Volume-Weighted Cap

A closer look at ranking coins using the Volume-Weighted Cap (VWC) methodology.

In a previous blog I looked into Bitcoin’s Market Dominance, finding it dominates over 80% of the market with liquidity taken into account.

Feedback from the blog was collected:

  1. Formalising a method of using and displaying VWC
  2. Accounting for wash trading, aka, fake volume
  3. Analysing all 2000 coins, instead of just Top 100
  4. Considering a larger date range for market dominance

This follow-up article treats these four points.

1. Formalising VWC

Volume-Weighted Cap accounts for liquidity by simply including it as a multiplier.

VWC = Market Cap * Liquidity

  • Market Cap should be circulating supply * price ($).
  • Liquidity is any form of standardised liquidity metric ($).

The VWC unit of account is naturally , but since this is hard to comprehend a VWC Index should be used instead. Simply divide the VWC of each coin by the total sum of all VWCs in the range tested. This leads to a much better ranking system for coins, since low volume coins place much lower.

Liquidity be changed out with buy-support liquidity, on-chain transaction volume, aggregate reported exchange volume, or even sanitised exchange volume. Since I believe VWC follows Power Law and is scale invariant any form of liquidity will result in roughly the same Index.

2. Removing Wash Trading

Although much has been said around fake volume such as this, this and this, it is *very* difficult to remove fake volume from aggregator sites, as Bitwise’s report found.

It is likely that wash trading affects all coins (within a couple order of magnitudes), and altcoins to a higher extent. This only reinforces the VWC metric. Additionally since VWC should be shown on a log-scale, any sub-orders of magnitude fluctuations should be minimised.

3. Analysing all Coins

CoinMarketCap has 2098 coins as at 24 March 2019, ranking Bitcoin at No 1. The lowest 350 coins had either no reported supply or volume so were not included.

I took 4 weeks of results of the Top 2000 coins and aggregated them together to remove outlier events.

The S Curve returns

Once again the mysterious S curve first noticed in the original analysis appeared. By testing different ranges, it was observed to be scale invariant. No matter the range tested, the S curve was always obvious.

To me, this highlighted a few things. Firstly that VWC is a naturally sound metric since it behaves fractally and is not disorderly on a log-scale. Secondly, it meant that my original hypothesis that it was a “Page 1 effect” was wrong, since it occurs across any range.

I attempted to model it, but it became very difficult. It isn’t strictly a Sigmoid Curve (classical S curve) since it 1) doesn’t correspond to the Y-axis and 2) occurs on a log-scale (not linear). It also wasn’t a Logit Curve, or an Inverse S Curve. Definitely over my head in maths at that point, I will leave modelling it to a future mathematician.

The VWC Index

The Top 25 were highlighted. Since the results are aggregrated from 4 weeks of results, the final ranking is indicated of sustained performance over a month.

The VWC Index

Of note:

  1. Bitcoin commands 86% of the market
  2. Ethereum is roughly 10% of Bitcoin
  3. Tether jumps 4 places to rank at #3 — due to very high sustained volume
  4. BNB keeps its position at #10. This may be indicative that BNB volume is healthy to its market cap.
  5. Monero, Cardano and NEM are relatively illiquid compared to their immediate peers.
  6. Any coin outside the Top 25 is chronically illiquid.

4. Expanding the Date Range

The final exercise was to expand the Date Range to consider data for 5+ years.

VWC Index since 2014

Since anything below the Top 5 is negligible, the data was limited to only Bitcoin, Ethereum, Ripple, Litecoin and “Others” which represented all altcoins.

Bitcoin maintained greater than 95% until 2017, when it dropped to 85%. Since then it has regained market share to 86%.

Summary

Exchanges and Coin Aggregrator sites should stop ranking coins based on a Market Cap which is vulnerable to manipulation, and instead switch to Volume-Weighted Cap. Any liquidity metric can be used in calculating VWC.

VWC on a log-scale mostly removes the influence of fake volume, though confirmed volume sources should be used as the rule.

VWC conforms to a higher Power Law, is orderly on a log-scale and is scale invariant. I was unable to find the actual mathematical model however.

VWC Ranking corrects placement for illiquid coins amongst their peers.

VWC Market Dominance over the last 6 years shows Bitcoin is really the only serious coin on the market.

Finally, Market Dominance metrics should definitely switch to the VWC Index, since ignoring liquidity is misleading.


Follow me on twitter, I research and write about Bitcoin.

https://twitter.com/jpthor__