The data is good for the stock market of the sudden market: Which is the liquidity of the exchange?

This paper analyzes and compares the liquidity of major exchanges on April Fool's Day.

In the short time from 12:30 to 13:22 Beijing time on April 2 (Beijing Pacific Summer Time, April 1st), the price of Bitcoin rose 21% to $5,079, since November 2018. The highest level since. Since then, the entire encryption community has been screaming that the bull market is coming.

Understandably, the soaring prices have caused many people to re-interest in Bitcoin, leading to soaring market demand. As prices climb, more and more people are madly trying to log in to exchanges and apps, trying to catch up with this round of bounces, hedging risks or selling tokens in their hands to make a profit. As the shorts were closed and the buy stop orders were triggered, the price rose faster and faster, resulting in a snowball effect on the buying. The large purchase or sale of bitcoin usually leads to large price fluctuations, and the mode of operation of individual investors following the trend (chasing up and down) and the withdrawal of market makers will also increase the volatility.

Therefore, it is very important for exchanges to provide investors with sufficient liquidity. The liquidity problems exposed in this upswing provide us with an excellent opportunity to assess the true quality of the exchange.

1_PtG1cpsMTMzad5IbEaQBJA

Image source: bitcoinist ; cryptobasicpodcast

How do the major exchanges respond to the market?

After the rebound of Bitcoin (BTC), it is time to take a deeper look at the response of major exchanges. According to the hourly k-line of the BTC, we can easily see that at 12 o'clock on April 2, Coin recorded a total of 12,384 BTC transactions in one hour, and soon reached 21,248 at 13 o'clock. BTC. For comparison, the previous hour's trading volume was only 1,510 BTCs. Similarly, at 12 o'clock on April 2, Coinbase-Pro traded 6,889 BTC, Bitstamp was 3,798, and Kraken was 4,121 BTC, both an order of magnitude or more.

1_PB4gePebzppZ1cJkb9s5Jw

The currency is on the hourly K line of the BTC/USDT transaction pair

In fact, as shown in the chart above, the trading volume of almost all major exchanges has soared 10 times after 12:00.

Although from a qualitative point of view, the volume of transactions of major exchanges is similar, we also need to pay attention to how much the trading volume of each exchange has increased and how much the order book has decreased (ie the price impact in different exchanges). ).

1_hVZ1SWiONrVaXA3U1_RaDg

The change in the hourly trading volume of the exchange recorded before and after the bitcoin rise. On mainstream exchanges, trading volume has increased more than 10 times

The volume of transactions is the total amount of transactions the trader makes in a particular market trading window. It sometimes reflects the potential volatility of the market. The high volume of trading indicates that there are two groups of people who have very different opinions on future market trends. Trading volume data can be collected from a variety of sources.

Price impact is the correlation between an order received (buy or sell) and subsequent price changes. It can be measured by several different statistical models, such as the famous Kyle's lambda or Amihud model. In this article, we use this term broadly to describe real-time price changes caused by 10-BTC market orders that reflect the thickness of the order book. We first measure the half-price difference (half the price difference between the bid price and the ask price), then we remove the trading volume of the top 10 BTCs from the order book of both parties and measure the half-price difference again. Their differences can be used as a reasonable measure of the thickness of the order book, reflecting the liquidity of the market ( providing greater flexibility to investors with as little value loss as possible ). In other words, the lower the price impact, the more stable the market.

By observing these two aspects, we can understand which exchanges are more stable, which exchanges only perform well during the “normal” period, and once the big market is coming, they will destroy their gorgeous coats.

1_5DqxFpjaEy3XvbS6NcxC1g

Price changes of major exchanges over time

In the price impact-time chart above, we found that Gdax (Coinbase-pro), coin security, fire coins, and Liquid maintained a relatively low price impact before and after the event . In addition, the currency security performance is the most prominent, and its price impact increases the least during the market, which shows its robustness in the big market. We will do more quantitative analysis later.

Quantitative analysis and comparison of the liquidity of encrypted exchanges

Sophon Tech Inc. records L2 data (activities categorized by order) from several large exchanges around the world, including China, the United States, Japan, and Europe. We quantitatively analyze the order data before and after this April Fool's Day, and compare the exchanges from different aspects.

Observation 1: The order book is reduced to varying degrees

Some exchanges have thick order books at normal times, but they are quickly reduced in the big market. We have observed this qualitatively in the price impact chart above. However, we can better observe this by the peak price impact vs. 24-hour median price impact chart (which illustrates the normal trading and behavior during peak trading). The ideal exchange with the best liquidity and the strongest crisis resistance will be located in the lower left corner of the chart.

1_qkI_7KhqEdFLxSq9LK3d_w

Ratio between peak price impact and 24-hour median price impact

For normal time trading, the price impact of many exchanges is very low, such as currency security, Gdax, fire coins, Liquid and OKEx, which are located on the left side of the chart, and the average price impact (measured by the 10-BTC method) is basically small. At 10bps.

However, when the market appears, the price impact rises sharply, and the faster the market maker withdraws, the higher it will be. For example, for Gdax and Fire, they are one of the exchanges with less average price impact, but the peak price impact is significantly increased. In these exchanges, the order book has been reduced very quickly, and many market makers are likely to leave the exchange without hesitation.

On the other hand, a small ratio does not necessarily indicate that the exchange has a strong resilience to risk. For example, the ZB exchange's peak price impact has the smallest ratio to the 24-hour price impact, but it is still one of the less liquid exchanges with a price impact of 60 + bps.

On the contrary, the currency has a lower 24-hour median price impact, and the price impact growth during the April Fool's Day market is also lower. The resistance of the currency to sudden price changes is very prominent, prompting more market makers to insist on trading.

1_d7EvWBcI_z-WFJmHcnTnWg

Ratio between peak price impact and 24-hour median price impact

Observation 2: Trading volume is fraudulent

The problem of fraudulent trading volume in the cryptocurrency community has always existed. There are many different ways to measure this. Interestingly, with the recent surge in BTC prices, we have another way to make this behavior useless. When the market occurs, the actual trading volume will increase significantly, but the false trading volume will not. Therefore, the trading volume of these fraudulent exchanges will be diluted by the false trading volume. Therefore, we can measure the “fake” volume of each exchange by measuring the ratio of peak to 24-hour median volume.

1_vqX8EQR3fFRh70vOOaqOcQ

Ratio between peak volume and 24-hour median volume

1_DbnUvsy5vsvUjYYRaj63nA

In the above picture, it is clear that the growth rates of Liquid, Coin and Gdax are very similar, at 23.5, 22.4 and 21.0 respectively. On the other hand, the 24h median volume of OKEx, Firecoin, and Hitbtc increased by only 7-9 times.

Based on our previous arguments and assuming that the volume of Gdax, Coin and Liquid is real, we can conclude that more than half of the OKEx, Firecoin and Hitbtc transactions may be false.

In fact, the trading rate of Gate.io and Kucoin is even higher. However, because of their very small trading volume, we consider these high ratios to be unrelated outliers.

April Fool's Day Aftermath

After the BTC rose, the order book was reduced, so the price impact increased significantly. More interestingly, the price impact continues to increase on most exchanges over the next few days , as shown in the 24-hour moving median price impact chart. This BTC price increase has a long-term impact.

WX20190409-144555@2x

After the BTC up event (dashed line), the 24h hourly average price of most exchanges we recorded affected the increase in mobile median

1_fL8D95BhjKSoWPvEcmeoDQ

The price impact after the rise of BTC generally increases 1.5 to 3 times

After this event occurs, the order book continues to decrease over the next three days. This can be demonstrated by the steady increase in the price of all exchange 10-BTC prices.

This shows that market makers are still very cautious even after a few days. If this trend continues, it means that the price fluctuates sharply, which may lead to another big price trend.

in conclusion

Through the analysis of BTC April Fool's Day, we observed the true quality of the current cryptomarket exchanges, which we may not be able to observe so clearly under normal circumstances.

Some exchanges have greater liquidity than other exchanges (such as currency security and Gdax). In addition, the different ratios of increased transaction volume indicate that there may be trading volume fraud issues from multiple exchanges. Finally, this BTC price spike has had a long-term impact on market makers, and they are very reluctant to put orders back to the top of the order book.