Tuesday, November 27, 2018

Cryptocurrency Collapse

Bitcoin 2-year "price" history
In the second half of last year Bitcoin experienced a massive pump-and-dump. The (heavily manipulated) "price" was pumped from under $2K in mid-July to almost $20K in mid-December. Then came the dump. Hannah Murphy's Bitcoin: Who really owns it, the whales or small fry? reports, based on data from Chainalysis, that in the dump phase:
longer-term holders sold at least $30 billion worth of bitcoin to new speculators over the December to April period, with half of this movement taking place in December alone.

“This was an exceptional transfer of wealth,” says Philip Gradwell, Chainalysis’ chief economist, who dubs the past six months as bitcoin’s “liquidity event”.
This dump drove the "price" down to the mid-$6K region by mid-June, where it stayed until mid-November. But in the last two weeks, things have become "dynamic". At 4pm yesterday on Coinbase the price of a bitcoin was $3718.96. It has lost more than 80% of its value this year. Below the fold, I look at why this might have happened.

David Gerard's The Bitcoin hash rate for the last year, and the squeeze that crypto miners find themselves in is a great place to start looking. Economic analysis of cryptocurrencies starts from the observation that mining is a competitive business, so miners margins tend over time to be competed away, so the cost of mining one coin tends over time to be a bit less than one coin.

Bitcoin 2-year hash rate
Over time, the Bitcoin protocol adjusts the difficulty, and thus the cost, of mining so that the current mining resources on average mine a new block every 10 minutes. The measure of the current mining resources is the "hash rate", and the graph shows that in response to the pump, miners invested in a lot more hardware.

Bitcoin 2-year difficulty
Given manufacturing lead times, the new hardware came on-line gradually over about 8 months. And this graph shows that the difficulty responded, so that despite all the additional hardware a new block was mined about every ten minutes and thus the rate of mining rewards stayed constant. But the price can change much faster than investment in mining resources. David Gerard points out the problem with this :
The price peak was December 2017 — but the hash rate was six times that by August 2018, at one-third of the price.

Mining 1 BTC cost way less than 1 BTC during the bubble. So, to compete, miners built out big.

The hash rate changes approximately every two weeks. But capital expenditure — building and deploying single-purpose mining hardware — has a rather longer lead time.
Bitcoin is a commodity like oil or iron ore, so like other commodities the lead time on investments causes booms and busts. Six times the supply at one-third the price means that miners are being forced out:
The price just crashed, and the hash rate is now dropping off sharply. So everyone’s getting squeezed — the cost of mining 1 BTC is circling 1 BTC, like it was in the doldrums of 2014-2015.

F2Pool founder Mao Shixing estimates 600,000 Bitcoin mining machines shut down in the last two weeks — mostly older, less efficient machines.
The interview with Mao Shixing gives us numbers for how cheap Chinese power actually is. Summer prices are around 2.9 cents per kilowatt-hour. We’re coming into winter, and it’s more like 4.3 c/kWh.

The current worldwide hash rate is 45×1018 hashes per second. If everyone is running an Ebang E-10 (which they’re not — most mining is less efficient), at 0.092 joules per gigahash — then in summer, that’s about $2750 a bitcoin for the electricity alone.

Around now, it’s getting to … around $4000 to mine a bitcoin.
In other words, right now the reward for mining a Bitcoin is less than the electricity costs even if you're using the most efficient miner in the market on cheap Chinese power. Let alone the capital cost of the miner which has to be amortized over say a year, and the other costs such as space, cooling, staff, bandwidth, etc. Gerard writes:
you can’t pay power bills with bitcoins, or Tethers. But this crash looks like selling pressure, with no matching buying pressure — all the suckers from the bubble have gone home now.
This is one of the fundamental problems of cryptocurrencies. Miners need "fiat currency" to pay their bills, so they need to sell their mining rewards. On the other side of those trades there must be buyers, selling "fiat currency" for coins. The buyers have to believe that the "price" of the coins will go up over time. After nearly a year in which the coins have lost 80% of their value this belief is getting hard to sustain.

Until May 27, 2020 there is a constant flow of 75BTC/hour coming in to the market. For the price to average $4K over the next 18 months, speculators have to inject on average $300K/hour in real money into the market. That is almost $3.9B. Right now its hard to see why they would do that. If they don't, the price crash will continue, making it even harder to see why they would pump money in. This is starting to look like a death spiral.


David. said...

Eric Lam's Battered Bitcoin Miners May Start Shutting Down makes the same point about price and hash rate:

"The Bitcoin network’s hash rate, one way of gauging the computing power dedicated to mining the digital currency, dropped about 24 percent from an all-time high at the end of August through Nov. 24, according to Blockchain.com. While the decline may have partially resulted from miners switching to other cryptocurrencies, JPMorgan Chase & Co. says some in the industry are losing money after Bitcoin’s price tumbled."

David. said...

Timothy B. Lee's headline says it all - Ethereum falls below $100—down 93 percent from its January high:

"The monthlong cryptocurrency slide continued overnight with ether, the cryptocurrency of the Ethereum network, falling below $100 for the first time since May 2017. Ether's value is down 93 percent from its January high above $1,400.

Bitcoin reached a 2018 low of around $3,500 last week. It's now trading at around $3,700, down 80 percent from its high of almost $20,000 last December."

David. said...

This year-old tweet from John McAfee is priceless:

"Bitcoin now at $16,600.00. Those of you in the old school who believe this is a bubble simply have not understood the new mathematics of the Blockchain, or you did not cared enough to try. Bubbles are mathematically impossible in this new paradigm. So are corrections and all else"

Tip of the hat to Robin Wigglesworth via Paul Krugman. Last I saw BTC @ $3,743.

David. said...

Jemima Kelly is, as usual, on point with “Cryptoassets” are crashing again. Is it time to start calling them cryptoliabilities instead?:

"If there were any doubt that the Great Cryptobubble of Late 2017-Early 2018 (catchy, we know) had burst, that has surely now gone. Bitcoin was down as much as 11 per cent on the day on Friday, falling below $3,400 for the first time since August 2017. That took its losses over the past month alone to almost 50 per cent."

Kelly takes off from there with a hilarious search for the appropriate name for these "things".

David. said...

The hash rate collapse continues. it peaked over 60M Terahash/s, it is now oscillating around 35M. It first got there in mid-May.

An Antminer S9 runs about 12 Terahash/s, so the peak was around 5M S9-equivalents. Currently about 3M of them. So the equivalent of 2M of them, maybe $2B-worth, got trashed.

BTC is still heading down - last I saw was $3252, around where it was early August.

David. said...

On the one-year anniversary of its "price" peak at just under $20K, BTC is trading just over $3.2K, so it has lost 84% of its "value" in the last twelve months.

David. said...

Nicholas Weaver points out that the falling Bitcoin hash rate is a threat to the cryptocurrency's security. The now-unused hashing power is in principle available to mount an attack. He draws a graph showing that the hash rate is dipping uncomfortably close to the level where 51% of the hashing power available at the peak is now unused and potentially available for a 51% attack. Such an attack could in theory be coordinated through NiceHash. In practice this wouldn't happen until the hash rate was well below that level for various reasons, as Graham Sutherland explains.