The crypto space is rife with examples of failed business strategy, from EtherDelta to the more recent example of 0x. What these failures show us are not that the strategists behind the projects were poorly prepared, but that cryptocurrency is an industry specifically primed for ultra-competitive market forces. This piece will go into some of the details around how competitive forces shift when we compare industries within the cryptocurrency space to their traditional counterparts, and how the levers of power shake out in the open-source world.
Industry Analysis in Cryptocurrency
The first question we have to ask ourselves before we look at industry attractiveness is, what is the ultimate judge of industry attractiveness in an industry that doesn’t necessarily target profit? In any other industry you can get a relative proxy for attractiveness by looking at the profitability of similar firms. If we looked at airlines, we would see that very few companies are strictly profitable and as such the industry is probably not attractive. We may be able to use market cap as a proxy, but considering the lack of fundamental analysis and true user adoption in this market it is unlikely this would give us any sort of relevant proxy for attractiveness. For this reason it makes more sense to begin with a qualitative analysis of the industry, and to understand what the makings of a competitive threat look like for a protocol.
For those of you who haven’t seen this chart, it is a representation of the five forces analysis pioneered by Michael Porter. This chart represents the five ways profitability is determined by forces outside of direct control by the form, with one slight difference: suppliers have been replaced by miners, and buyers have been replaced with users. We can think of ‘miners’ in this case being those who practice generalized mining (including staking, creating CDPs, providing bandwidth, etc.) as opposed to proof-of-work mining. Users are assumed to be those participating in the network from a demand side; more often than not being user aggregators like wallets and exchanges. Substitutes and new entrants are largely unchanged concepts. The far more interesting analysis, however, happens when we look at each of these circles in more depth.
We can think of miners as ‘suppliers’ of protocol networks. They perform almost identical functions as say, suppliers of crude oil in the oil refining value chain. One of the main differences between suppliers and miners are that protocols always set the rules. Miners may be able to influence these rules, but in practice their many attempts to control protocols have generally failed. Because of this we have seen miners migrate to different protocols in search of more profitable returns, but they have yet to act as a oligarchs and truly manipulate networks (except in the obvious case of 51% attacks). The following excerpts are examples of how the powers that traditionally shape profitability apply to cryptocurrency.
Differentiation of inputs: ASIC mining has a significant impact on input differentiation in terms of relative cost competition. In most protocols the inputs (proofs/transactions) are determined by the developers, so suppliers cannot differentiate per se, but they can compete on cost. For a while, Monero had secret ASIC mining being performed on it which allowed the suppliers an unknown leverage over the protocol. In Sia/Filecoin however, miners can directly differentiate with better performance. Miner differentiation is up to the discretion of the protocol, but at the end of the day it needs to create a profitable and productive network.
Switching costs of suppliers and firms in the industry: Determined by the industry landscape i.e. miners can switch easily between different chains with the same hash function and developers can switch easily between forks of the same protocol. This is a mild looming threat for protocols with significiant forks such as Ethereum, as many of the attempts to rescue user funds have resulted in developers threatening to move to Ethereum Classic.
Because EIP 867 could make such corrections more frequent, hundreds have stepped up to voice their opinions on Github — with some threatening to migrate to ethereum classic.
Supplier concentration: By default this should be less significant for protocols than it is for traditional businesses, but not irrelevant. This is simply because the suppliers compete on the terms of the protocol, so concentration does not give additional leverage in the same way it does elsewhere. In traditional markets if the supplier has a large market share they can enforce their own pricing schemes and logistic frameworks on their consumers, but in protocols they can only vote with their feet. This, however, can be an issue when we think about integration through the value chain.
Threat of forward vs. backward integration: DDEX is a great example for this one; they forward integrated into 0x by forking it because they didn’t want to deal with the token (plus some complaints about order matching). Now, they will have a protocol with half the liquidity and none of the ZRX token friction; It is much more difficult to do the opposite simply because you can’t fork liquidity.
To define users, in most scenarios we are not talking about the end consumer. The users we care most about are the ones who can drive the most value to our platform in the form of protocol demand. In many cases this will be the last layer of abstraction between the protocol and the end user, apps. In this scenario protocols like Bitcoin will have to contend with wallet apps that aggregate users, and sidechains like the lightning and liquid networks.
User concentration: Interestingly enough, this has been an issue for Bitcoin itself. Large exchanges like Coinbase and Binance have the ability to support / deny access to soft forks and hard forks to the network (i.e. segwit and bitcoin cash respectively). This type of power will mainly affect transactional currencies like Litecoin, however. Any currency that is looking to target mass adoption needs strong user adoption, and because of this the firms that control adoption will hold better hands.
Substitute products: Bitcoin cash is a substitute for Bitcoin only in performance. Substitutions have a heavy dose of memetics, at least at this stage in the market, so it may be harder than it seems to create a substitute that is actually used. At the end of the day, it actually is fairly easy to create a scalable blockchain for transactions from the tools we have today, the hard part is convincing everyone that it’s secure and worth using.
Backward integration: Direct control of the user experience is known in the tech world to be the single most significant driver of value. This is known as aggregation theory. Binance has already started to take advantage of this theory with its recent announcement of Binance Chain, a smart contract chain that will leverage Binance’s user base and liquidity pool to create a decentralized exchange. This is one of the most interesting developments of 2018 that managed to slide by relatively unnoticed.
Threat of New / Substitute Protocols
The threat of protocols usurping other protocols’ markets is ever-present, but it takes on different dynamics in the cryptocurrency space. For instance, there is the constant threat of protocol forks. These are often advantageous in the short-term to the forker, but long term it may actually contribute to the strategic worsening of the industry. Take the DDEX fork for example, while it makes sense for them to fork in the short term to improve their product with better order matching, it actually worsens the competitive position of decentralized exchanges relative to traditional ones. By forking the protocol they are splitting the liquidity into different pools, which even they themselves claim will be the ultimate determinant of the protocol’s success.
“We are calling this new protocol Hydro, to emphasize that liquidity matters above everything else.”
This is one example of the considerations to be made when creating new / substitute protocols. Here are some of the others:
Economies of scale: In theory there should be economies of scale only on the supplier side, not on the protocol side. PoW miners can create economies of scale by bulk purchasing mining equipment and electricity, but protocols should almost always produce at cost. In fact, one could argue there is a diseconomy of scale for something like Bitcoin. As more and more users pile onto the network, the price increases, the hashrate increases, the difficulty increases, and it actually becomes more expensive to perform one transaction. Proponents will argue that this is a feature, not a bug. I am inclined to agree in the case of Bitcoin, but would also caution that this dynamic can really only exist for one asset (digital gold) and will not be applicable to the wider market.
Brand identity: Brand identity is far more powerful in the cryptocurrency space than it is in traditional business. This is because brand drives price, and price drives utility in a weird circular argument. Imagine the barriers to entry of digital gold today. As Hasu puts it:
If you wanted to fork bitcoin — and not have the new network die immediately — you would have to fork the social contract first.
Put another way, in order to create a new cryptocurrency that has value it needs to have a social contract that appeals to people. This social contract is very often what creates a cryptocurrency’s brand. Nano is based off of the idea that the original distribution of a cryptocurrency should be more human-centric than Bitcoin’s was, and Ethereum is based off the idea that functionality is more important than monetary policy. Brand has had huge implications for the success of cryptocurrencies in the past and most likely will continue to into the future.
Proprietary product differences: As most protocols are entirely open-sourced, there is significantly less that can be done to protect a protocol than a traditional business. That being said, here are a couple things that apply to certain protocols:
- Liquidity pools — Applicable to decentralized exchange protocols like 0x, dy/dx, and airswap.
- Community Engagement — Applicable to research pools like Messari and blogging platforms like Steemit.
- Unforkable UI — Certain platforms create their own UI with proprietary codebases. This includes platforms like Polymath and Santiment.
Capital Requirements: Protocols do not need capital in the same way that other businesses do. The only capital they truly require is human capital in the form of developers. This, however, is a significant barrier to entry given the general shortage of developers in the space. Certain protocols (Dash, Litecoin) have leveraged marketing costs as part of their competitive advantage to some success, but in a market polluted with Bitcoin killers we will have to see if this strategy is successful.
Switching Costs: Switching costs between protocols for both suppliers and users are defined by a couple different types:
- Implementation costs: The cost it takes to re-implement one protocol in favor of another in terms of development time, cost of developers, and loss of functionality.
- Brand costs: The costs associated with differences in brand between one protocol or another.
- Operational costs: The costs associated with increased prices / decreased profit for the usage of each protocol, i.e. if transactions cost more on Bitcoin than Bitcoin Cash.
So what does a moat look like?
Based on this analysis we can think about moats across a couple of dimensions:
- Unforkable user experience
- Attractive social contract
- High and necessary protocol complexity
- Low / no miner economies of scale
- Strong developer team / community participation
These are the factors that can prevent a protocol from being outcompeted. If we look at a currency like Bitcoin on all of these dimensions, we can see why it is the most valuable. Bitcoin outperforms all its competitors mainly with its social contract, lack of miner economies of scale, and its strong developer team. With these combined it has dominated the store of value market and has created a community that will strive to improve its usefulness. A protocol like 0x, on the other hand, had low bars in all of these categories. The lack of liquidity gave the ability for a main liquidity provider to fork without much opportunity cost, and its low complexity allowed them to develop it in parallel. 0x had also not created any sort of user experience that could protect them in the event of a fork, nor did it have a particularly attractive social contract. 0x has great developers, however, so the jury is not completely out on them yet. Ultimately it will be up to the market to decide.