Your ASIC is Probably Mispriced
In our paper (joint work of Aviv Yaish & Aviv Zohar) we argue that ASICs used for cryptocurrency mining have been mispriced since most miners only consider their expected returns and usually ignore the effect of risk and volatility. To properly account for these, mining hardware should be treated as a bundle of financial options, that allow miners to convert electricity to virtual coins. We use option-pricing tools to correctly price ASICs, and compare the result to historical market prices. We find that ASICs were usually overpriced. Contrary to the widespread belief that ASICs are worth less if the cryptocurrency is highly volatile, we show the opposite: volatility significantly increases an ASIC’s value. Finally we construct a portfolio of coins and bonds that provides returns imitating an ASIC, and evaluate its behavior. We find that it historically outperforms ASICs due to higher-than-expected increases in competing hash rate.
In our recently written paper, Pricing ASICs for Cryptocurrency Mining we take a look at ASICs from the perspective of option pricing. ASICs are very similar to European options in which the holder of the option can choose to convert it to some other asset at a predetermined strike price. For ASICs, that convert electricity to coins, the strike price is the cost of electricity.
The implications are that the loss from owning an ASIC is bounded from below, but the upside (in case the value of the coin spikes up) is potentially unlimited — especially if the coin is highly volatile. The figure below shows how Bitcoin’s hash rate reacts (sometimes with delay) to changes its highly volatile exchange rate.
How to Price ASICs
To properly price ASICs we break them down to discrete “mining opportunities” and apply option pricing tools to price each one. The main technique is to construct a risk free portfolio consisting of coins and the ASIC, i.e., one that yields the exact same profit regardless of how the coin fluctuates. See the full details in the paper.
Below we compare the prices of an Antminer S9 on Amazon to prices calculated according to the expected value, and according to our own method. We find that usually Amazon prices are below the expected value, but higher than the value we derived.
Mining Without an ASIC?
An age old question that bugs cryptocurrency enthusiasts is whether you can get the same profits of a miner from just holding the coin itself.
In our paper we construct a dynamic portfolio that simulates the profits from an ASIC but consists only of the underlying cryptocurrency and risk-free bonds. To construct it, we estimate the expected future growth in hash rate based on the rate of historical growth. We find that the rate of growth actually increased (which hurt the profit of physical ASICs, but did not lower the profit of our imitating portfolio). As a result, we see higher profits from the portfolio while initial purchase costs are lower.
The Effects of Volatility
Since Bitcoin’s volatility is expected to decrease as it gains adoption, it is interesting to explore how the value of an ASIC would change if volatility was lower. The figure below shows that volatility accounts for up to 30% of the value of an ASIC when considering historical prices.
Delay and its Consequences
Due to the popularity of ASICs, they are often out of stock, and waiting two to three months for a top-of-the-line machine is not exceptional. When we quantify the loss of value from late delivery of an ASIC we find that it is high. This is due to the fact that there is growing uncertainty about the exchange rate farther into the future.
The figure below looks at the percentage of an ASIC’s value that is lost if it is delivered late. A two month delay can result in a loss of up to 40% of its value.
ASIC pricing, like most of the market for Bitcoin is still not fully matured. As the market becomes more evolved we can expect prices to change. An interesting issue may arise if the volatility of cryptocurrencies decreases, and with them the profits of miners and the coin’s security.
In our full paper we hypothesize that it might be worthwhile to artificially add volatility to miner rewards in order to counteract this effect, and to protect a coin from having its miners migrate to a competing coin with the same hash function.