Tl;dr Bitcoin maximalists argue that the energy consumption of Proof-of-Work provides the network with a security moat. In reality, it introduces political risk because of the highly centralized nature of energy infrastructure and markets. Evidence points to Bitcoin being susceptible to sovereign censorship in the future as its network growth continues along existing geopolitical fault lines. Alternatives to completely miner-driven blockchains, like Lightning Network or Proof-of-Stake, reduce or avoid this issue by relying on endogenous factors (i.e. the native currency) for transaction validation.
Bitcoin has redefined what money means in the digital age. But as the “censorship-resistant” cryptocurrency enters its second decade, its advocates must reconsider the narrative that has fueled its rise. Bitcoiners will roll their eyes at that last sentence: they have spent ten years successfully weathering concerns over block size, transaction throughput, and energy consumption, after all. Why should this alarm sound any different?
- In a word: asymmetry.
- In a phrase: the different profiles of bitcoin miners and holders.
- In a sentence: the geographical distribution of Bitcoin’s hashrate creates an attack surface for sovereign censorship that scales with network value.
Bitcoin might not depend on any nation-state, but it does not exist in a political vacuum. The network addressed the aforementioned issues through the market dynamics enshrined by Nakamoto consensus (i.e. miners continued to approve transactions, holders continued to store value in bitcoin). Because censorship is exogenous to the network, however, the integrity of the network need not be the only deciding factor for both miners and holders.
This article argues that if Bitcoin’s current growth pattern holds, miners’ and holders’ incentives will diverge such that censorship resistance will no longer be a natural Schelling point. That divergence will come as a result of an influx of western institutional investment whose interests misalign with those of mining pools concentrated in China. If sovereign actors can incentivize miners with political value greater than that endogenous to the network, Bitcoin could be weaponized into a vehicle of economic sabotage.
When a Node is More Than the Sum of Its Parts
Bitcoin has successfully established itself as a global currency. It doesn’t take orders from a central bank, it ignores capital controls, and it is open to everyone. It has made Americans fortunes, aided Russian conspiracies, and given Venezuelans lifelines — all in programmed, emotionless fashion. But its growth from cypherpunk experiment to global asset class has come at the cost of shedding some of that apolitical design.
We can trace the roots of this change to the emergence of ASIC (application-specific integrated circuit) miners in January 2013. Given the zero-sum nature of bitcoin mining, their performance advantage over general CPUs/GPUs set off an arms race. This has made it progressively more prohibitive to run a full node, as you’re less likely to recoup profits if you’re never able to win a block reward. Five years on and Bitcoin has grown entirely reliant on ASIC manufacturers like Bitmain and Bitfury, and the mining pools that serve as their hungriest customers, to supply its hashrate.
Bitcoiners argue that this degree of miner specialization guarantees security by better aligning miner incentives to the network and raising the cost to attack it. While the article referenced makes the explicit caveat that this would only be the case in a political vacuum, even when it accounts for the threat of politically-motivated actors, it stops at the ASIC manufacturers themselves. And because ASIC manufacturers make money hand-over-fist with their current business model, they’re incentivized to preserve the network’s integrity and not to make malicious use of their centralized power. This argument makes sense within the context of on-chain politics — it does not when one considers off-chain political incentives. Not to mention, new competition in the manufacturing space threatens to do away with that golden goose altogether.
Geographic Profiles of Miners and Holders
The idyllic image of the family farm no longer represents modern agriculture. The narrative of dissident cypherpunks running Bitcoin nodes in their closets rings similarly hollow. Industrial-scale mining pools, either public or private, power virtually all of the network’s hashrate. They make operating decisions in pursuit of profit like any business would, as evidenced by a) their concentration around cheap sources of energy and b) mining hardware lifecycles. Importantly, this leaves them at the mercy of their local business environments.
As of September 2018, Chinese mining pools comprised nearly 2/3rds of Bitcoin hashrate. While any Bitcoin node can join most of these from around the world, cheap coal and hydroelectric energy ensures that Chinese miners take home the lion’s share of pool earnings. This is in spite of the government’s endless litany of supposed bans and clampdowns on the cryptocurrency. Their market dominance has shrunk since, but F2Pool and BTCC led the rise of Chinese mining back in 2013 and 2014, respectively.
Now, contrast this with the geographic distribution of bitcoin holders (using node distribution as a proxy**). It comes as little surprise that the majority of holders hail from OECD countries with mature financial systems: fiat on- and off-ramps have primarily required traditional banking infrastructure. Clues from Bitcoin address allocations suggest a similar distribution. The network’s early community hailed predominantly from English-speaking locales, and it’s these early adopters who own a majority of the bitcoins mined thus far. The noteworthy exceptions to this observation are countries like Venezuela, Argentina, South Africa, and Turkey, where a combination of weak currencies and stringent capital controls leave Bitcoin as a lucrative alternative.
Comparing these two distributions reveals an unsettling fact: Proof-of-Work consensus suffers from an inherent geographic path dependency. Satoshi’s innovation was not flawed for rewarding early adopters and, later, specialized miners; rather, it was the disregard for how these groups could be different subsets of individuals that leaves the network vulnerable. Before ASICs, the network favored English-speakers from developed economies who had the technical know-how to navigate early Bitcoin clients. Afterwards, the network’s hashrate came to reflect a distribution of the world’s cheapest energy markets to optimize the profits of industrial-scale miners. This suggests that, at any given moment, exogenous factors inadvertently bias Bitcoin in favor of nodes with a specific geospatial profile.
** Pools are represented by single nodes.
Stratification of Miner Generations over Time
Why does it matter that Bitcoin mining tends to concentrate in certain geographies when holders can transact bitcoins anywhere in the world? Because Proof-of-Work consensus rewards earlier miners exponentially more than it does later miners, later miners have progressively less skin in the game, in absolute terms, to secure the network. That is, they have less of a vested investment in the network and would require less external incentive to act against it. If economic and technological factors lead to periodic changes of the guard in who is mining Bitcoin, that disparity in on-chain incentive could be exploited by off-chain actors.
The original Bitcoin miners who used CPUs and GPUs have seen the utility of their bitcoins as stores of value increase relative to their initial cost to mine over time. Even if they secured free electricity, ASIC-utilizing miners could never match their predecessor’s gains, given that over 80% of all bitcoins have been mined already. Moreover, those miners-turned-holders have little incentive to sell their bitcoins: as a deflationary asset, they can assume that the value of one will increase going forward, all else held constant. Miners of the network’s different periods are thus stratified, with post-2013 miners reduced to second-class citizens and the miners of the inevitable post-ASIC period doomed to a similar fate.
This refutes the idea that Proof-of-Work’s electricity consumption creates a security moat around the network. By tying Bitcoin to an external resource, it instead introduces friction that warps the incentive design of the network. That friction stems from the fact that electricity is inherently geopolitical: location, public infrastructure, and utility regulation all influence its price and availability. It calls into question the “one-CPU-one-vote” assertion of the Bitcoin white paper by pricing out all but the dominant homogenous (re: networked-preferred characteristics) nodes from validating the network.
Node homogeneity reduces Bitcoin’s censorship resistance by creating a honeypot of miners whose interests increasingly :
- Diverge from those of former miners over time.
- Converge with those of miners subject to the same jurisdictions.
Political Force as an Off-Chain Incentive
Bitcoin’s shift to ASIC mining created a house of cards in which the majority of the network’s value depends on a second-class minority for security. The energy required to power these specialized, industrial-scale miners has biased the network to favor those nodes located in a small subset of locations with cheap electricity. As the geographic diversity of Bitcoin’s hashrate decreases, the network grows more vulnerable to sovereign censorship by its miners’ resident jurisdictions.
In theory, high capital barriers and the absence of a central point of failure combine to dissuade censorship. Not only must an attacker spend an inordinate amount to commandeer more than 51% of network hashrate for an indeterminate period of time, but the prospect of a hard fork provides honest nodes with a safety net. The cost of attacking Bitcoin has successfully warded off would-be attackers for a decade because it’s always made more sense to either invest in the network or settle scores off-chain. Given growing adoption by traditional finance and business interests, however, Bitcoin stands to lose this flexibility as greater economic leverage generates a compelling enough political attack surface.
“Leverage is an investment strategy of using borrowed money — specifically, the use of various financial instruments or borrowed capital — to increase the potential return of an investment. “ -Investopedia
Leverage as it relates to cryptocurrency today remains the domain of exchanges like BitMEX and Kraken which offer margin trading to their customers. While their financial instruments often elicit comparisons to gambling, the leverage they offer does not transfer risk in or out of Bitcoin. Because their leverage is synthetic, financed by parimutuel pools, failed longs and shorts simply result in transfers of ownership to opposing traders. At worst, these platforms rely on insurance funds or socialized losses to fulfill the gains of successful, but market-cornering, options — both of which are denominated in BTC.
Now contrast that with the derivatives offered by traditional financial institutions. Predicated on the western preference for fractional reserve banking, these instruments provide little more than paper claims to the underlying assets they represent. Certain practices in particular allow for this indirect leveraging of underlying assets, i.e.:
- Rehypothecation: same asset used to collateralize multiple contracts
- Commingling: different assets used to stand in for the initial underlying
- ETFs: multiple assets tracked by a singular exchange-traded fund
This phenomenon results in a paper market that far outpaces its underlying physical market. For example, the market for notional gold saw trade volumes totaling over $9 trillion in 2016, 223 times greater than the $42 billion in gold investment that occured that year. There’s little reason to doubt that institutional investors would handle digital gold any differently.**
Assuming we see the requisite near-term progress in asset custody solutions and regulatory clarity, financial institutions across the OECD will look to get bitcoins on their books. This influx will initially drive large-scale transfers of wealth in the OTC markets to fuel the rapid expansion of Bitcoin-based financial instruments in traditional capital markets. Even if only a fraction of Bitcoin’s supply enters into circulation, a primary market worth some hundreds of billions could foster a secondary market that surpasses the trillion-dollar mark. Bitcoin hashrate will follow this demand increase in lockstep due to its strong correlation with price.
The parallel rise in Bitcoin derivatives and network hashrate will magnify existing geographic fault lines. The aggregate value of the network will skew further towards developed western economies; its mining power will further concentrate in dominant energy markets. What will warp the incentive design of the network, however, will be the financial jurisdictions barring the majority of miners from participating in the secondary markets. If mining continues to concentrate in countries with tenuous to flat-out hostile US relations, like China, Venezuela, or Russia, western sanctions and restrictions will only amplify the disparity between miners and the wealth they’re securing.
** See Caitlin Long’s responses to ICE’s Bakkt saying that they wouldn’t here.
Sovereign actors could exploit the divergent concentrations of Bitcoin holders and miners across adversarial jurisdictions, but any attack would require that the political gain of doing so outweighs its economic cost. In a scenario where the attacker has little to no exposure to the secondary market, an attack on the underlying asset guarantees that their expenditure results in economic damage equaling a higher figure. The question remains, however: how do you quantify the political incentive to do so?
Two phenomena from international relations theory can provide a bridge between political and economic value.
- The Thucydides Trap asserts that as a runner-up country gains economic and military might, it invariably results runs into conflict with the status quo’s dominant one. Of 16 identified iterations, it’s proven true 12 times.
- The G-Zero World is Ian Bremmer’s argument that there no longer exists a singular source of global leadership like that of the US in the post-WWII era. Without a hegemon to back international institutions, countries will revert to regional and bilateral coordination.
Large countries that have lived in the shadow of US foreign policy have the incentive to undermine both it, and the international systems it has designed. They want to weaken the country’s hold on international trade and diplomacy to hedge against possible future conflict with the US. Moreover, they know that no dominant replacement could intervene in such conflict, leaving whatever spoils they snag from American decline as theirs.
Because the existing financial system is predicated on American leadership, it represents a target for countries looking to get out of its shadow. A secondary Bitcoin market with underlying assets secured by miners in a number of countries with such ambitions is a possible attack vector.
These assumptions allow us to walk through a possible scenario in which the political gain of censoring Bitcoin outweigh the costs of sabotaging network:
- Institutional investors buy up some fraction of Bitcoin’s circulating supply and create a multi-trillion dollar derivatives market accessible to accredited investors in OECD countries.
- Increased demand leads to a) higher prices and b) greater hashrate. This further incentivizes industrial-scale mining in cheap energy markets and ASIC manufacturing by landed incumbents.
- The norm becomes for all western financial institutions and companies to have some degree of Bitcoin exposure within their treasuries. Meanwhile, China has continued its “blockchain not Bitcoin” campaign, minimizing the country’s commercial exposure to the cryptocurrency.
- Conflict breaks out between the United States and China sometime before 2030. Reasons why could range widely: the South China Sea, the Indo-Sino border, and the North Korea dilemma all represent possible flashpoints. American military might forces China to pursue indirect attacks: cyber-warfare, trade disruption, political manipulation, etc.
- China identifies Bitcoin as an attack vector. It commits the tens of billions necessary to commandeer the network by ransoming electricity generation, paying off domestic miners, and taking ASIC manufacturers’ supply chains hostage.
- Bitcoin itself recovers from the attack through a hard-fork, but the damage has been done. The secondary market evaporates, the price of the forked Bitcoin plunges, and its perception as a store-of-value thesis disappears. China successfully unleashes a financial contagion similar in design (but not in scale) to the 2008 mortgage-backed securities crisis. Its economy remains relatively insulated from the fallout, as its investors were barred from secondary markets and miners had cashed in profits prior to attacking the network.
Should this scenario fall under Black Swan categorization? Absolutely. But it demonstrates that while Bitcoin can survive just about anything as a P2P transactions network, recent attempts to pass it off as digital gold or a store-of-value with a 10,000 year shelf-life are misguided. Its dependence on external resources introduces a geopolitical dimension to the network that mirrors the very systems it hopes to replace. This the technologist’s fatal flow: to think that their technology exists in a vacuum, when in reality, it amplifies existing societal and political realities.
Addressing Solutions and Counterarguments
Make no mistake: cryptocurrencies are here to stay. They provide the digital, cross-border infrastructure our increasingly digital global economy demands. But if cryptocurrencies are to replace state-backed ones, they will need to minimize their exposure to off-chain political risk. Otherwise, blockchain networks will simply provide new battlegrounds to rehash the same old power struggles.
Can Proof-of-Work accomplish this separation of crypto and state? In its current form, no. Its demand for electricity and dependency on hardware will always result in some dominant geopolitical distribution of the network’s miners and holders, which introduces the risk of political manipulation. An even more intrinsic impediment, however, is that miners and holders can be different subsets of people in the first place. While it has allowed the network to scale to its current heights, the distinction separates transacters from processors in such a way that sustaining a network-beneficial Schelling point becomes difficult over time.
Potential solutions to this security flaw focus on making holders less dependent on miners, or on external resources all together:
- Lightning Network is a layer 2 solution for Bitcoin that allows holders running nodes to open transaction channels between one another. Lightning transactions are paid for by proportional fees, miners are only involved at the opening and closing of channels, and mutual nodes can support channels across each other a la six degrees of connection. This makes holders less reliant on miners.
- Practical Byzantine Fault Tolerance (pBFT) is a complex concept that is best left to be explained by this article. What it solves, however, is the arms race for bigger and better mining rigs. pBFT blockchains like the one created by Helium rely on Proof-of-Work for Sybil attack prevention, but not for node consensus, meaning that nodes that are determined to be honest no longer compete against one another. This gets rid of the need for ASIC style specialization and counteracts industrial-scale centralization.
- Proof-of-Stake Consensus replaces miners with stakers, network holders who submit their native currency to a smart contract for a set period of time, granting them with a chance at validating transactions. Ethereum is leading research on this design with the impending release of “Shasper” but it remains highly unlikely that Bitcoin would make the transition away from its Proof-of-Work origins. This style of consensus algorithm gets rid of the exogenous dependency that plagues Proof-of-Work as we know it.
Bitcoin has evolved from a cryptographically-secured P2P payments network to the basis of a new asset class. Its proponents have justified this shift with the introduction of new narratives centering on its role as digital gold, suggesting that it will become the global benchmark for storing value. More recently, Bitcoiners have argued that its energy consumption is what backs the cryptocurrency, like how USD was once backed by the gold standard. They reason that the expenditure of physical resources represents a tax the network pays for miners to secure it, providing a realistic baseline for the price of one BTC.
This narrative suffers from a naive understanding of how energy markets work. It ignores the fact that electricity is political: by and large, it is subsidized by governments, produced by monopolistic utilities, and delivered through public infrastructure. While Bitcoin in theory is distributed across enough jurisdictions, to the point where singular domestic grids have no sway over the network, the rise of industrial-scale mining has proven otherwise. The network’s miners have concentrated in a select few of the world’s cheapest energy markets, and will continue to do so as hashrate increases, leaving specific grid operators with disproportionate control.
The cost of censoring the network and the inability to target specific nodes or addresses will prevent an attack by a sovereign actor the vast majority of the time. However, Bitcoin does not exist in a political vacuum. With the cheapest mining costs to be found in countries like China, Venezuela, and Iran, but the majority of network wealth held by OECD citizens, the political dimension introduced by electricity dependence presents clear friction. This friction will only increase as secondary Bitcoin markets explode and miners from the aforementioned countries remain barred from participating.
Nakamoto consensus fails to consider how network growth could mirror off-chain politics. Granted, Satoshi might never have wanted Bitcoin to become an asset class for traditional financiers to play with — but their prediction of specialized mining farms reveals that they should have known how geography would play a role in Bitcoin. This lack of foresight will weaken Bitcoin’s censorship resistance over time unless the network works to implement soft forks that reduce the need for centralization in specific geographies. Until then, it’s only a matter of time until a sovereign actor has the incentive to attack the network through the very mechanism meant to secure it.
This section will be reserved for counterarguments to be added after the initial publication. Responses will be edited in and shared via the original Twitter thread.