How is CoinMarketCap Combating Fake Crypto Trading Volumes with its New Liquidity Metric?
Crypto data aggregator, CoinMarketCap (CMC), held its inaugural large-scale event, The Capital, bringing together leading stakeholders in the blockchain and cryptocurrency space. Held at the Victoria Theatre in Singapore on Nov. 12-13, CMC released its new metrics for assessing liquidity for the industry to solve the problem of volume inflation caused by crypto exchanges and market makers. Blockchain.News sat down with Carylyne Chan, the Chief Strategy Officer of CMC in Singapore, for an interview on the new liquidity metric and other initiatives CMC is putting forward for a more transparent crypto ecosystem in terms of data.
Liquidity metric: Combating inflated volume reporting by crypto exchanges
It was announced at The Capital that CMC has revealed that the website has released a new metric to rank crypto exchanges by using liquidity in addition to the ranking by trading volume. The change in ranking methodology focused on combating inflated volume reporting by crypto exchanges. Chan believed that the previous solutions such as picking a few “trusted” exchanges or unscientific correlations like web traffic were not comprehensive to address the root cause of the issue.
Chan stated that there are three components to the liquidity metric system, the distance of the orders from the mid-price, the size of the orders, and the relative liquidity across different market pairs. The liquidity metric has been designed to measure liquidity adaptively while polling market pairs at random intervals over 24 hours and the result is based on the average. This way, time zone differences and change in order book depth due to immediate market conditions would be accounted for.
Chan further explained, “Why we got to liquidity in the first place is because liquidity is the most important aspect for traders. For us, we are moving away from volume; the volume has lost a little bit of its purpose to gauge real trading interest. Liquidity metric itself can help us to ensure that we account for orders that are close to the mid-price and higher, and orders that are further from the mid-price, we discount that market.”
On the relative liquidity aspect across different market pairs, Chan said: “When we think about liquidity across different digital assets, we have to account for liquidity for different types of assets. For example, if you have a BTC/USDT paired on a very liquid exchange, its liquidity will be higher than if you have a much more illiquid market, let's say that 2000th ranked coin on another exchange. What we’re really trying to say is that we can account for each of these different assets, because the liquidity metric adapts to each of them based on how the absolute liquidity is.” There will be three phases on the new Liquidity metric on CMC’s website: firstly on the ranking of trading pairs, followed by exchanges and finally, cryptoassets.
Chan explained that CMC’s acquisition of the Hashtag Capital team was for the company’s pricing algorithm. “When I described the graph-based algorithm, the genesis came from the Hashtag team. After we brought them in, we also did a lot of research with them on liquidity. They have also been critical in the creation of the Liquidity metric,” said Chan. She also added that the Hashtag team had made very “big contributions” there.
CMC’s graph-based algorithmic pricing model
Regarding CMC’s graph-based algorithmic pricing model, Chan said that the current way of pricing is mostly based on volume. “We look at the volume of that particular market pair and how much it contributes to volume as a percentage of everything, and we average out the price (i.e., volume-weighted average price),” Chan elaborated.
Under the new graph-based model, the way that pricing works will be different. “We put all of the assets in a graph model. We start mapping the relationship between each of them to see how each asset’s price affects another’s price.” She also mentioned that regression and solving simultaneous equations are used to investigate how the models could be mapped out. “We do that over and over across multiple iterations until we get to the best and the most stable price.”