On the intraday return curves of Bitcoin: Predictability and trading opportunities

Elie Bouri; Chi Keung Marco Lau; Tareq Saeed; Shixuan Wang; Yuqian Zhao

doi:10.1016/j.irfa.2021.101784

On the intraday return curves of Bitcoin: Predictability and trading opportunities

Elie Bouri, Chi Keung Marco Lau, Tareq Saeed, Shixuan Wang, Yuqian Zhao

Research output: Contribution to journal › Article › peer-review

91 Downloads (Pure)

Abstract

Motivated by the potential inferences from intraday price data in the controversial Bitcoin market, we apply functional data analysis techniques to study cumulative intraday return (CIDR) curves. First, we indicate that Bitcoin CIDR curves are stationary, non-normal, uncorrelated, but exhibit conditional heteroscedastic, although we find that the projection scores of CIDR curves could be serially correlated during some certain periods. Second, we show the possibility of predicting the CIDR curves of Bitcoins based on the projection scores and then assess the forecasting performance. Finally, we utilize the functional forecasting methods to explore the intraday trading opportunities of Bitcoins and the results provide evidence of profitable trading opportunities based on intraday trading strategies, which confronts the efficient market hypothesis.

Original language	English
Article number	101784
Journal	International Review of Financial Analysis
Volume	76
Early online date	27 Apr 2021
DOIs	https://doi.org/10.1016/j.irfa.2021.101784
Publication status	E-pub ahead of print - 27 Apr 2021

Access to Document

10.1016/j.irfa.2021.101784

Full textAccepted author manuscript, 641 KBLicence: CC BY-NC-ND

https://www.sciencedirect.com/science/article/pii/S1057521921001228

Cite this

@article{a649dbc3c59549388af4ad40740bd457,

title = "On the intraday return curves of Bitcoin: Predictability and trading opportunities",

abstract = "Motivated by the potential inferences from intraday price data in the controversial Bitcoin market, we apply functional data analysis techniques to study cumulative intraday return (CIDR) curves. First, we indicate that Bitcoin CIDR curves are stationary, non-normal, uncorrelated, but exhibit conditional heteroscedastic, although we find that the projection scores of CIDR curves could be serially correlated during some certain periods. Second, we show the possibility of predicting the CIDR curves of Bitcoins based on the projection scores and then assess the forecasting performance. Finally, we utilize the functional forecasting methods to explore the intraday trading opportunities of Bitcoins and the results provide evidence of profitable trading opportunities based on intraday trading strategies, which confronts the efficient market hypothesis.",

author = "Elie Bouri and Lau, {Chi Keung Marco} and Tareq Saeed and Shixuan Wang and Yuqian Zhao",

year = "2021",

month = apr,

day = "27",

doi = "10.1016/j.irfa.2021.101784",

language = "English",

volume = "76",

journal = "International Review of Financial Analysis",

issn = "1057-5219",

publisher = "Elsevier",

}

TY - JOUR

T1 - On the intraday return curves of Bitcoin: Predictability and trading opportunities

AU - Bouri, Elie

AU - Lau, Chi Keung Marco

AU - Saeed, Tareq

AU - Wang, Shixuan

AU - Zhao, Yuqian

PY - 2021/4/27

Y1 - 2021/4/27

N2 - Motivated by the potential inferences from intraday price data in the controversial Bitcoin market, we apply functional data analysis techniques to study cumulative intraday return (CIDR) curves. First, we indicate that Bitcoin CIDR curves are stationary, non-normal, uncorrelated, but exhibit conditional heteroscedastic, although we find that the projection scores of CIDR curves could be serially correlated during some certain periods. Second, we show the possibility of predicting the CIDR curves of Bitcoins based on the projection scores and then assess the forecasting performance. Finally, we utilize the functional forecasting methods to explore the intraday trading opportunities of Bitcoins and the results provide evidence of profitable trading opportunities based on intraday trading strategies, which confronts the efficient market hypothesis.

AB - Motivated by the potential inferences from intraday price data in the controversial Bitcoin market, we apply functional data analysis techniques to study cumulative intraday return (CIDR) curves. First, we indicate that Bitcoin CIDR curves are stationary, non-normal, uncorrelated, but exhibit conditional heteroscedastic, although we find that the projection scores of CIDR curves could be serially correlated during some certain periods. Second, we show the possibility of predicting the CIDR curves of Bitcoins based on the projection scores and then assess the forecasting performance. Finally, we utilize the functional forecasting methods to explore the intraday trading opportunities of Bitcoins and the results provide evidence of profitable trading opportunities based on intraday trading strategies, which confronts the efficient market hypothesis.

U2 - 10.1016/j.irfa.2021.101784

DO - 10.1016/j.irfa.2021.101784

M3 - Article

SN - 1057-5219

VL - 76

JO - International Review of Financial Analysis

JF - International Review of Financial Analysis

M1 - 101784

ER -

On the intraday return curves of Bitcoin: Predictability and trading opportunities

Abstract

Access to Document

Fingerprint

Cite this