2026-09-19T05:01:34Zhttps://ecommons.cornell.edu/server/oai/requestoai:ecommons.cornell.edu:1813/443652026-05-14T13:51:10Zcom_1813_57466com_1813_57353com_1813_11653com_1813_35col_1813_57468col_1813_47
Miao, Qisiyu
Turvey,Calum G.
Bogan,Vicki L.
2016-07-05T15:30:13Z
2021-05-30T06:00:13Z
2016-05-29
https://hdl.handle.net/1813/44365
https://doi.org/10.7298/X4639MP0
9597211
Momentum, the strategy of capitalizing on the ongoing trend in the stock price movements, has been one of the most puzzling market anomalies in modern finance. This paper seeks to exploit the momentum profitability from the perspective of the excursion patterns in the stock price movements. A theoretical framework is developed for momentum strategy analysis and the long memory process in the financial markets. To test the null hypothesis of the Random Walk Hypothesis and the Efficient Market Hypothesis, we employ the Hurst exponent to detect the long-term memory existed in the stock return series. A time series with 0 < H < 0.5 shows negative correlations between points and a mean-reverting behavior, while a series with 0.5 < H < 1 indicates positive correlations and a long memory process. Basic momentum strategies are further applied to the past stock price data, and the back testing results show that there is a U-shaped relationship between the strategy returns and Hurst exponent. This paper also builds on earlier model of a rule-based naïve trading strategy using Hurst exponent as a signal. The strategy generates remarkably higher profitability compared with the benchmark returns. These findings provide new evidence against the random walk assumption and present challenges to a number of rational asset pricing theories. ii
en_US
Momentum Trading
Fractional Brownian Motion
Hurst Exponent
Memory Vs Momentum - Exploring Momentum Strategies With The Hurst Exponent
dissertation or thesis
application/pdf
Agricultural Economics
Cornell University
Master of Science
M.S., Agricultural Economics
Publication
Applied Economics and Management MS Theses
Cornell Theses and Dissertations
Miao, Qisiyu
47601f18-d3de-4171-9187-8e5283cf3a8f
5893a6ea-7af3-41d7-abc6-04bcd26ab5df
open.access
open.access
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="03640e79-f246-4bd2-af08-41b079899aa7">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>en_US</Language>
<Title>Memory Vs Momentum - Exploring Momentum Strategies With The Hurst Exponent</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2016-05-29</PublicationDate>
<DOI>https://doi.org/10.7298/X4639MP0</DOI>
<Authors>
<Author>
<DisplayName>Miao, Qisiyu</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<OrgUnit />
</Publisher>
</Publishers>
<Keyword>Momentum Trading</Keyword>
<Keyword>Fractional Brownian Motion</Keyword>
<Keyword>Hurst Exponent</Keyword>
<Abstract>Momentum, the strategy of capitalizing on the ongoing trend in the stock price movements, has been one of the most puzzling market anomalies in modern finance. This paper seeks to exploit the momentum profitability from the perspective of the excursion patterns in the stock price movements. A theoretical framework is developed for momentum strategy analysis and the long memory process in the financial markets. To test the null hypothesis of the Random Walk Hypothesis and the Efficient Market Hypothesis, we employ the Hurst exponent to detect the long-term memory existed in the stock return series. A time series with 0 < H < 0.5 shows negative correlations between points and a mean-reverting behavior, while a series with 0.5 < H < 1 indicates positive correlations and a long memory process. Basic momentum strategies are further applied to the past stock price data, and the back testing results show that there is a U-shaped relationship between the strategy returns and Hurst exponent. This paper also builds on earlier model of a rule-based naïve trading strategy using Hurst exponent as a signal. The strategy generates remarkably higher profitability compared with the benchmark returns. These findings provide new evidence against the random walk assumption and present challenges to a number of rational asset pricing theories. ii</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>