stated factually: it was not the complete form of order book matching. dwdollar self-described in msg=265 on February 6, 2010, at 22:37:44 UTC: "ONLY the limit orders work. Market orders will come later." [10]—the platform only supported limit orders, market orders were not implemented; settlement relied on PayPal intermediation, not pure on-chain settlement. Nevertheless, Bitcoin Market was the first platform to publicly display BTC quotes in an order book form; this form itself has mechanistic significance.
Parallel to Bitcoin Market during the same period was P2P physical matching on Bitcointalk. Hanyecz's pizza event is a representative case of this form. The two paths coexisted in parallel from March to July 2010—the dwdollar path proved BTC could be matched with USD to produce quotes, the Hanyecz path proved BTC could be matched with physical goods to produce quotes.
2.3 Midpoint · The Pizza Transaction (May 22, 2010)
Hanyecz's pizza transaction sits at the midpoint of the second tier in the mechanism timeline. Reconstructed in UTC chronological order:
-
2010-05-18 00:35:20 UTC: Hanyecz issued the invitation in Bitcointalk topic=137 msg=1141 [3]. Hanyecz was located in Jacksonville, Florida, USA (EDT, UTC−4), corresponding to the local time of 2010-05-17 20:35. This timezone conversion explains why some secondary narratives date the initiation to May 17.
-
2010-05-22 18:16:31 UTC: txid
a1075db55d416d3ca199f55b6084e2115b9345e16c5cf302fc80e9d5fbf5d48dwas included in block #57043 [11]. -
2010-05-22 19:17:26 UTC: Hanyecz reported completion in msg=1195—"I just want to report that I successfully traded 10,000 bitcoins for pizza. Thanks jercos!" [4]
The 61-minute difference between the on-chain block time and the Bitcointalk confirmation post time is consistent with the narrative of "posting confirmation after pizza delivery." This time difference also independently confirms that the Bitcointalk forum displays timestamps in UTC, not Eastern Time.
The real identity and location of the counterparty jercos are often misstated in secondary narratives. jercos is Jeremy Sturdivant, who stated in a written interview with Bitcoin Who's Who on January 30, 2016: "I have yet to travel outside of the US, and am living on the west coast, near Santa Cruz, California." [12] According to Sturdivant's own account, he is an American, living near Santa Cruz, California, was 19 years old on May 22, 2010, and had never left the United States. The widely circulated description of a "19-year-old British person" is inconsistent with firsthand records.
Hanyecz withdrew the open invitation on June 4, 2010, at 17:51:05 UTC [13], stating "I can't afford that many more." Eight days later, on June 12, he reopened the invitation in msg=1526 and gave the USD anchor for the first time: two pizzas costing about $25, maybe $30 with a tip [5].
2.4 Tier Three · Centralized Matching with Continuous Quotes (From July 2010)
On July 18, 2010, at 01:57:19 UTC, username mtgox posted an announcement in Bitcointalk topic=444 msg=3866: "Hi Everyone, I just put up a new bitcoin exchange." [14] The operator behind this account was Jed McCaleb—founder of eDonkey, the original holder of the mtgox.com domain (an acronym for Magic: The Gathering Online eXchange). The Wayback Machine's earliest recorded capture of mtgox.com is August 17, 2007 [15], about 17 months before the BTC genesis block; this domain was repurposed as a BTC exchange in July 2010.
McCaleb described the platform's quoting architecture in msg=3873 (2010-07-18 02:15:09 UTC) of the same thread [14]: "Last Price ... High ... Low ... Volume ... Current Lowest Buy Price; Current Highest Sell Price ... All trades are between users."—"Last Price," "24-hour high/low," "volume," "best bid/ask"; this set of terms appeared for the first time in the Bitcoin context, marking that BTC now had a standard format for continuous quoting. The mechanistic difference was structural: matching between users rather than intermediary settlement, automatic continuous operation of the order book, 24/7 availability.
From this moment, BTC gained external referentiality in the unit of account dimension—others could say "my thing is worth X BTC," and X BTC had an uninterrupted USD identifier. Mt.Gox was announced to be transferred by McCaleb to MagicalTux (later Mark Karpelès) on March 6, 2011 [16]; its subsequent fate is another independent timeline not expanded here.
Chapter Three · Structural Similarity—The Same Curve Across Three Hundred Years
The proposition of this chapter is: The three-tier mechanism evolution completed by BTC between October 2009 and July 2010—cost-of-production anchoring → P2P matching and discovery → centralized matching with continuous quotes—has comparable structural forms with the evolutionary paths of various asset classes throughout history. This chapter presents two cases for structural comparison and does not construct a quantifiable cross-asset comparison model.
3.1 Amsterdam · The 17th-Century VOC Secondary Market
The Dutch East India Company (VOC) was established in 1602, issuing shares to 1,143 subscribers; share transfers required in-person registration by the company bookkeeper at the East India House [17]. This is generally regarded as the earliest identifiable scenario of an equity secondary market [17].
The maturation of its price discovery mechanism occurred about half a century later. Lodewijk Petram, in his 2011 PhD dissertation "The world's first stock exchange" at the University of Amsterdam, states: the stock market "evolved into a modern securities market" between 1630 and 1650, and "a continuous price discovery process took shape in that period" [17]. Petram's judgment is based on 851 sets of 17th-century stock price observation data. These data were not from official exchange records—neither the VOC nor the exchange systematically recorded prices; data were scattered in merchant correspondence, notarial deeds, and occasional newspapers. The very "dataset" of 17th-century VOC share prices grew from private bookkeeping and oral exchange rates.
The structural similarity with BTC's early path is specific. The NLS formula is to BTC what privately circulated VOC quotes among 17th-century merchants were: unilaterally published, private bookkeeping, lacking matching endorsement. Bitcoin Market is to BTC what informal broker matching within the Amsterdam Beurs was: public but lacking standardized clearing. Mt.Gox is to BTC the starting point of continuous quoting institutionalization—not the end point.
3.2 Chicago · 19th-Century CBOT Standardized Futures Contracts
The second comparison comes from 19th-century American grain trading. The Chicago Board of Trade (CBOT) was founded in 1848, initially as a spot market for commodities [18]. Around 1851, "to-arrive" forward contracts began circulating at CBOT—traders promised to deliver grain at a future date at a specified price [18].
The key institutionalization date for the price discovery mechanism is 1865. CME Group historical archives and the official history of the U.S. Commodity Futures Trading Commission (CFTC) both record [18][19]: CBOT introduced standardized futures contracts in May 1865 and established formal trading rules on October 13, 1865. There is academic controversy over this—one view holds that mature futures trading did not appear until around 1874 [18]. This study adopts 1865 as the starting point and acknowledges this periodization dispute.
The structural similarity between the CBOT path and the BTC path lies in this: the evolution from dispersed bilateral forward contracts to standardized, tradable futures contracts with margin and delivery rules follows the same curve of price discovery mechanism institutionalization. Chicago took 17 years, Bitcoin took 9 months. The difference in time scale does not dissolve the structural similarity. Price discovery in all markets does not start from an exchange; it grows from specific bilateral transactions, oral exchange rates, and private bookkeeping.
3.3 The Explanatory Power and Boundaries of Structural Similarity
Placing BTC's early path within the three-hundred-year sequence of Amsterdam–Chicago–Bitcoin yields a judgment at the mechanism level, not a quantitative model. Three assets, three institutional environments, three technological conditions, yet the price discovery mechanisms all follow the same coarse-grained path: private bookkeeping → publicized quotes → continuous matching. In other words, crypto derivatives infrastructure was not invented out of thin air; it lies on a financial institution evolution line spanning centuries.
This structural similarity argument is not a causal argument. It does not claim "BTC inevitably followed the same path as VOC and grain futures," nor does it claim "all future digital assets will follow the same curve." It claims that: when an observer faces a new asset class, the "prehistory of price discovery mechanisms" is a more enduring subject of study than "asset price trends."
Chapter Four · Acknowledging Limitations and Leaving Gaps
This piece does not expand on contemporary price discovery mechanism comparisons. ETF fund flows, CME open interest, Perpetual Futures funding rates, on-chain market-making—these are the current forms of this 16-year path, but the space and methodology required to expand on them are left for subsequent research.
There are three reverse signals for this article's core argument. First, if academic or on-chain archaeology research reveals an earlier effective entity publishing exchange rates before NewLibertyStandard, the positioning of the "first externally publicized exchange rate" needs revision. Second, if the timestamps of the three core Bitcointalk threads—topic=20, topic=137, topic=444—are proven false at the archival layer, the event anchors need resetting. Third, if the "semantic transition of unit of account" framework is replaced by a more precise monetary function evolution theory, the core argument requires revision.
This article acknowledges the following boundaries. First, "BTC's first functional performance as a unit of account" is an interpretive proposition, not the full establishment of the function in monetary definition. Second, labeling Bitcoin Market as "P2P matching and discovery" is a simplification. Third, the structural similarity argument with Amsterdam and Chicago serves only as a mechanism-level comparison and does not construct a quantifiable cross-asset comparison model. This article does not predict BTC price movements and makes no statements regarding BTC's investment attributes as a single asset.
