Skip to main content

Table 6 Robustness tests using quantile regressions

From: The witching week of herding on bitcoin exchanges

 

Before expiration

 

After expiration

 

Rm2 Dexp

p-value

Rm2 (1 − Dexp)

p-value

Rm2 Dexp

p-value

Rm2 (1 − Dexp)

p-value

D0

− 0.013186

0.91

0.106122***

0.00

    

D1

− 0.142918

0.29

0.113435***

0.00

0.981784

0.23

0.107023***

0.00

D2

− 0.170789***

0.00

0.113514***

0.00

0.782161

0.34

0.107138***

0.00

D3

− 0.085632

0.25

0.112832***

0.00

− 0.043864

0.97

0.107066***

0.00

D4

− 0.089831

0.11

0.112729***

0.00

− 0.099513

0.79

0.106936***

0.00

D5

− 0.085813

0.15

0.111250***

0.00

0.367007

0.78

0.107200***

0.00

D6

− 0.055984

0.29

0.112948***

0.00

0.661392

0.59

0.107234***

0.00

D7

− 0.055442

0.33

0.111882***

0.00

− 0.425801***

0.00

0.106638***

0.00

D8

− 0.090121**

0.02

0.112287***

0.00

− 0.410641***

0.00

0.106639***

0.00

D9

− 0.128446*

0.05

0.113445***

0.00

− 0.337711

0.39

0.106667***

0.00

D10

− 0.128106**

0.02

0.113516***

0.00

0.034026

0.53

0.114780***

0.00

D11

− 0.097982**

0.01

0.112689***

0.00

0.044743

0.18

0.113423***

0.00

D12

− 0.089636**

0.02

0.112217***

0.00

0.080440

0.27

0.113600***

0.00

D24

− 0.128187***

0.00

0.113447***

0.00

0.035516

0.23

0.107183***

0.00

D36

− 0.098464***

0.00

0.112197***

0.00

0.059866***

0.00

0.105874***

0.00

D48

− 0.087003***

0.00

0.110951***

0.00

0.060119***

0.00

0.113675***

0.00

D60

− 0.083854***

0.00

0.110769***

0.00

0.058194***

0.00

0.113647***

0.00

D72

− 0.054904

0.84

0.111672***

0.00

0.056161***

0.00

0.114722***

0.00

D84

− 0.045361

0.85

0.110657***

0.00

0.059461***

0.00

0.113765***

0.00

D96

− 0.078991***

0.00

0.109752***

0.00

0.060059***

0.00

0.113441***

0.00

D108

− 0.081603***

0.00

0.109956***

0.00

0.059868***

0.00

0.113845***

0.00

D120

− 0.035914

0.88

0.110315***

0.00

0.059316***

0.00

0.113899***

0.00

D132

− 0.040240

0.86

0.110850***

0.00

0.061740***

0.00

0.113383***

0.00

D144

0.005853

0.61

0.111669***

0.00

0.063808***

0.00

0.111452***

0.00

D150

0.004261

0.71

0.112111***

0.00

0.064497***

0.00

0.110977***

0.00

  1. The table shows the estimates of Eq. (3) including five lags of CSAD \(CSAD_{t} = \gamma _{0} + \gamma _{1} D_{{\exp }} \left| {Rm_{t} } \right| + \gamma _{2} (1 - D_{{\exp }} )\left| {Rm_{t} } \right| + \gamma _{3} D_{{\exp }} Rm_{t}^{2} + \gamma _{4} (1 - D_{{\exp }} )Rm_{t}^{2} + \varepsilon _{t}\)
  2. Dexp is the dummy variable, defined differently, that takes value 1 in specific times around expiration and 0 otherwise. Each raw contains the estimated parameters of the model for one dummy variable associated to Dexp. For example, the dummy variable D1 takes a value of 1 one hour before (after) expiration and 0 otherwise; the dummy variable D2 takes a value of 1 2 h before (after) expiration and 0 otherwise and so on, until D150 that takes a value of 1 150 h before (after) expiration and 0 otherwise. Results using Newey–West heteroscedasticity and autocorrelation consistent estimators. ***, **, * indicate significance at 1%, 5% and 10% respectively