Brian: Based on the survey, what were some of the most notable geographic trends in Bitcoin mining?
Apolline: I find the cost structure breakdown by geography particularly interesting. This year is the first time we were able to provide that level of granularity thanks to a larger mining sample. Regarding miners’ cost structure, it was interesting to see general assumptions being confirmed, such as North American miners allocating a greater share of their expenses to capital expenditures compared to their Asian counterparts (52% vs. 37%). This is even starker when focusing exclusively on responses from Chinese and US hashers, whereby the average Chinese miner allocate about 30% of their costs to acquire machines, vs. 56% for the average US miner. Although we do not have absolute figures, having informative underlying data to validate anecdotical evidence is another step towards greater transparency/clarity.
Quite a few people were also surprised to see that Latin American miners reported the lowest share of utilities cost as part of their total cost structure and the lowest median electricity price ($0.025/kWh vs. $0.046/kWh for the global average). The clustering of respondents at the bottom part of the distribution for Latin American miners suggests that a select few have access to very cheap electricity. Some even reported access to nearly free electricity!
Brian: What changed versus previous surveys and what do you think the trends you've noticed can tell us about where hashrate might disperse over the next 1-5 years?
Apolline: Couple of things changed since the last survey which mostly relate to the larger pool of respondents we obtained for the mining survey in 2020 allowing for a more accurate geographic distribution. We were able to provide aggregate findings on areas such as cost structure, hedging strategies, etc. These topics were covered in the survey from the previous years, but we did not collect enough responses to disclose the results. Another research project we worked on, the mining map of Cambridge Bitcoin Electricity Consumption Index, shows that the growth in hashrate that we would ordinarily expect in China seems to have relocated in other non-Chinese regions between late 2019 and mid-2020. This trend will probably continue, but the pace of it remains to be seen. We might also observe a change in the distribution of hashrate across Chinese provinces, if the seasonal migration of miners stops.
Brian: What trends are you seeing in terms of composition of energy consumption? Are miners increasingly using renewables?
Apolline: The samples from 2018 and 2020 are quite different and do not allow for a complete like-to-like comparison. The 2018 sample found that on average 28% of mining comes from renewables vs. 39% in 2020. It is certainly an indication that miners are increasingly relying on renewables, but it might be accentuated by sample bias we cannot control for (because samples are still too small). But generally, a meaningful part of mining comes from renewables, primarily hydropower (62% of miners indicated using hydroelectricity). Unsurprisingly, apart from hydropower, the use of other sources of energy (e.g. coal, wind, natural gas) vary quite a lot between regions. This distribution might evolve in the coming years. For instance, in Asia-Pacific, an equal share of miners reported using hydropower and coal (primarily in provinces like Inner Mongolia, Xinjiang, or countries like Kazakhstan) in 2020. But if, as some Chinese miners indicated to us in separate interviews, it is no longer profitable for them to move to Sichuan or Yunnan during the rainy season, they would most likely stay in Chinese provinces that are mostly producing coal.
Brian: What can you tell us about how Bitcoin miners manage their reserves? For example, how often do miners hold bitcoin versus immediately selling to capture the dollar spread?
Apolline: There is not sufficient reliable data from our survey to fully address this question. The pro-cyclical effect of miners and inventory management are certainly very important topics that have been understudied as of now. Apart from anecdotical accounts, I am not sure I have come across any comprehensive study that looks at it in details. This is certainly an area we’d like to investigate further in our future studies.
Brian: Do miners take out loans against their bitcoin instead of selling? How often do miners use derivatives to manage risk? Do you think miners can gain an edge using derivatives versus those that don't?
Apolline: Survey data suggests that 15% collateralize their cryptoassets to mitigate risks. I thought this figure would be slightly higher, but I guess my perception was biased by exposure to mostly Western miners. With regards to derivatives, although the use of financial products has been widely discussed throughout 2020, data suggests that only a handful of miners make use of them. The vast majority of miners rely on rather “elementary” strategies to mitigate risks, like holding fiat or cryptoasset reserves. I guess there is a place for derivatives, but a few pending questions would need to be solved/answered (maturity, valuation, etc.), but also greater regulatory certainty around these financial products is needed.
Brian: Are there any other quick mining highlights you found particularly interesting?
Apolline: A finding that was particularly unexpected was the share of miners that receive support from governments or local authorities (23% of surveyed miners). This aid primarily takes the form of locally-focused support, such as electricity price subsidy for users within a region. Of this 23%, 38% of them operate in China, 19% in Kazakhstan and 12% in Canada. I guess that was quite surprising given miners’ perception of governments. For instance, miners in Asia-Pacific have ranked “government seizure of mining facility” as their top concern, along with regulations.