Bitcoin Mining Revisited: Asymmetric Risk in Perpetual Mining Bonds

Those selling Bitcoin PMBs (perpetual mining bonds) or offering ‘mining contracts’ directly to the public employ a business model in which all risk lies with those putting up the cash. Normally, those taking on exceptional levels of risk might expect to be compensated with exceptional levels of return, but in this case that seems unlikely in the extreme.

It’s now been over a year since Butterfly Labs starting accepting millions of dollars worth of pre-orders for dedicated ASIC-based Bitcoin mining hardware, and many would-be miners have begun celebrating the fact that several weeks’ worth of some types of devices (and a far smaller number of other devices) have finally been delivered. However, there is not yet any direct evidence that the thousands of remaining pre-orders will be filled in a timely fashion. Rather, most evidence seems to suggest the company is unlikely to reach its forecast hardware production rate of 400 units per week — or was it per day? Or does it even matter? As I outlined in an earlier article, every single day of delay in the delivery of Bitcoin mining hardware reduces its value, pushing it one step closer to its virtually guaranteed destiny as nothing more than a glorified space heater. (See “The Economics of Investing in Bitcoin Mining”.) The story is similar for other major sellers of ASIC-based mining hardware (KnCMiner, Avalon, etc.), where either nothing has been shipped at all, or only a relatively small number of customers have received hardware.

Yet many investors and other hopeful members of the public have been eager to pay third parties to place new pre-orders for this type of hardware, on the understanding that the third parties will eventually mine with it and share the proceeds. Because this is hardware with an unknown delivery date which will eventually be operating in a context of currently unknown network power, it offers an unknown return on investment, a return which may wind up being literally nothing at all. On Bitcoin exchanges, these arrangements are referred to as PMBs, or perpetual mining bonds — where the word perpetual denotes a loan that never has to be paid back. Investors essentially loan money to the PMB issuer forever, and the issuer in theory uses the money to acquire mining hardware. Eventually, one day, maybe the hardware is delivered, and creditors then receive a share of the mining proceeds. To be fair, some PMBs are at least partially backed by mining hardware that actually exists, and the issuers are tapping the markets in order to expand their operations; in other cases, it does not yet exist, and nobody knows when or whether it will ever come to exist.

In another variation on the scheme, the site CloudHashing.com sells one-year or two-year “mining contracts” to members of the public and plans to use the money received to begin mining…one day. For the proprietors — private UK company Technology IQ Ltd., which came into existence 4 months ago — it seems to be going very well. Back on 3 April 2013, they claimed to have 4.5 terahashes of mining power on pre-order, and they had sold 3020 out of 4500 maximum available contracts which would eventually share that power. Just 10 days later, they listed 6 terahashes of power on pre-order, with 5249 contracts sold out of a maximum of 6000 available. Since then, Cloud Hashing’s claimed sales have exploded well beyond the million dollar mark, with the company now showing over 10,000 contracts sold from a total 39,945 which they intend to sell.

As with many such schemes, actual mining is going to commence…real soon now. Maybe it was June, then it was July. Fast forward to today, and now official word is that mining might commence in September. (If you think a couple of months of moving the goalposts is no big deal, check the slope of the curve showing total network hashing: the higher the graph goes, the less profitable mining becomes for a given amount of computing power.)

The CloudHashing.com site includes some rosy projections about expected return from their different variations on mining contracts, with the company suggesting that people might reasonably expect to receive a return of upwards of 500% of invested capital just in the first year of mining. Unless the Cloud Hashing folks have discovered a way to jump the queue and nab their pre-ordered mining hardware before mountains of other mining rigs hit the network, the mathematician in me views these projections as altogether ludicrous. But crucially, nobody knows whether these projections will come true; nobody knows when mining will really commence; nobody knows how much other mining power will hit the Bitcoin network between now and the time that Cloud Hashing’s pre-ordered equipment arrives. You can check for yourself what you think of the projected 500% 1-year return using a profitability calculator combined with estimates of the total hashing power which will be hitting the network. But because nobody knows for certain whether any of Cloud Hashing’s projections will actually come true, it seems to me there is little to be gained from arguing the finer points.

What we can say, however, and without much argument about anything, is that the distribution of risk in this business model is entirely one-sided: proprietors of schemes such as CloudHashing.com’s get paid right up front, and they will retain their share of the money handed over by hopeful members of the public regardless of when or even whether any hardware is actually delivered. (Note that CloudHashing.com sells pre-orders for two years of mining power for nearly twice what Butterfly Labs charges for the underlying piece of hardware. A roughly 90% markup for a pre-order built atop a pre-order is nothing to sneeze at.) Investors in PMBs and those members of the public who hope CloudHashing.com will generate a 500% return by mining for them bear every last iota of risk that hardware will never be delivered at all; that hardware will be delivered late; or that hardware will be delivered around the same time as millions of dollars of other people’s hardware, potentially adding thousands of terahashes of power and radically decreasing the expected economic return of mining.

My point is not to pick on the Cloud Hashing folks, or even on issuers of PMBs. (After all, some of the latter at least do have a demonstrable history of performing actual mining.) Rather, my point is simply that the entire model embodies risk that is asymmetrical and, at least on the face of it, highly unfavorable to investors. Generally speaking, I believe wildly lop-sided risk coupled with eye-popping levels of “estimated” return are a sign of something other than a great investment. There are always exceptions, but to my mind, these types of schemes have little or nothing in common with real investing or even speculating; even the word ‘gambling’ seems too charitable.

• Bitcoin