Exchange Liquidity Ranking

Order book depth is a fairly new metric to compare exchanges. It measures the volume of orders in exchanges’ order books and is intended to be a better comparison method of exchanges’ liquidity than a ranking by trading volume. I introduced order book depth in January 2019 and also analyzed in a follow-up article that there is no correlation between trading volume and order book depth. In this article, I want to apply this comparison method and present a ranking of exchanges based on a data snapshot I collected recently.

Fortunately, with Coingecko and Coinpaprika, two exchange ranking sites have stepped up this year and implemented order book depth as an additional measurement for liquidity. This empowers users to compare liquidity across exchanges in a much simpler way than to open the exchange and compare order books manually. Since Coingecko integrated the order book depth feature for 50 more exchanges than Coinpaprika, I chose to use Coingecko’s data for this analysis.

I made a snapshot of all exchanges’ order book depth as displayed on Coingecko on August 25, 2019. Of a total of 354 exchanges listed on Coingecko, 183 exchanges’ order books were integrated into Coingecko. Of these 183 exchanges, I excluded 4 exchanges because they are (partly or entirely) using the same order book than a larger exchange: Upbit has Bittrex’ order book integrated into their exchange and Bequant is using HitBTC’s order book. The order books of Huobi Korea, Huobi US and Huobi Global also looked similar. Therefore, I decided to exclude Upbit, Bequant, Huobi US and Huobi Korea from this analysis and only display the order book depth of Bittrex, HitBTC and Huobi Global. Furthermore, I decided to manually adjust HitBTC’s order book depth, since their reported liquidity was way higher than their balance in hot and cold wallets suggested. This leaves us with a dataset of 179 exchanges.

Coingecko offers data on order book depth of the buy side and sell side within a 2% range of the current price. So, “-2% depth” is the sum of all buy orders in USD not further away than 2% of the current price, while “+2% depth” is the sum of sell orders in USD not further away than 2% of the current price. Just to give you an example: With a market sell of BTC worth 4.3 million USD on Bitstamp at the time of the snapshot you would have hit all buy orders, that were not lower than 98% of the current price.

While Coingecko offers both buy orders and sell orders on integrated exchanges, I decided to only use buy orders for this analysis, since it seemed like most coins and tokens, especially the ones with smaller market caps, had way more orders on the sell side than on the buy side. To compare exchanges, I therefore used the sum of all buy orders on an exchange within a range of 2% of the current price in each trading pair.

Let’s start with the lowest liquidity exchanges. 22 exchanges of the considered 179 exchanges had less than a total of 10k USD in buy orders within 2% of the current price across all trading pairs. Most of them offered hardly any liquidity at all, while others only had decent liquidity in trading pairs with their own exchange token. See all 22 exchanges in the following graphic:

45 of the 179 considered exchanges had a total sum of buy orders within 2% of the current price across all trading pairs between 10k USD and 100k USD. This list contains decentralized exchanges such as Binance DEX or Switcheo but also newer exchanges like Beaxy. You can also find exchanges in this list, that claim to have a daily trading volume of over 100 million USD, but their combined order book depth implies, that most of their trading volume is artificial. See the results:

Of the 179 considered exchanges, 61 had a total sum of buy orders within 2% of the current price across all trading pairs between 100k USD and 1 million USD. Those are already respectable numbers, especially given that some of these exchanges offered just a small number of trading pairs. Coincheck is one of those exchanges, that had only one trading pair listed (BTC/JPY) but buy orders worth 773k USD within 2% of the current price. Other exchanges were building up a bigger total of buy orders across many trading pairs (for example Hotbit or CREX24 with ~660 trading pairs). Find the ranking below:

41 exchanges in the data sample had a sum of buy orders within a 2% range of the current price between 1 million USD and 8 million USD. Those numbers already indicate a very high liquidity per exchange. Again, some of these exchanges had this amount of liquidity even though they only offered a few trading pairs (itBit, Gemini or Bitflyer), while others built up a higher order volume with hundreds of different trading pairs (e.g. HitBTC, Gate.io or Kucoin). See the ranking below:

The TOP 10 exchanges in this ranking all had buy orders within 2% price range worth 8 million USD or more. Binance offered by far the highest liquidity with buy orders worth 51 million USD. Kraken was placed in the second spot while Huobi Global was placed third. Bittrex, OKEx and Liquid also had a total of over 10 million USD in buy orders, while Bitstamp, Digifinex, Bitfinex and Coinbase Pro all had buy orders over 8 million USD. Find the results in the graphic:

Ranking exchanges by overall order book depth with data provided by Coingecko reveals the exchanges with high liquidity way better than a ranking by (inflated) trading volume. Most exchanges seem to be honest about the liquidity in their order books, although a few of them might have inflated their books with fake orders.

The results of this analysis display, that liquidity of an exchange can be divided among few very liquid pairs or many different smaller liquidity pairs. It can be discussed, whether the sum of buy orders per exchange (within 2% of the current price) is the best comparison method – one could also use the average or median buy order volume per trading pair. However, such a ranking would favor exchanges that only list few selected high market cap coins, while it would disadvantage exchanges with many different small cap coins. Therefore, I think the sum of buy orders within a different range is the better comparison method.

This analysis can’t guarantee, that the collected data on order book depth is 100% correct. Some exchanges might have inflated their numbers or there might have been errors in the data collection process, since order book depth is still a new feature. Adding to that, even if the numbers are correct, order books change every second and therefore the numbers in this analysis might be different from current order books.

Furthermore, a high ranking in this analysis only indicates a high reported liquidity but doesn’t necessarily indicate whether an exchange is great from an overall perspective. A high-liquidity exchange can still block withdrawals or get hacked at any time. I would advise everyone to DYOR before using an exchange and to not keep too many funds in the exchanges’ wallet.

The recent implementation of order book depth per trading pair and per exchange on different coin comparison sites enable users to rank exchanges by liquidity instead of ranking them by inflated trading volumes. In this analysis, buy orders not further away than 2% of the current price of a coin were collected from 179 exchanges and the exchanges were then ranked by the sum of those buy orders per exchange. The results show, that order book depth can differ vastly per exchange and I hope this article emphasizes the importance of this new metric.