27 February 2014

Bitcoin, the Supernova of Deflation

Wherein we discuss the confluence of quants, computers, macroeconomics, and SQL.

Herbalife is a ponzi scheme, although it remains legal. So is bitcoin. Whether either will survive 2014 is up for debate. For this piece, bitcoin is discussed.

For those who're lazy (or don't like my prose), here's the punch line: should bitcoin's promoters get their way, i.e. bitcoin becomes the reserve currency of record, the world will fall into a permanent deflationary spiral. The sole winners will be the early buyers (Winklevii, I'm talking to you). Just as they are in a Ponzi scheme. There is historic precedent, for those who bother to look. 19th century USofA was a morass of falling prices and wealth concentration, and for the obvious reason; gold standard currency is dependent on native lodes of the metal and appeared only a very few times. The entire amount of gold extracted from the planet for all time has been described as about enough to fill an olympic swimming pool. The Wiki says a 21 meter cube. Near enough.

Here's an academic-ish review. On balance, the 19th century was volatile to a degree no one alive today, or their grandparents, have experienced first hand. But there was more deflation and depression than the 20th by a long shot.

By now, the Mt. Gox fiasco is well known. What isn't known, as I type, is whether the entirety of deposits (said to be 750,000 bitcoin) has vanished. The metaphor for bitcoin, mining, to gold was intentional. The difference is that the amount of recoverable gold is unknown, while with bitcoin it is fixed and published: 21 million of the little boogers, 2.5/minute and soon even less. There are reported to be 12 million already mined. 9 million to go, and a long way to get there: 2140. Unlike gold, which submits to standard mining techniques (pretty much) once located and whose total is unknown, bitcoin mining is intentionally non-linear with time, number of miners, and of a single source. The notion that such a snail's pace of currency generation can support expanding global commerce is somewhere between silly and evil. Making value independent of governments, sovereignty, geology, and place of residence (e.g., external factors which individuals can't easily change) sounds like an egalitarian motive. But the implementation is decidedly autocratic. The gating factor is bitcoins/unit time; this is fixed, until it gets smaller. The reason mining has become so compute intensive is that the gating factor is a constant, irregardless of the amount of mining attempted (the machine increases difficulty as the number of attempts increases to hold static the number of bitcoins released, a pure positive feedback loop); not the move from 50 to 25 per unit time. Moreover, by making the game one of rewarding timeliness rather than actual effort, the game is perverse. Kind of like being born 7 feet tall with a deadly 18 footer; pure chance to get massive reward.

As with the archetype Ponzi, the later you join the harder it is to succeed. Herbalife, but with a much larger cost of entry.

Which all reminds me of the taunt from grad school (voiced by MBA wannabes to the rest of us): "if you're so smart, why ain't ya rich?" Well, here's the other side:
At the Tokyo office building housing Mt. Gox, bitcoin trader Kolin Burges said he had picketed outside since Feb. 14 after traveling from London in an effort to get back $320,000 he has tied up in bitcoins with Mt. Gox.

Bitcoin is a ponzi scheme, and makes little effort to hide this fact. They've just used veiled language. At the beginning, mining was easy with few computers involved, so the originators could make all they wanted. Soon enough, godzilla sized computers are needed. In short order, mining them becomes nearly impossible, on a cost/benefit basis. It's already reached the point that the cost of the electricity alone to run the required computing power exceeds the value of mined coins. The system, by design, adjusts to the amount of mining by making the mining process more difficult. Thus, early miners who were few in number got lots of bitcoin with minimal effort. Since the release rate is what's fixed, more or less, at a point in time, then the effort/reward metric skyrockets as more mining is attempted. It's as if God goes about hiding the gold as more miners enter the lode. How Darwin of him. And so Ponzi. Today's serious miners are betting that bitcoin's value will move inversely to national currency (mostly, the US buck). That deflation will occur globally, and raise the purchasing power of bitcoin. Gold bugs have the same point of view.

The SQL part? MtGox has been hacked with sql injection before. And current reporting says that the missing bitcoins dribbled out over the years, not as a single event robbery. No Ocean's 666. I certainly wouldn't bet against the possibility that the sql injection vector has been ongoing since first admitted.

As to the macroeconomic effect of bitcoin uber alles? The main two are that wealth would concentrate to a degree not seen since the pharaohs, and deflation becomes the norm. Why, one might ask? Well, for the same reason as in the 19th century: with a fixed amount of legal tender, any expansion in output has to be accommodated within that currency store. Price * output == money supply. Price, in the aggregate, has to fall. Falling prices are an incentive to hoard, and thus a disincentive to produce. Death spiral. Of course it can be argued that if output *does* increase (the necessary motive for deflation), then no harm no foul as it's just a wash. History says it's never been a wash. The Fortune X00 are hoarding trillions of $$$ as we speak, CPI/PPI flip-flop around the 0 mark, and we've not yet entered the bitcoin spiral. The issue is that deflation must, necessarily, reduce output increases due to hoarding and wealth concentration.

Since the bitcoin mining supply rule is intentionally asymptotic to 0, and is well into its declining stage due to the release criteria: every 10 minutes 25 bitcoins appear. Since this is not a reward for effort, but for timeliness, the proportional return on effort is inverse to the total amount of mining computer power employed. And, as we know, that has skyrocketed. Just as in real world gold rushes, the more the sadder; all but a handful waste not only time and energy, but funds which could have been put to productive use. One might as well spend the moolah on Powerball. Odds might even be better. In real gold rushes, those that made money were the ones who sold eggs and beer to the miners. All that gold, and only so many eggs and kegs of beer. One might see the correlation in the move from cpu to gpu to asic processing in mining: the ones making the money are the egg and beer guys.

Tulips, anyone?

No comments: