On The Bitcoin Halving
What effect will the 2016 halving of bitcoin mining rewards have on the cost of bitcoin?
You'll perhaps remember that the amount of bitcoin released to miners as reward for settling transactions has been preprogrammed to cut in half every four years approximately, until the eventual cessation altogether of newly introduced coin at some point around 2140. For more detailed information, check out our introduction to bitcoin series.
At first glance, the bitcoin halving would not seem to have much effect on the cost of bitcoin. After all, there are currently close to 15 million in circulation, with approximately 1%, or 150,000 coins, changing hands daily at the exchanges. What difference does it make then if the daily amount added to such volumes is 3600 bitcoin or 1800 bitcoin?
Tim Swanson, the astute bitcoin critic best known for selling perpetual-motion shared databases branded as "permissioned blockchains" to banks, sums up this argument best. He explains that only the future money supply is impacted by the halving, and thus has zero effect on the existing monetary supply. In other words, it is foolish to expect a diminishing future supply to have an effect on the present-day cost of bitcoin. Only increased demand or a considerable drop in current supply would have such an outcome.
He goes on to argue that the speculative nature of the market, in which people are not actually buying bitcoin to spend but to hoard, undermines upward demand pressure as transactional demand which would be generated by a vibrant bitcoin merchant economy is basically nonexistent. Thus, speculative demand undermines the transactional demand, on which it is actually dependent for success.
The alternative viewpoint suggests that current demand is fully absorbing the newly introduced 3,600 coins daily, whilst maintaining, and even gaining a bit, in price. Bitcoin is currently in its least volatile period historically (well, up until press time at which point it has risen about 70% in a few weeks), which perhaps suggests some sort of equilibrium. See below two horribly designed charts, the left showing transactional volumes over the last six months, the right showing price over the same period.
On the price side, we see a trend line with a slight upwards bend, but notable for impressive flatness. On the transaction side we see massive growth which tells us...well, nothing definitively. It is incredibly difficult to extrapolate hard and fast meaning from transactional amounts. A transaction could mean a big holder splitting up one wallet into a dozen; it might show an exchange moving coin between its hot and cold wallets, it could show joe shmo moving .001 btc from primedice back to his wallet - could be anything really. We also know that transactional volume between traders within the exchanges is increasing as well - but who knows how much of that is due to, for instance, the emergence of leveraged trading, or the use of volume bots by sneaky exchange operators to prop up their volume statistics.
However, the substantial increase in transactional volume is at least a pretty strong indicator that activity in the bitcoin space is increasing, and the flatness and growth in price shows that the ecology is absorbing healthily the additional 3,600 newly minted coins each day. Or in other words, demand is keeping up with the marginal increase in supply. But what actually is the real supply?
The question of actual supply brings us tidily to the bitcoin hoarding phenomenon. Mr. Swanson's description of the speculative nature of the market is particularly relevant to this discussion. 70% of all bitcoin has been held for more than 6 months without moving, with 30% falling into the so-called "zombie bitcoins" category, or those that have not moved for more than two years. Please see below a graphical illustration of the amount of time bitcoin has been held, with thanks to John Radcliffe. Orange and above is 6 months or more, the pukish green two years or greater.
It is reasonable to assume that these 70% of bitcoin are held for speculative purposes. Of the zombie bitcoins a great deal of the coin, perhaps the vast majority, are simply lost forever, visible in public addresses but out of reach. These coins were mined and earned in the very early days when bitcoins were worth fractions of pennies, and so were treated pretty much the way you'd treat a handful of pennies. However, very notable wallets are included in this zombie category, including the wallet of Satoshi Nakamoto himself, who accumulated almost one million bitcoins and has never moved a single one since. It would indeed be tragic if the father (or mother) of bitcoin lost the private key to his stash; it is far more likely that he/she not only is bullish about its future, but wants to avoid flooding a market not yet ready for flooding.
According to Mr. Radcliffe, who popularized the term zombie bitcoins with a series of fascinating articles, 10% of the bitcoin in circulation are involved in a transaction monthly - approximately 1.45 million bitcoins. If we take this number at face value, it means that a full 48,000 bitcoins are traded per day - meaning that the effect of a halving from 3600 to 1800 bitcoin, while not insignificant, does not represent a great market-moving shift.
However, this 10% number comes with caveats. Mr. Radcliffe explains that the entire bitcoin holding of any particular wallet is counted in this number, regardless of the actual amount moved. So, for instance if a wallet with 20 bitcoin spent or otherwise moved 0.5 bitcoin, the entire 20 bitcoin would could towards this 10% figure. And just as importantly, the bitcoin involved could just be moving from one wallet to another - meaning it does not necessarily represent bitcoin available for purchase.
So we have no idea of the actual amount of bitcoin available for sale, only that the upper threshold of the daily supply is 48,000, the lower threshold is 3,600, and the actual amount fluctuates somewhere in between. In this author's opinion, the actual number is much closer to the lower threshold, to the extent that the mining reward released daily into the ecosystem accounts for a substantial percentage of supply.
So why is all of this important, aside from the obvious desire to see bitcoin price increase?
From a network perspective, it is pretty crucial that there is a considerable appreciation in the cost of bitcoin around the time of the halving. The amount of bitcoin rewarded to miners, which is used to fund the mining operations necessary for efficient and decentralized settlement of bitcoin transactions, will be cut in half overnight. This will theoreticall be offset by a price increase, whatever that may be. If this is not sufficiently offset, we will likely see the most inefficient miners ceasing operations leading to greater centralization, less hashing power, and less network security.
This is the point Mr. Swanson is getting at in his article, which reeks a bit of concern trolling. In his introductory paragraph he poses, and then proceeds to bash, the proposition that "[bitcoin advocates state] the market price is supposed to immediately double". He then concludes with a quote from Jonathan Levin of the now non-existing website Coinometrics, stating "the hashrate to halve as the reward halves. Essentially the argument would go if the price has not doubled the hashrate must fall". So basically, if there is not an immediate doubling of price, the hashrate will halve. I suppose this is hyperbolic flipside to his introductory strawman. In any event, we would assume the halving will cause a gradual increase in price: much of it prior to the actual event due to market speculation, with additional market movement in either a positive or negative direction coming after the event, as the actual effect on supply is felt by the market.
An additional risk, of course, is that there is a zombie coin holder patiently waiting for just such an event to unload a hoard of coins at increased profit, and that the market will be unable to absorb these coins, leading to a large price drop.
Bitcoin Price (USD): 853.48