Abstract
This paper applies a new proposed multivariate score-type test against spurious long memory to a group of cryptocurrency market returns. The test statistic developed by Sibbertsen et al. (J Econ 203(1): 33–49, 2018) is based on the multivariate local Whittle likelihood function and is proven to be consistent against the alternative two cases of random level shifts and smooth trends. We apply the test to the returns, absolute returns, and modified absolute returns. Overall, the recently developed test statistic fails to reject the null hypothesis of true long memory for most cryptocurrencies, except for the Stellar market. Therefore, applying the new test statistic supports the argument that the long memory in the cryptocurrency markets is real and is not a spurious one. Our results are further supported by applying other consistent local Whittle methods that allow for the estimation of the memory parameter by accounting for the presence of perturbations or low-frequency contaminations.
Similar content being viewed by others
Availability of data and materials
The data used for the study and codes are available from the corresponding author upon request.
Notes
Similar to checking for a break in persistence, it is worth mentioning that some tests have also been proposed for checking for a change in mean to account for the case of long-memory processes. A review is given in Wenger et al. (2019). Examples of the tests fall under the CUSUM principle and include the CUSUM-LM test of Horváth and Kokoszka (1997) and Wang (2008), the (simple) CUSUM test based on fractionally differenced data by Wenger et al. (2018), and the CUSUM fixed bandwidth tests by Wenger and Leschinski (2019). We thank a referee of pointing at this kind of tests.
Sibbertsen et al. (2018) show via simulations that the test works well in the low non-stationary range for d. The test is also proved not to be requiring Gaussianity, allowing for conditional heteroskedasticity, not requiring a precise specification of the low-frequency contamination and displaying high finite sample power results compared to competing tests (see Leccadito et al. 2015).
Estimators are provided in Appendix.
Another approach that tries to approximate the perturbation by a constant rather than a polynomial is suggested by Hurvich and Ray (2003) who proposed the local polynomial Whittle estimator with noise.
Hou and Perron (2014) provide an additional extension of the modified local Whittle estimator to account for both the perturbation and the low frequency contamination.
Other proxies of volatility were used in the literature and include the log-squared returns ln(rt2 + a) as suggested by Stărică and Granger (2005) with the parameter a being close to zero such as 0.001.
It is worth noting that a nonlinear process such as a Markov-switching model (MSW) and threshold autoregressive models can be mistaken for a long memory process (see Diebold and Inoue (2001) and Kuswanto and Sibbersten (2008). Studies conducted by Hyung and Franses (2005), Chung (2006), Baillie and Kapetanios (2007, 2008), and Kuswanto and Sibbersten (2008) have considered forecasting performance from models with both long memory and nonlinear features such as Smooth Transition Autoregressive (STAR), Self-Exciting Threshold Autoregressive (SETAR), Markov-switching (MSW) and artificial neural network. The proposed tests in this paper have some limitations in terms of accounting for nonlinearity in the data generating process. We thank a referee for pointing at this shortcoming.
Other approaches that try to approximate the perturbation by a constant rather than a polynomial are, for example, suggested by Hurvich and Ray (2003) who proposed the local polynomial Whittle estimator with noise (LWN). This estimator is nested in the LPWN estimator if the polynomials are chosen of order zero.
References
Abuzayed B, Al-Fayoumi N, Charfeddine L (2018) Long range dependence in an emerging stock market’s sectors: volatility modelling and VAR forecasting. Appl Econ 50(23):2569–2599
Achard S, Gannaz I (2016) Multivariate wavelet whittle estimation in long-range dependence. J Time Ser Anal 37(4):476–512
Achard S, Bassett DS, Meyer-Lindenberg A, Bullmore E (2008) Fractal connectivity of long-memory networks. Phys Rev E 77(3):036104
Al-Yahyaee K, Walid M, Seong Min Y (2018) Efficiency, multifractality, and long-memory property of the Bitcoin market: a comparative analysis with stock, currency, and gold markets. Fin Res Lett 27:228–234
Andrews D, Sun WK (2004) Adaptive local polynomial whittle estimation of long-range dependence. Econometrica 72:569–614
Anyfantaki S, Arvanitis S, Topaloglou, N (2018) Diversification, integration and cryptocurrency market. Bank of Greece Working Paper No. 244.
Arteche J (2004) Gaussian semiparametric estimation in long memory in stochastic volatility and signal plus noise models. J Econom 119:131–154
Aslan A, Sensoy A (2020) Intraday efficiency-frequency nexus in the cryptocurrency markets. Fin Res Lett 35:101298
Baillie RT, Kapetanios G (2007) Testing for neglected nonlinearity in long-memory models. J Bus Econ Stat 25:447–461
Baillie RT, Kapetanios G (2008) Nonlinear models for strongly dependent processes with financial applications. J Econom 147:60–71
Baillie RT, Morana C (2009) Modelling long memory and structural breaks in conditional variances: an adaptive FIGARCH approach. J Econ Dyn Control 33(8):1577–1592
Balcilar M, Bouri E, Gupta R, Roubaud D (2017) Can volume predict bitcoin returns and volatility? A quantiles based approach. Econ Model 64:74–81
Bariviera AF (2017) The inefficiency of Bitcoin revisited: a dynamic approach. Econ Lett 161:1–4
Bariviera AF (2020) One model is not enough: heterogeneity in cryptocurrencies’ multifractal profiles. Fin Res Lett 39:101649
Bariviera A, Basgall M, Hasperue W, Naiouf M (2017) Some stylized facts of the bitcoin market. Phys A 484:82–90
Bariviera AF, Zunino L, Rosso OA (2018) An analysis of high-frequency cryptocurrencies prices dynamics using permutation-information-theory quantifiers. Chaos Interdiscip J Nonlinear Sci 28(7):075511
Batten JA, Ellis C, Fetherston TA (2005) Return anomalies on the Nikkei: are they statistical illusions? Chaos Soliton Fract 23(4):1125–1136
Batten JA, Ellis CA, Fethertson TA (2008) Sample period selection and long-term dependence: new evidence from the Dow Jones index. Chaos Soliton Fract 36(5):1126–1140
Baur D, Lee A, Hong K (2015) Bitcoin: currency or investment? Working paper, SSRN: https://ssrn.com/abstract=2561183
Berkes I, Rorvath L, Kokoszka P, Shao QM (2006) On discriminating between long-range dependence and changes in mean. Ann Stat 34(3):1140–1165
Blau B (2017) Price dynamics and speculative trading in Bitcoin. Res Int Bus Fin 41:493–499
Blau B, Griffith T, Whitby R (2020) Comovement in the cryptocurrency market. Econ Bull 40(1):448–455
Bloomfield P (1973) An exponential model in the spectrum of a scalar time series. Biometrika 60:217–226
Bouri E, Gupta R, Tiwari A, Roubaud D (2017) Does Bitcoin hedge global uncertainty? Evidence from wavelet based quantile-in-quantile regressions. Fin Res Lett 23:87–95
Caporale GM, Gil-Alana L, Plastun A (2018) Persistence in the cryptocurrency market. Res Int Bus Fin 46:141–148
Charfeddine L (2016) Breaks or long range dependence in the energy futures volatility: out-of-sample forecasting and VAR analysis. Econ Model 53:354–374
Charfeddine L, Guégan D (2012) Breaks or long memory behavior: an empirical investigation. Phys A 391(22):5712–5726
Charfeddine L, Maouchi Y (2019) Are shocks on the returns and volatility of cryptocurrencies really persistent. Fin Res Lett 28:423–430
Cheah ET, Mishra T, Parhi M, Zhang Z (2018) Long memory interdependency and inefficiency in Bitcoin markets. Econ Lett 167:18–25
Chung S-K (2006) The out-of-sample forecasts of nonlinear long-memory models of the real exchange rate. Int J Fin Econ 11:355–370
Ciaian P, Rajcaniova M, Kancs A (2018) Virtual relationships: Short-and long-run evidence from Bitcoin and altcoin markets. J Int Fin Mark Inst Money 52:173–195
Corbet S, Lucey B, Yarovaya L (2018) Datestamping the Bitcoin and Ethereum bubbles. Fin Res Lett 26:81–88
Davidson J, Rambaccussing D (2015) A test of the long memory hypothesis based on self-similarity. J Time Ser Econ 7(2):115–141
Deo RS, Hurvich CM (2001) On the log periodogram regression estimator of the memory parameter in long memory stochastic volatility models. Econ Theory 17:686–710
Diebold FX, Inoue A (2001) Long memory and regime switching. J Econ 105(1):131–159
Diebold FX, Rudebusch GD (1991) On the power of the Dickey Fuller test against fractional alternatives. Econ Lett 35(2):155–160
Dolado JJ, Gonzalo J, Mayoral L (2005) What is what? A simple time-domain test of long-memory vs. structural breaks. Unpublished Manuscript, Department of Economics, Universidad Carlos III de Madrid.
Dwyer G (2015) The economics of Bitcoin and similar private digital currencies. J Finan Stabil 17:81–91
Dyhrberg AH (2016) Bitcoin, gold and the dollar: a GARCH volatility analysis. Finan Res Lett 16:85–92
Frederiksen P, Nielsen FS, Nielsen M (2012) Local polynomial Whittle estimation of perturbed fractional processes. J Econ 167:426–447
Gil-Alana LA (2004) The use of the Bloomfield (1973) model as an approximation to ARMA processes in the context of fractional integration. Math Comput Model 39:429–436
Giraitis L, Kokoszka P, Leipus R, Teyssière G (2003) Rescaled variance and related tests for long memory in volatility and levels. J Econ 112(2):265–294
Gourieroux C, Jasiak J (2001) Memory and infrequent breaks. Econ Lett 70(1):29–41
Granger CW, Ding Z (1996) Varieties of long memory models. J Econ 73(1):61–77
Granger CW, Hyung N (2004) Occasional structural breaks and long memory with an application to the S&P 500 absolute stock returns. J Emp Fin 11(3):399–421
Haldrup N, Kruse R (2014) Discriminating between fractional integration and spurious long memory. Unpublished Manuscript, Department of Economics, University of Aarhus.
Hassler U, Wolters J (1994) On the power of unit roots against fractional integrated alternatives. Econ Lett 45(1):1–5
Horváth L, Kokoszka P (1997) The effect of long-range dependence on change-point estimators. J Stat Plan Infer 64(1):57–81
Hou J, Perron P (2014) Modified local Whittle estimator for long memory processes in the presence of low frequency (and other) contaminations. J Econ 182:309–328
Hurvich CM, Ray BK (2003) The local Whittle estimator of long-memory stochastic volatility. J Fin Econ 1:445–470
Hurvich CM, Moulines E, Soulier P (2005) Estimating long memory in volatility. Econometrica 73:1283–1328
Hyung N, Franses PH (2005) Forecasting time series with long memory and level shifts. J Forecast 24:1–16
Iacone F (2010) Local Whittle estimation of the memory parameter in presence of deterministic components. J Time Ser Anal 31:37–49
Ji Q, Bourie E, Lau CK, Roubaud D (2019) Dynamic connectedness and integration in cryptocurrency markets. Int Rev Fin Anal 63:257–272
Jiang Y, Nie H, Ruan W (2018) Time-varying long-term memory in Bitcoin market. Fin Res Lett 25:280–284
Katsiampa P (2019) An empirical investigation of volatility dynamics in the cryptocurrency market. Res Int Bus Fin 50:322–335
Künsch HR (1987) Statistical aspects of self-similar processes. Proc First World Congr Bernoulli Soc 1:67–74
Kuswanto H, Sibbertsen P (2008) A study on “spurious long memory in nonlinear time series models. Appl Math Sci 2:2713–2734
Lahmiri S, Bekiros S, Salvi A (2018) Long-range memory, distributional variation and randomness of Bitcoin volatility. Chaos Solitons Fract 107:43–48
Leccadito A, Rachedi O, Urga G (2015) True versus spurious long memory: Some theoretical results and a Monte Carlo comparison. Econ Rev 34(4):452–479
Lee D, Schmidt P (1996) On the power of the KPSS test of stationarity against fractionally integrated alternatives. J Econ 73:285–302
Lobato IN (1999) A semiparametric two-step estimator in a multivariate long memory model. J Econ 90(1):129–153
Lobato IN, Savin NE (1998) Real and spurious long-memory properties of stock market data. J Bus Econ Stat 16(3):261–268
Long H, Zaremba A, Demir E, Szczygielski JJ, Vasenin M (2020) Seasonality in the cross-section of cryptocurrency returns. Fin Res Lett 35:101566
Lu YK, Perron P (2010) Modeling and forecasting stock return volatility using a random level shift model. J Emp Fin 17(1):138–156
McCloskey A, Perron P (2013) Memory parameter estimation in the presence of level shifts and deterministic trends. Econ Theory 29:1196–1237
Mensi W, Al-Yahyaee KH, Kang SH (2019) Structural breaks and double long memory of cryptocurrency prices: a comparative analysis from Bitcoin and Ethereum. Fin Res Lett 29:222–230
Mikosch T, Stărică C (2004) Nonstationarities in financial time series, the long range dependence, and the IGARCH effects. Rev Econ Stat 86(1):378–390
Mokni K, Ajmi AN (2021) Cryptocurrencies vs. US dollar: Evidence from causality in quantiles analysis. Econ Anal Policy 69:238–252
Mokni K, Ajmi AN, Bouri E, Vo XV (2020) Economic policy uncertainty and the Bitcoin-US stock nexus. J Multinatl Fin Manag 57:100656
Nadarajah S, Chu J (2017) On the inefficiency of Bitcoin. Econ Lett 150:6–9
Nakamoto S (2008) Bitcoin: a peer-to-peer electronic cash system. Unpublished Manuscript. http://pdos.csail.mit.edu/6.824/papers/bitcoin.pdf. Accessed on 26 March 2019
Nielsen MØ (2007) Local Whittle analysis of stationary fractional cointegration and the implied–realized volatility relation. J Bus Econ Stat 25(4):427–446
Noda A (2020) On the evolution of cryptocurrency market efficiency. Appl Econ Lett 28(6):433–439. https://doi.org/10.1080/13504851.2020.1758617
Ohanissian A, Russell JR, Tsay RS (2008) True or spurious long memory? A new test. J Econ Stat 26(2):161–175
Omane-Adjepong M, Alagidede P, Akosah N (2019) Wavelet time-scale persistence analysis of cryptocurrency market returns and volatility. Phys A 514:120
Osterrieder J, Lorenz J, Strika M (2017) Bitcoin and cryptocurrencies-not for the faint-hearted. Int Fin Bank 4(1):56–94
Perron P, Qu Z (2010) Long-memory and level shifts in the volatility of stock market return indices. J Bus Econ Stat 28(2):275–290
Phillip A, Jennifer C, Shelton P (2019) On long memory effects in the volatility measure of Cryptocurrencies. Fin Res Lett 28:95–100
Qu Z (2011) A test against spurious long memory. J Bus Econ Stat 29:423–437
Rambaccussing D, Mazibas M (2020) True versus spurious long memory in cryptocurrencies. J Risk Fin Manag 13(9):186
Robinson PM (1995) Gaussian semiparametric estimation of long range dependence. Ann Stat 23:1630–1661
Robinson PM (2008) Multiple local Whittle estimation in stationary systems. Ann Stat 36(5):2508–2530
Selmi R, Mensi W, Hammoudeh S, Bouoiyour M (2018) Is Bitcoin a hedge, a safe haven or a diversifier for oil price movements? A comparison with gold. Energy Econ 74:787–801
Shimotsu K (2007) Gaussian semiparametric estimation of multivariate fractionally integrated processes. J Econ 137(2):277–310
Shimotsu K (2012) Exact local Whittle estimation of fractionally cointegrated systems. J Econ 169(2):266–278
Shimotsu K, Phillips PC (2005) Exact local Whittle estimation of fractional integration. Ann Stat 33(4):1890–1933
Shimotsu K (2006) Simple (but effective) tests of long memory versus structural breaks. Department of Economics, Queen’s University. Working Paper 1101
Sibbertsen P, Leschinski CH, Holzhausen M (2018) A multivariate test against spurious long memory. J Econ 203(1):33–49
Soylu PK, Okur M, Çatıkkaş Ö, Altintig ZA (2020) Long memory in the volatility of selected cryptocurrencies: Bitcoin, ethereum and ripple. J Risk Fin Manag 13(6):107
Stărică C, Granger C (2005) Nonstationarity in stock returns. Rev Econ Stat 87(3):503–522
Thies S, Molnár P (2018) Bayesian change point analysis of Bitcoin returns. Fin Res Lett 27:223–227
Tiwari AK, Jana RK, Das D, Roubaud D (2018) Informational efficiency of bitcoin: an extension. Econ Lett 163:106–109
Tran VL, Leirvik T (2020) Efficiency in the markets of crypto-currencies. Fin Res Lett 35:101382
Urquhart A (2016) The inefficiency of Bitcoin. Econ Lett 148:80–82
Urquhart A (2017) Price clustering in Bitcoin. Econ Lett 159:145–148
Vandezande N (2017) Virtual currencies under EU anti-money laundering law. Comput Law Sec Rev 33:342–353
Wang L (2008) Change-in-mean problem for long memory time series models with applications. J Stat Comput Simul 78(7):653–668
Wendt H, Roux SG, Jaffard S, Abry P (2009) Wavelet leaders and bootstrap for multifractal analysis of images. Sig Process 89(6):1100–1114
Wenger K, Leschinski C, Sibbertsen P (2018) A simple test on structural change in long-memory time series. Econ Lett 163:90–94
Wenger K, Leschinski C, Sibbertsen P (2019) Change-in-mean tests in long-memory time series: a review of recent developments. AStA Adv Stat Anal 103(2):237–256
Wenger K, Leschinski C (2019) Fixed-bandwidth cusum tests under long memory. Econ Stat, Forthcoming.
Werner K, Bouri E, Tetsuya T (2020) Cryptocurrencies and equity funds: evidence from an asymmetric multifractal analysis. Phys A 545:123711
Xu J, Perron P (2014) Forecasting return volatility: Level shifts with varying jump probability and mean reversion. Int J Forecast 30(3):449–463
Yau CY, Davis RA (2012) Likelihood inference for discriminating between long memoryand change-point models. J Time Ser Anal 33(4):649–664
Yermack D (2015) Is Bitcoin a real currency? An economic appraisal. Handbook of digital currency. Academic Press, Cambridge, pp 31–43
Zhang YJ, Bouri E, Gupta R, Ma SJ (2020) Risk spillover between Bitcoin and conventional financial markets: an expectile-based approach. N Am J Econ Fin 55:101296
Funding
This research did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Contributions
AA did the coding, data analysis and empirical results. LG contributed to the literature, editing and supervision of the overall paper and helped in drafting the manuscript. KM contributed to the literature, editing, writing and supervision of the overall paper.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Human participants or animals performed
This article does not contain any studies with human participants or animals performed by any of the authors.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: The local whittle estimator
Appendix: The local whittle estimator
The Whittle estimator is initially proposed by Künsch (1987) and then modified by Robinson (1995). It is used to estimate the long-memory parameter d. Ford ∈ (-1/2, 1/2), it is given by:
where \(\hat{G}(d) = \frac{1}{m}\sum\nolimits_{j = 1}^{m} {\lambda_{j}^{2d} I_{T} ,y(\lambda_{j} ),I_{T} ,y} (\lambda_{j} ) = (2\pi T)^{ - 1} \left| {\sum\nolimits_{t = 1}^{T} {y_{t} e^{{it\lambda_{j} }} } } \right|^{2}\) is the periodogram of the series yt and \(\lambda_{j} = (\frac{2\pi j}{T})\) represent the Fourier frequencies.
Hou and Perron (2014) suggest a modified version and call it the modified local Whittle estimator. The local Whittle estimator is consistent in the case of two scenarios: the presence or absence of mean shifts and is given by:
where \(\hat{G}(d,\theta ) = \frac{1}{m}\sum\nolimits_{j = 1}^{m} {\frac{{I_{T} ,y(\lambda_{j} )}}{{\lambda_{j}^{ - 2d} + \theta \lambda_{j}^{ - 2} /T}}}\) where θ reflects the scaling of the shifts. Hou and Perron (2014) prove the consistency and asymptotic normality of the modified local Whittle estimator for the case when m > T5/9.
Frederiksen et al. (2012) fit the following two polynomials: \(\log \phi_{y} \left( \lambda \right) \simeq \log G + h_{y} \left( {\theta_{y} ,\lambda } \right)\) and \(\log \phi_{w} \left( \lambda \right) \simeq \log G + \log \theta_{p} + h_{w} \left( {\theta_{w} ,\lambda } \right)\) to approximate the logarithms of \(\phi_{y} \left( \lambda \right)\) and \(\phi_{w} \left( \lambda \right)\) in\(f_{Z} \left( \lambda \right) = \phi_{y} \left( \lambda \right)\lambda^{ - 2d} + \phi_{w} \left( \lambda \right) + \phi_{u} \left( \lambda \right)\lambda^{ - 2} /T\). Here \(\theta = \left( {\theta^{\prime}_{y} ,\theta_{p} ,\theta^{\prime}_{w} } \right)^{^{\prime}}\), \(\theta_{p} = \phi_{w} \left( 0 \right)/\phi_{y} \left( 0 \right)\) is the long-run signal-to-noise ratio,\( {h}_{a}\left({\theta }_{a,}\lambda \right)=\sum_{r=1}^{{R}_{a}}{\theta }_{a,r}{\lambda }^{2r}\), and\( a \epsilon \left\{y,w\right\}\). Then, \({\widehat{G}}_{\mathrm{LPWN}}=\frac{1}{m}\sum_{j=1}^{m}\frac{{\lambda }_{j}^{2d}{I}_{\mathcal{z}}({\lambda }_{j})}{{g}_{\mathrm{LPWN}}(d,\theta ,{\lambda }_{j})}\). Then, \({g}_{\mathrm{LPWN}}\left(d,\theta ,\lambda \right)=\mathrm{exp}\left({h}_{y}\left({\theta }_{y},\lambda \right)\right)+{\theta }_{\rho }{\lambda }^{2d}\mathrm{exp}({h}_{w}\left({\theta }_{w},\lambda \right))\) and the estimator is given by:
where \(0<{d}_{1}<{d}_{2}<1\) is assumed to be stationary, and \(\Theta \) is a compact and convex set in \({\mathbb{R}}^{{R}_{y}}\times (0,\infty )\times {\mathbb{R}}^{{R}_{w}}\)
Rights and permissions
About this article
Cite this article
Assaf, A., Gil-Alana, L.A. & Mokni, K. True or spurious long memory in the cryptocurrency markets: evidence from a multivariate test and other Whittle estimation methods. Empir Econ 63, 1543–1570 (2022). https://doi.org/10.1007/s00181-021-02165-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-021-02165-6