Paper: arXiv 2307.08612

Authors: Jessica Morales Herrera, Raúl Salgado-García

Abstract

In this paper, we present a measure of time irreversibility using trend pattern statistics. We define the irreversibility index as the Kullback-Leibler divergence between the distribution of uptrends subsequences (increasing trends) and the corresponding downtrends subsequences distribution (decreasing trends) in a time series. We use this index to analyze the degree of irreversibility in log return series over time, specifically focusing on five cryptocurrencies: Bitcoin, Ethereum, Ripple, Litecoin, and Bitcoin Cash. Our analysis reveals a strong indication of irreversibility in all these cryptocurrencies and the characteristic evolves over time. We additionally evaluate the market efficiency for these cryptocurrencies based on a recently proposed information-theoretic measure. By comparing inefficiency and irreversibility, we explore the relationship between these statistical features. This comparison provides insight into the non-trivial relationship between inefficiency and irreversibility.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper introduces a formal irreversibility index based on Kullback-Leibler divergence, showing mathematical density with stochastic process theory and derivations. It applies the method to real cryptocurrency data with statistical tests, providing backtest-ready analysis but lacks code or implementation details.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Assess Irreversibility in Cryptocurrencies"] --> B["Data Inputs:<br>Log Return Series of 5 Cryptocurrencies"]
  B --> C["Key Methodology:<br>Trend Pattern Statistics"]
  C --> D["Computational Process:<br>Calculate Irreversibility Index<br>(Kullback-Leibler Divergence)"]
  D --> E["Comparative Analysis:<br>Irreversibility vs. Market Inefficiency"]
  E --> F["Key Finding 1:<br>Strong Irreversibility in all Cryptocurrencies"]
  E --> G["Key Finding 2:<br>Non-trivial Relationship between<br>Irreversibility and Inefficiency"]
  F --> H["Outcome:<br>Dynamic Evolution of Irreversibility over Time"]
  G --> H

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper introduces a new method to evaluate time‑irreversibility through trend pattern statistics, defining an irreversibility index as the Kullback‑Leibler divergence between up‑trend and down‑trend distributions.

Source sentence
Firstly, we introduce a new method to evaluate time-irreversibility through what we refer to as trend patterns.

Assumes 4

  • Missing log‑return values are replaced with independent normal random numbers using the sample mean and standard deviation of the observed returns. data
    Source sentence
    11 To address this issue, we replaced the missing data in the log returns with independent random numbers following a normal distribution. We used the sample mean and sample standard deviation of the corresponding log returns as the mean and standard deviation of this normal distribution, respectively.
  • A sliding window of 91 days provides a sample size large enough to minimize statistical errors for the irreversibility estimation. statistical
    Source sentence
    We chose this window size to ensure that the sample size is large enough to minimize statistical errors, as our tests in Section 3.2 showed.
  • Surrogate data generated by randomly shuffling log returns represent a null hypothesis of zero irreversibility. statistical
    Source sentence
    To determine the statistical significance, we performed a surrogate data analysis by randomly shuffling the original data (log returns) to eliminate any sign of irreversibility in the sample.
  • The log‑return series of each cryptocurrency can be treated as a realization of a stochastic process suitable for trend‑pattern analysis. model
    Source sentence
    We define a trend pattern as a sequence of log returns that either increase or decrease over time. We then measure the duration of increasing and decreasing patterns in a given time series.

Data

  • Cryptocurrency minute‑frequency price data (Bitcoin, Ethereum, Ripple, Litecoin, Bitcoin C · CryptoDataDownload market-prices public
    varies; Bitcoin 2015‑2022, others simila · five major cryptocurrencies
    Source sentence
    Some proponents argue that it is an improved version of the original cryptocurrency and that its goal is to stay true to the vision of the creator, Satoshi Nakamoto [47] We acquired data from the CryptoDataDownload website [48], which offers historical databases from various exchanges such as Bitstamp, Bitso, Gemini, Binance, and others. Data Preprocessing We were able to access several years of data for each
  • Synthetic autoregressive (AR) and nonlinear autoregressive (NAR) time series synthetic unknown
    Source sentence
    (25) To test the irreversibility index based on trend patterns, we simulate both the AR and NAR processes numerically. In the case of the NAR process, which is bivariate, we obtain a single time series by defining ut = x2 t + y2 t .

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • All five cryptocurrencies exhibit strong signs of time‑irreversibility, as indicated by irreversibility indices exceeding the 95% statistical threshold. statistical figure from elsewhere in the paper
    Source sentence
    Our analysis reveals a strong indication of irreversibility in all these cryptocurrencies and the characteristic evolves over time. From these irreversibility curves we can observe that these cryptocurrency markets have strong signs of irreversibility, something which is clear when comparing the estimated curves (solid black lines) with the threshold of statistical significance (solide red lines), which as we
  • Irreversibility levels are not constant over time, showing periods of high and low values across the sample. statistical
    Source sentence
    We should also notice that this irreversibility exhibited in all these cryptocurrencies is not constant: we 14 02/09/2018 01/16/2019 12/22/19 11/26/20 11/01/2021 10/07/2022 0 2×10 -4 4×10 -4 6×10 -4 8×10 -4 1×10 -3 IT 09/27/2019 07/04/2020 04/10/2021 01/05/2022 The solid black lines represent the irreversibility indices, while the solid red lines indicate the threshold of statistical confidence at 95%. clearly see
  • A global decreasing trend in irreversibility is observed for most cryptocurrencies, though not monotonic. statistical
    Source sentence
    Also, but less clear at first sight, we could even say that the irreversibility seems to decrease globally overtime although not in a monotonic nor a systematic way.
  • Bitcoin Cash consistently shows low irreversibility, with its index often below the significance threshold. statistical
    Source sentence
    However, in contrast to the other cryptocurrencies, we did not observe a systematic decrease in the irreversibility over time for BCH and LTC.
  • The proposed trend‑pattern irreversibility index correctly distinguishes reversible synthetic AR processes from irreversible NAR processes. theoretical
    Source sentence
    (25) To test the irreversibility index based on trend patterns, we simulate both the AR and NAR processes numerically. In the case of the NAR process, which is bivariate, we obtain a single time series by defining ut = x2 t + y2 t .

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 20 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.