I study the effect of a change in positive and negative sentiment on the weekly change in different Cryptocurrency returns. For my regression I'm using lagged values of the independent variables to determine the effect on the Cryptocurrency returns for the following week. Note that Weeky returns and Realized Volatility in each row are the values for that cryptocurrency. Now to my question: Can I only say that a postive change in PSVI correlates with a positve change in BTC returns for the following week? Or am I also able to say that a change in PSVI has a stronger effect on the BTC returns than for example Doge returns? I tested for Granger causality and found that PSVI granger causes all Crypto returns while NSVI does not granger cause returns besides for XRP.
Would appreciate your help!
My Regression looks as follows:
===========================================================================================
Dependent variable:
------------------------------------------------------------
BTC_Ret ETH_Ret Binance_Ret XRP_Ret Cardano_Ret Doge_Ret
(1) (2) (3) (4) (5) (6)
-------------------------------------------------------------------------------------------
L(PSVI, 1) 0.064*** 0.084*** 0.077*** 0.044*** 0.026*** 0.013***
(0.018) (0.024) (0.027) (0.012) (0.009) (0.004)
L(NSVI, 1) -0.045** -0.046* -0.057** -0.039*** -0.020** 0.003
(0.018) (0.025) (0.027) (0.012) (0.009) (0.004)
L(Weekly_returns, 1) 0.076 0.073 0.019 -0.021 0.116* -0.038
(0.070) (0.069) (0.075) (0.034) (0.068) (0.030)
L(Real_Volatility, 1) 0.105 0.173 0.292** -0.237** 0.019 -0.116*
(0.136) (0.148) (0.121) (0.099) (0.087) (0.067)
Constant -0.005 -0.015 -0.020 0.007 0.0002 0.002
(0.015) (0.020) (0.018) (0.007) (0.005) (0.002)
-------------------------------------------------------------------------------------------
Observations 207 207 207 207 207 207
R2 0.074 0.071 0.066 0.118 0.059 0.088
Adjusted R2 0.055 0.053 0.048 0.101 0.040 0.070
Residual Std. Error (df = 202) 0.104 0.138 0.152 0.069 0.053 0.023
F Statistic (df = 4; 202) 4.013*** 3.871*** 3.596*** 6.770*** 3.159** 4.872***
===========================================================================================
Note: *p<0.1; **p<0.05; ***p<0.01
