Perpetual swaps has a funding mechanism which keep the perpetual price in line with the spot price. When funding is positive long positions pay shorts and when funding is negative shorts pay longs. Funding mechanism works hourly in FTX, while it is executed in 8 hour blocks in other exchanges.
In general, traders would like to utilize the fat tails of being long as much as possible especially in bull markets. Instead of buying the spot, they can buy the perpetuals futures with leverage which creates temporary price deviations. These deviations lead to higher funding rates. But does it make sense to maximize the gainz of obvious pumps while paying the high funding fee or is it just an overcrowded trade that is waiting to be shaken out?
In the above figure, orange lines are the hourly funding values and blue lines are the hourly closing prices of the ETH-PERP. At first glance it looks like the price dips correspond to below 0 funding rates. In January, most of the high funding regions marked the local price tops, however in February highest funding regions resulted in the continuation of price increase. We can calculate what would be the percentile return(for long positions) after a few hours for each hourly data point and do a scatter plot of funding rate versus the percentile returns. Price charts are shown for the 2021–01 to 2021–04 period for easier visibility while the scatter plot below uses the data between 2020–09 to 2021–04.
In Figure 2, you can see that most of the data points clustered around 0,0 point. As the funding rate increases to the right of the graph, returns are not clustered around 0. There are high returns and very low returns therefore we can suggest that there is a higher volatility when the funding rate is high. There is also a line showing the linear relationship between the funding and the return, it is almost horizontal however it is a little tilted downwards with a small negative slope which suggests a small inverse relationship between the funding rate and the returns for ETH-PERP.
We can calculate the slope of the linear relationship between the funding rate and the returns. Resultant slope coefficients for the major perpetual contracts are given below:
The most inverse relationship between the returns and the funding is with SUSHI-PERP while the most positive relationship is with the TRX-PERP. We can investigate both of these pairs further.
As you can see in the scatter plot there are extreme cases in SUSHI-PERP scatter plot that has highly negative funding rate and have very high positive returns, which is the reason for negative slope. Extreme sell pressure got absorbed at these points and price bounced very quickly.
Scatter plot of the TRX-PERP is more circular, there arent any outliers or extremely high/low funding that is distrupting the market order. One reason for this might be that there arent strong emotional trades for this contract yet. Each FOMO buy is accompanied by a forced panic sell at a price wick. If there are no FOMO buys or panic sells there is no pressure to absorb.
If you check the Figure 3, in general the coins on the left side of the x-axis have pumped far more than the coins on the right side except for a few outliers. They had more momentum that is accompanied by overleverage. When these highly appreciated coins such as SUSHI dumped, sell pressure got absorbed far more than the relatively not as much appreciated coins such as TRX. At the end, it all comes to the simple idea of “buying the dip”.
As a final comparison of BTC, ETH and the majors listed in Figure 3, we can put each funding value in bins and check the average return of value of futures in aggregate:
It can be easily seen in the chart that highest funding rates with high buy pressure does not end well for longs especially for BTC and ETH futures, whereas the mean returns for longs are highest when the sell pressure is at its peak at the most negative funding rates.
Winners buy the dips.