Does herding behavior exist in cryptocurrency market?

Figures & data

Table 1. Descriptive statistics (1 January 2015–31 January 2019)

Name	Ticker	Weight (MVIS)	Obs.	Mean	Std. Dev.	Kurtosis	Skewness
Bitcoin	BTC	59.3%	1,495	0.001712	0.047095	141.2214	−6.395736
XRP	XRP	12.3%	1,495	0.003793	0.088477	135.4764	6.140011
Ethereum	ETH	10.7%	1,277	0.004904	0.078463	33.34436	−1.397016
Bitcoin Cash	BCH	2.0%	561	0.000271	0.104559	22.66432	−.3199769
EOS	EOS	2.0%	583	0.004826	0.12155	73.06374	4.053202
Tether	USDT	2.0%	1,440	−0.00057	0.034552	604.0049	−14.27206
Litecoin	LTC	1.9%	1,495	0.002851	0.067532	48.28515	−.5584429
TRON	TRX	1.7%	509	0.010598	0.145112	28.8541	2.668094
Stellar	XLM	1.5%	1,495	0.004441	0.091978	45.18002	2.801441
Bitcoin SV	BSV	1.1%	87	−0.00089	0.193697	14.35462	−.4281331
Cardano	ADA	1.0%	491	0.003844	0.12288	50.44482	3.245584
Binance Coin	BNB	0.8%	559	0.010287	0.111274	27.94556	.8317939
Monero	XMR	0.7%	1,495	0.004881	0.077902	32.57896	.1744137
IOTA	IOT	0.7%	601	0.000806	0.098609	22.69746	−1.066513
Dash	DASH	0.6%	1,495	0.003566	0.067814	40.93346	−.9420432
NEO	NEO	0.4%	878	0.007126	0.116592	31.5135	2.063862
NEM	XEM	0.4%	1,405	0.006976	0.10272	68.37551	3.845374
Ethereum Classic	ETC	0.4%	925	0.005228	0.134577	362.9901	14.61604
ZCash	ZEC	0.3%	828	−0.00111	0.10966	107.281	4.750347
Tezos	XTZ	0.2%	490	−0.00172	0.097341	33.2143	−1.186083
MVIS large	–	100%	1,492	0.00249	0.041374	9.123678	−.1035642

Figure 1. MVIS variation (1 January 2015–31 January 2019)

Figure 2. MVIS variation (1 January 2015–2 January 2016)

Figure 3. MVIS variation (3 January 2016–28 February 2017)

Figure 4. MVIS variation (1 March 2017–31 January 2019)

Table 2. Regression results of CSAD_t

Period	${\alpha }_0$	${\alpha }_1$	${\alpha }_2$	R^2
1 January 2015–31 January 2019	0.0337629*	0.7216843*	1.224422*	0.436*
1 January 2015–2 January 2016	0.0310459*	0.5450117*	1.948044*	0.5461*
3 January 2016–28 February 2017	0.0299549*	0.6549929*	1.966564*	0.3244*
1 March 2017–31 January 2019	0.0423536*	0.6103543*	1.606074*	0.3767*

*Significant at level 5%.

Figure 5. Scatterplot of Equation (3)(3) $CSA{D}_t={\alpha }_0+{\alpha }_1\left|R_{m,t}\right|+{\alpha }_2{R}_{m,t}^2+{\epsilon }_t$(3) estimation’s residuals

Table 3. BDS test

Dimension	BDS Statistic	Std. Error	z-Statistic	Prob.
2	0.029823	0.002665	11.19015*	0.0000
3	0.060749	0.004235	14.34584*	0.0000
4	0.081469	0.005043	16.15490*	0.0000
5	0.093226	0.005257	17.73223*	0.0000
6	0.097566	0.005072	19.23729*	0.0000

*Significant at level 1%.

Figure 6. The coefficient ${\alpha }_2$based on rolling window (30 observations)

Figure 7. T-statistics based on rolling window (30 observations) and 5% critical value

Table 4. Herding periods

Year	FromTo
2015	 31 May–29 June 28 August–26 November
2016	25 February–20 November
2017	20 May–18 July
2018	14 February–14 April 14 June–11 September

Table 5. Estimation of logistic model

Comfort index	Constant	LogL	LR chi2	P-value
−.0653062*	3.330517*	−142.6786	8.87*	0.0029

*Significant at level 1%.

Table 6. Estimation of logistic model

Inflation	Crude oil	Constant	LogL	LR chi2	P-value
−.5460505	−.1302266	4.012242*	−139.41828	1.98*	0.0071

*Significant at level 1%.