Graphic by Garrett Kinsman

Contesting Token Velocity as an Issue

Velocity-based valuation approaches may be fundamentally wrong

About the writer

Madhumitha Harishankar is a PhD Candidate in Carnegie Mellon University studying the application of network economics in wireless networks for resource sharing. She is advised by Prof. Patrick Tague and work closely with Prof. Carlee Joe-Wong. This work is in collaboration with blockchain-based IoT service provider Nodle.

Token staking and burn-and-mint mechanisms are used to incentivize users of a cryptocurrency to lock up their token holdings for periods of time, and have become popular ways of controlling token velocity. Lowering token velocity is considered necessary to ensure that the network’s valuation grows proportionally with transaction volume, especially for utility tokens. The argument typically goes something like this:

  • Assume there are no speculators in the market; i.e., only the vendors providing services and customers interested in these services exist in the marketplace for a decentralized company called AppCoinCompany. AppCoinCompany may require vendors to charge a uniform price (e.g. if these are substitutable services) or allow different vendors to price differently. Assume the former for ease of argument.
  • AppCoinCompany has x customers who purchase y AppCoins each to avail the vendors’ services and immediately exchange these coins for the services. Vendors then immediately go on the exchange to sell the AppCoins (x*y in total) and realize revenues in, for instance, USD.
  • Assume a pure exchange economy; i.e., fixed total circulating supply. Let this total available supply of coins in the market be x*y.
  • Now, say AppCoinCompany’s customers grow from x to 2x. Given that there are only x*y coins available, which are fully required by the x customers who were already present, one may expect the AppCoin to increase in (USD) value, simply because demand > supply now, assuming vendors’ prices (charged in AppCoins) remain fixed.
  • However, this may well not be the case. Previously, when the system had x customers, suppose they availed the vendors’ services on a daily basis at some time t of the day, and the vendors put the coins back on the marketplace at time t+1, after their transactions with these x customers are done. Assume the newly added customers also have a daily demand for the vendors’ services. In that case, they could well come to the exchange everyday at t+2, and procure these coins at the same rate that the first x customers that came in at t procured them, since at t+2, the instantaneous demand is only x (instead of 2x) and the supply is still x*y.
  • Hence, the problem seems to be: growth in the network use and cash flows do not translate directly into growth in coin value.

And hence the various mechanisms to force users to hold on to their AppCoins for a while (rather than dumping them straight on the exchange as soon as they are done using the company’s services), which lowers supply and presumably gives the marketplace a chance to reflect the increased demand.

But is this really a problem?

Decreasing token velocity in response to this “problem” appears inappropriate. For dividend-paying stocks, it indeed makes sense for the value of the stock to be proportional to the cash flows (and thereby reflect such growth in customers and service demand as in our previous example). However, tokens are not stocks, they do not pay dividends, and their value should not grow directly with growth in demand.

Coins buy privileges to access the services of a platform, not the platform itself. So it makes sense that their value will fluctuate with real-time dynamics of supply/ and demand for the platform’s services.

As we already see from the simple example discussed previously, AppCoin’s value should, in fact, grow with growth in instantaneous demand. As evident, this is harder to achieve than mere growth in overall product consumption and is one of key tradeoffs in considering a crypto-funding model vs. a stock model. Coin holders do not directly benefit monetarily from an increase in consumption (since they are not entitled to any dividends). Therefore, the consumption growth must be significant enough to raise the overall instantaneous coin demand for the token to have meaningful rise in its exchange rate. If it is likely that many consumers will ask for the platform’s services at the same time and continue to return with high frequency to procure services, then there is a good reason to believe instantaneous demand for AppCoins will grow and to therefore invest in it.

There is a second way that the token can organically increase in value as well. If the quality of the underlying services provided increases/the target consumer base is diversified (i.e. higher product differentiation), then the fiat value of the provided services increases. AppCoinCompany, in that case, should increase its price (in AppCoins) to reflect its evolved product quality under the competitive market. For instance, x customers may now require 2y coins instead of y to procure AppCoinCompany’s services, thereby increasing instantaneous demand for the coin (without necessarily increase in customers) and hence the coin value.

Is the mechanism design feasible?

The real market has several types of investors and speculators with largely unpredictable and possibly manipulative market behaviors. Further, most AppCoinCompanys have an increasing amount of total circulating supply (until all coins are minted) and hence they are not pure exchange economies that we saw in the simple example above. Given these, is it possible for an AppCoinCompany to influence its service pricing, rewards and coin policy such that AppCoin exhibits a meaningful growth in value against these two factors described above?

I will delve into this further in future work, but here is a (possible) broad-strokes answer:

  • At any given time, the market should price AppCoins for the (discounted) final future customer-facing prices (in fiat) of AppCoinCompany’s final products and services and final future instantaneous demand. Consequently, AppCoinCompany should charge consumers the competitive fiat value of their services, dynamically priced in AppCoins since the market rate for AppCoin may be highly volatile.
  • When all coins are mined, AppCoinCompany’s services are “final”, and customer base is more or less constant, speculators’ best strategy is likely to sell their coins at the current market rate since there is no reason to expect further coin appreciation (based on the two factors discussed above).
  • At that time, AppCoinCompany may revert to a fixed-price model where its charged prices (in AppCoins) takes into account the total circulating supply (most of which should be highly liquid at this point).

In Summary

  1. Tokens are not stocks. By putting in techniques to artificially decrease token velocity, AppCoinCompany forces AppCoins to appreciate directly with growth in cash flows which is inappropriate for a non-dividend model and lends AppCoins’ value uninterpretable. If AppCoins are used to access AppCoinCompany’s services but valued for a market cap that does not translate into dividends on the token, the resulting value of AppCoin is un-interpretable.
  2. There is still reason good to invest in tokens if you believe customers will, at some point in future, be willing to pay more for underlying service than they do today. This could happen due to increase in long-term instantaneous demand or increase in product quality or both.
  3. As an investor in AppCoins, what you are investing in/betting on is different compared to stock investments, though these are related. When you invest in stocks you are acquiring a stake in the company’s profits in the form of dividends. When you invest in tokens, you are acquiring a stake in the consumers’ willingness to pay function for the underlying product/services that can be redeemed with AppCoins.



Madhumitha Harishankar
Writer for

I am a 4th year PhD Candidate at Carnegie Mellon University. I enjoy applying economics, optimization and machine learning to wireless networks and sensor data.