Crash prediction based on big data analysis of fluctuation in supply and demand balance for liquidity

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K18439
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 7:Economics, business administration, and related fields
• Research Institution: Osaka Metropolitan University (2022-2023); Osaka City University (2021)
• Principal Investigator: Takada Teruko 大阪公立大学, 大学院経営学研究科, 教授 (30347504)
• Project Period (FY): 2021-07-09 – 2024-03-31
• Keywords: 流動性 / 需給バランス / 暴落予測 / 指値 / 確率密度推定

Outline of Final Research Achievements

Aiming at achieving higher crash prediction accuracy, this research extracts maximum information about crash from the world biggest limit order book data, NYSE Openbook, by applying robust nonparametric methods advantageous for crash events. First, the liquidity demand-supply curves are nonparametrically estimated based on the statistics by buyer and seller, and several statistical patterns of time series changes and facts related to the degree of buy/sell pressure are found. Next, I developed a machine which learns market conditions to predict future crash. Lastly, the developed machine is extended to learn the newly found facts from the liquidity demand-supply curve analysis; long-term trend direction and the degree of buy/sell pressure, which achieves the earlier crash prediction with higher accuracy.

Free Research Field

Computational finance

Academic Significance and Societal Importance of the Research Achievements

株価暴落のような異常事象は、深刻な社会的ダメージをもたらすため、その予測の社会的意義は高い。近々の米国株式市場暴落の可能性を憂う声もあり、本研究提案は時宜も得ている。しかし、暴落のような異常事象の統計分析のためには、現代科学が抱える様々な難問の解決を要するため、これまでほとんど行われてこなかった。本研究の行った「データのみからの暴落予測の実現」は、従来法のパラダイムより新しい第４のパラダイムと呼ばれる「データ中心科学」のアプローチにより難題を解決するものであり、学術的意義も高い。