Incentive Fundamentals for a Token Economy

A year ago I took a hard right with my career and dove fully into the blockchain space. After months of networking and MeetUps — hoofing it an hour-and-a-half to some and starting my own in part to get something closer— I was fortunate to find myself in a position to think critically about cryptocurrencies and what factors drive value in permissionless blockchain systems.

That lead to other opportunities to think big about incentive systems and now I find myself in the happy position of talking to founders of start-ups in all types of industries looking to apply this technology to their ideas.

What I’ve found is that it is definitely not just about the technology but equally about the economics. True economics is the study of human action and good token models acknowledge the power of bottom-up economic activity. That I spent twenty-three years in investment management modelling value factors and questioning the top-down economic dogma of the industry from the perspective of classical liberal economics seems to have helped me understand the blockchain space (at least permissionless systems). However, regardless of the domain, these are still businesses started by individuals with profit motives. As I talk to these entrepreneurs, they often get stuck on how they can benefit financially from doing business on blockchain.

To be sure, cryptocurrency and blockchain technology is not driven by people with purely altruistic motives. They may have additional motives distinct from profit but they all have strategies to benefit financially. What they realize is that there is a different way to generate and capture value. By building networks that incentivize certain types of work, they monetize the value of the work performed on the network — the token or cryptocurrency is how they monetize that value. For the typical entrepreneur, this requires a very different perspective on revenue and profit models.

With respect to the economics of the ecosystem that drive the value of the token, there are two fundamental models at work. One relates to the Quantity Theory of Money and the other relates to Network Theory of Value. The Quantity Theory of Money is a fundamental approach that can be used to explain a tokens intrinsic value. The Network Theory of Value effectively explains it’s total value, which can drive a premium to the fundamental value.

The Quantity Theory of Money has been around for decades and, when I started in this business, it was still relevant to analyzing monetary policy (though it definitely fell out of favor over the last three decades). It states that the total money supply (M) multiplied by annual turnover of the money supply, or velocity (V), equals the price of the average goods and services (P) times the total quantity of goods and services purchased during the year (Q). Thus, MV=PQ. MV is the total quantity of money demanded over a period of time and PQ is the total value of goods demanded over that same period. Traditionally, PQ is nominal GDP where P is the level of the inflation index and Q is real GDP. M is a measure of the supply of U.S. dollars in circulation, for instance, and V is the annual turnover of the dollars required to meet demand for the dollars to buy goods and services.

Now let’s look at a token that represents a unit of data either stored or otherwise linked to a blockchain record. In the case of this token, M is the total circulating supply of tokens (only issued, not held in reserve or otherwise locked out of circulation), P is the number of tokens required to purchase a unit of data (think of it as the average price of a unit of data across the system), Q is the total units of data exchanged on the system during the course of a year and V is the number of times the token supply turns over during that same time period.

Note that P is inversely related to token value. If today one token buys one unit of data, then a holder of the token can command one unit of data. If tomorrow it takes two tokens to buy one unit of data, then the holder of one token can only command a half-unit of data, which is a depreciation in value. Alternatively, if tomorrow it takes ½ a token to buy one unit of data, the holder of one token now commands two units of data, so his token has appreciated and is now more valuable.

We can rearrange the formula to get P on one side of the equation, so MV/Q = P. P will decrease if our token appreciates in value so the incentives should be the following:

Incentives within the system should be aligned accordingly. Staking, burning or recycling (circular model for tokens) all help minimize M. Increasing the quantity of high value data increases Q so there should be mechanisms and/or incentives to validate the quality and to contribute data — this is where curation models come in. Velocity is really about expectations of increasing value and if people expect the value to increase, they will hold it rather than exchange it. If you do all of these thing right, and grow your network, then P will decline and the value of token will increase.

The Network Theory of Value is based on Metcalf’s Law, which states that the value of the network is related to the square of the number of users (U). To Metcalf’s law has been added the number of transactions, though there are many ways to estimate transactions. I describe my application of the model in the article linked here, where I use average number of token transactions per user (address) priced in USD (Tx) as my measure of transactions. To model the total market value of a token (total token supply * value per token in USD), or token market cap (MCT), we get the following:

MCT = U^2 * Tx

The implication is simple, increasing either the number of users (as measured by addresses) or dollar value of each transaction drives up the value of the token. The value implied by the network model can easily exceed the intrinsic value of the data on the system, which can be thought of as speculative value. Whether it’s pricing the present value of the future state of the data or simply a greater fool dynamic among traders is irrelevant. More people trading the token means more addresses and higher dollar value transactions per address, which drive the network effect higher, even if artificially above actual value of transactions for data on the chain.

So, as founders of the data ecosystem itself, there isn’t a revenue model per se, but there is a profit model. Sell tokens from reserves (or held by founders) gradually over time as the US dollar price of the token appreciates significantly beyond the value of the data within it.

As for revenue, there should be many actors within the system hoping to profit, such as data contributors, data analysts, data market makers, etc., and their revenue models will be to earn tokens. The more they hold onto these tokens rather than exchange them immediately for dollars, the better (see Velocity discussion above). Founders can then also earn revenues by contributing to the system going forward.

I personally believe that founders should avoid taxing the system directly to the benefit of any single actor (which may be defended as a formal revenue model), as it violates a principle of permissionless, open-source token economics. The monetization is a result of building the right incentives and profiting from token appreciation. I believe revenue should be derived entirely by competition on the chain. Otherwise, it feels like an entitlement that represents an unfair advantage against other players and could disincentivize their involvement. The incentive to drive the value of the token higher is simple and straightforward.

Thus, in this case, to model the total value of the system you estimate how much data you can get on the chain, the value of that data and the number of times that data turns over, or is demanded, as several players can demand the same data set. Then you estimate the growth of the data set. Given the size of the data market and value of that data, the numbers can quickly get staggering. If a founder is confident they can craft solid incentives and build a robust network of users and contributors, they may just have a solid blockchain business.

Related posts:

I want to go to space. Probably not words you are used to hearing from a girl, but I do. And, to be clear, I am not your typical STEM girl. I am not a master coder, math is challenging for me and…

Python is an high level programming language. It is easy to understand and used to develop programs both in small and large scales. If a person didn’t know any programming language. So, he/she can…

Every single time Coinbase lists a new coin, its market cap goes through the roof, even Ethereum Classic (ETC),which has basically no community or developer support, value skyrocketed after being…