Measuring Liquidity for Governance Tokens

Governance Tokens issued by DAOs may suffer from low liquidity and may struggle to find price discovery for their Governance Tokens (aka Native Tokens).

This makes it difficult for potential buyers, some of whom may want to participate in DAO governance (and governance over their owned protocols), to enter the market for these tokens.

In this article, I explore some measurements for liquidity which could be employed by DAOs to measure how liquid their governance tokens are in the secondary market.

I begin by introducing the most commonly used metrics, why some of these metrics can't yet be used for on-chain data, and then provide additional metrics which could be used in their absence.

Here is a list of commonly used liquidity measures which can be easily calculated from Automated Market Making (which are typically used for on-chain asset exchanges) data:

• Volume: number of market participants and the size of their TXs

• Turnover Rate: rate of outstanding volume changing hands

• Average Trade Size: average cost of trade; ratio of daily volume to trade count

I was not able to compile the following measurements with on-chain data, and would love to find a source for doing so for future research:

• Market Depth: measures market breadth; the market's ability to settle large trades

• Bid v. Ask spread: the gap between the highest bid and lowest ask price on an order book

This is due to the fact that limit order book data does not exist for AMMs, central limit order books use a totally different mechanism for matching buyers to sellers.

This trading mechanism is not typically used on-chain due to discrepancies with transaction cost and throughput on L1 chain solutions compared to centralized central limit order book exchanges.

Alternative measures for liquidity need to be employed in their absence using price and volume data which is more readily available.

Proposed by Hasbrouck and Schwartz (1988) [1], this is the ratio of the variance of long period returns and the product of the variance of short period returns by the number of short periods (e.g. days) used for this metric.

Values slightly below one denotes high liquidity levels, values significantly below one denotes low liquidity. This short period volatility could be due to (implicit) execution costs (e.g. spread, slippage), which this metric assumes.

This metric also builds on the intuition proposed by Amihund and Mendelson (1986) [2] that, ceteris paribus, bid-ask spread size is proportional to returns.

Danyliv et. al (2014) [3] shows how suitable this metric is for measuring the impact of a price change relative to a given volume, aka price impact. For this, they propose the Liquidity Index (LIX).

They state that low liquidity assets would have a LIX of ~5, assets with high liquidity typically have a LIX of ~10.

Its main contribution is estimating how much volume is required to move the price of an asset by $1, calculated as 10 raised to the LIX value. This is useful for potential traders seeking a strategy for buying/selling an asset.

A few studies were consulted for finding the best estimators for bid-ask spread to measure the tightness of crypto assets on-chain.

The best performing ones seem to be the Tobek (2016) [4] Volume over Volume estimator based on daily high/low price data (VoV_HL) and the Corwin & Shultz (2012) [5] spread estimator (CS) which is also based on high/low prices.

However, when testing these metrics with on-chain DEX data via Dune Analytics, DAO native tokens which seem to have vastly different liquidity levels were found to have CS and VoV_HL values very low and close to each other.

This suggests that the trading mechanism used may be inefficient in minimizing tx costs, or that these estimators are not well suited for AMM data. Further research will be required before these estimators can be used to measure liquidity.

As mentioned earlier, the difficulty in measuring implicit tx costs using low frequency measures may be due to their ineptitude as a liquidity measure or inefficient trading mechanisms which make implicit tx costs high.

Next steps then would be to compare these measurements to daily price and volume data from central limit order book exchanges (e.g. via Kaiko data or dYdX trading data) and compare them to trading data from AMMs.

A lot of research has been conducted in finding ways to measure liquidity in financial markets using low frequency measurements.

If these measurements are accurate and can be applied to DAO native tokens, it gives potential buyers and sellers for these assets more information on how to strategize a trade or choose a trading mechanism, respectively.

Reducing the friction between buyers and sellers leads to increased liquidity for these assets. It is also key to promoting decentralization in the ownership of DAOs as the market for their native token is better able to attract a more diverse set of buyers.

So, do these tokens suffer from low liquidity, can this issue be remedied with the existing DeFi trading mechanisms? Or are we still waiting for the dawn of a new trading mechanism that can effectively reduce this friction?

[1] J. Hasbrouck and R. A. Schwartz, “Liquidity and execution costs in equity markets,” The Journal of Portfolio Management, vol. 14, no. 3, pp. 10–16, Apr. 1988. doi:10.3905/jpm.1988.409160

[2] Y. Amihud and H. Mendelson, “Asset pricing and the bid-ask spread,” Journal of Financial Economics, vol. 17, no. 2, pp. 223–249, Dec. 1986. doi:10.1016/0304-405x(86)90065-6

[3] O. Danyliv, B. Bland, and D. Nicholass, “Convenient liquidity measure for financial markets,” SSRN Electronic Journal, Dec. 2014. doi:10.2139/ssrn.2385914

[4] K. Y. Fong, C. W. Holden, and O. Tobek, “Are volatility over volume liquidity proxies useful for global or US research?,” SSRN Electronic Journal, Jun. 2017. doi:10.2139/ssrn.2989367

[5] S. A. CORWIN and P. SCHULTZ, “A simple way to estimate bid‐ask spreads from daily high and low prices,” The Journal of Finance, vol. 67, no. 2, pp. 719–760, Mar. 2012. doi:10.1111/j.1540-6261.2012.01729.x