How Something With No Price Acquires One
Before there is a market there are two people arguing about what a fair swap looks like, and everything else grows from that.
“The first price” is a phrase that appears in almost every history of this subject and it is a category error. A price is an output of a market. Before there is a market there are no prices, there are trades, and a trade is an anecdote about two people.
The distance between those two things is where most bad writing about this period lives. What follows is the sequence by which a thing with no price acquires one, in four stages, with a primary document sitting at the second stage. There are no numbers in this piece, and the last section explains why that is a decision rather than a gap.
Stage one: two people and an argument
Somebody holds a thing. Somebody else wants it. Neither has a reference point, because no reference point exists. What happens next is a negotiation with no anchor, and the anchors people reach for in that situation are revealing.
The commonest is cost. What did it take to make this. That is not a valuation argument and it does not become one by repetition, but it is available, it is checkable by both parties, and it produces a number both can reason about. For a thing produced by leaving a computer running, the cost is electricity and time, and both are quantities somebody can look up.
The second commonest is a swap against something whose price already exists. Not “what is this worth” but “how many of these for one of those”, which is a narrower and more answerable question, and it is how the first trades in almost every new instrument get done.
The third is nothing at all. Two people pick a number because they both want the trade to happen. This is more common than the tidy versions admit, and it leaves no trace distinguishing it from the other two.
A trade at this stage tells you that on one occasion two specific people agreed on one specific ratio. It tells you nothing about what a third person would have paid, nothing about whether a fourth could have sold at that ratio, and nothing at all about a quantity larger than the one that changed hands.
Which anchor was used in any given early trade is usually unrecoverable, and that is the general problem with this period rather than a problem with prices. Where somebody happened to write their reasoning down as they went, as one early participant did, you can read it. Where they did not, what survives is a number and a story attached to it afterwards by somebody who knew how it turned out.
Stage two: a posted quote
The next thing to appear is not a market. It is one party publishing a number and standing behind it.
Here is one, archived on 29 December 2009. It is the exchange rate page of a site calling itself New Liberty Standard, and it does three things a bilateral negotiation does not: it publishes a rate to buy, a rate to sell, and the method by which both were derived.
The method is the reason to read it. Quoted in full, and note that not one input is a trade:
Our exchange rate is calculated by dividing $1.00 by the average amount of electricity required to run a computer with high CPU for a year, 1331.5 kWh, multiplied by the 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.
A national electricity statistic, a national electricity price, and one person’s own production over the previous month. The doubled “the the” is in the original. The operator adds that he has ordered a single outlet electrical meter and will probably change the calculation to an average production cost over a year, starting on the first of January.
The rest of the page is equally instructive about what stage this is. To transact you emailed him. Payment was accepted by one third party payment processor and no other. And:
The rate changes at midnight Greenwich Mean Time regardless of whether the new rate has been posted. The time at which the financial transaction begins determines the rate for the transaction.
That sentence exists because a posted quote is a commitment by one person, and one person needs a rule for when their own commitment expires. A market does not need such a rule. A market has a next quote.
So what stage two adds is not price discovery. It is availability and commitment. Anyone can now see a number without knowing anyone, and one party has said in public that they will transact at it, in some size, subject to their holdings and their patience. That is genuinely new and it is not a market price, because the number was not produced by anyone’s willingness to pay. It was produced by a division.
The page carries a buy rate, a sell rate, an available balance and a history table going back weeks. I have all of it. None of it is here.
A quoted rate that nobody transacted at is not evidence about price, it is evidence about what one person was willing to state. Reproducing it in a history would launder a calculation into an observation, and that laundering is precisely the error this piece is about. The methodology is quoted because the methodology is the finding. The output of the methodology is not a finding, it is an arithmetic result whose inputs are printed right next to it.
The same reasoning applies to every early figure in this subject, including the famous ones, and it is why this site does not print them.
Stage three: a venue that matches
A venue is different in kind from a posted quote, and the difference is that the venue does not have an opinion.
It accepts an order to buy at a limit and an order to sell at a limit from unrelated parties, it holds them, and it matches them when they cross. The number that results is not anybody’s calculation. It is the point at which one person’s willingness to pay met another person’s willingness to accept, and neither of them had to find the other, negotiate, or trust a stated method.
Three things arrive at once with the venue and it is worth separating them.
A continuous number. There is a last trade, and then another one, and the sequence is a time series rather than a set of anecdotes.
A depth behind the number. The resting orders on either side say something about size, which no single trade ever does. This is the first stage at which “the price” is a claim about more than one unit.
A single point of failure. Everything the venue reports is a report by the venue. This site is going to spend an entire era on what that turned out to mean.
Stage three is also where the reader’s intuitions from ordinary markets start to apply, and where they start to mislead, because a thin venue’s number looks exactly like a deep venue’s number and behaves nothing like it.
Stage four: a reference rate
The last stage is when other people start quoting a venue’s number as the price, rather than as that venue’s price.
A shop pricing goods, a second venue setting its own opening quote, a newspaper printing a figure: each of them is treating one venue’s matching engine as an oracle about a thing that has no location. This is the stage at which a price stops being an observation and becomes an institution, and it is also the stage at which the circularity begins, because the second venue is now quoting the first, and the first is being fed by traders who read the second.
Nothing in the sequence rules that out and nothing corrects it. A reference rate can be a faithful summary of deep two sided trading, or it can be a number that several venues are copying from each other, and from the outside those two look identical.
Four sentences that get written and should not be
“Bitcoin was worth X in year Y.” Almost never a supportable sentence about the early years. Worth to whom, in what size, at which venue, and was anybody on the other side. The supportable version names the venue and the date and says what it is, which is one venue’s quote.
“The first exchange rate was set at X.” A rate somebody calculated is not a rate somebody paid. The New Liberty Standard page above is a calculation, and it says so, in a sentence anybody can read.
“Early holders got in at X.” This treats a giveaway, a mining reward and a purchase as the same event. The distribution mechanisms of the period handed units to people at no price at all, which is a different fact about a different thing.
“The price was X, which shows the market believed Y.” A price is not a belief and a thin price is not even a price. This construction takes an anecdote, promotes it to a measurement, and then reads a collective mental state off the measurement.
What was expanding, what was contracting
What expanded, across these four stages, was the number of people who could act on a number without knowing the person who produced it. That is the actual achievement of a market and it is a social achievement before it is a technical one. At stage one you needed a counterparty and a conversation. By stage four you needed neither, and the intermediate stages are the machinery that removed them.
What contracted, at every single step, was how much you could tell about where the number came from. The bilateral trade is fully legible and useless. The posted quote publishes its method and is not a market. The venue produces a real price and shows you almost nothing about how. The reference rate shows you nothing at all, and it is the one that ends up in the history books.
Who could tell at the time is the answer this whole piece exists for. The people transacting could tell, perfectly, because they were the market and they knew exactly how thin it was. Everyone downstream lost that knowledge at each stage, and the numbers that survive from this period, the ones repeated in every account, are stage four artefacts describing stage one events.