Analysis of trade packages in the Chinese stock market

Abstract

This paper conducts an empirical study on trade packages of 23 stocks of the Chinese stock market, each composed of a sequence of consecutive purchases or sales of a stock. We investigate the probability distributions of the execution time, the number of trades, and the total trading volume of trade packages, and analyse the possible scaling relations between them. Quantitative differences are observed between institutional and individual investors. The trading profile of trade packages is investigated to reveal the preference for large trades with respect to trading volume and transaction time of the day, and the different profiles of the two types of investors imply that institutions may be more informed than individuals. We further analyse the price impact of both the entire trade packages and the individual transactions within trade packages. We find that the price impact of trade packages is non-negligible over the period of the execution time and it may have a power-law relation with the total trading volume. The price impact of the transactions within trade packages displays a U-shaped profile with respect to the time t of the day, and also shows a power-law dependence on the trading volume. The trading volumes of the transactions within trade packages made by institutions have a stronger impact on current returns, but the following price reversals persist over a relatively shorter horizon in comparison with those of individuals.

Identification of clusters of investors from their real trading activity in a financial market

“…The main reason is the extreme heterogeneity in the trading activity of individual investors (heterogeneity is not so significant for market member data). For this reason, in this paper we use the more sophisticated SVN to identify clusters of investors.Other studies have had access to databases with the resolution of the individual investors (see, for example, [39] for a profit analysis of Taiwanese investors, [40] for an analysis of the Flash Crash of 6 May 2010 or [41] for an investigation of order splitting for individual investors). The database used in this paper has been investigated extensively by Grinblatt and Keloharju [42,43] in a series of studies on the trading profile of individual and institutional investors and on behavioral aspects of individual investors.…”

Quantitative Multi-factor Stock Selection Strategy Based on Internet Service Industry

“…Commonly used machine learning stock selection algorithms include neural networks, deep neural networks, support vector machines, random forests, and XGBoost algorithms, etc. [8], which can be more sensitive to capture investment opportunities brought about by unreasonable or irrational factors in the market. Furthermore, in terms of constructing investment portfolios, Li made predictions based on the parameter regression of the logistic model, and utilized the clustering stock selection strategy on the basis of selecting the stock pool by the regression method, so as to construct a reasonable investment that can obtain excess returns portfolio [9].…”

Patterns of trading profiles at the Nordic Stock Exchange. A correlation-based approach.

“…Several studies have considered investment profiles of individual investors. Examples are [16] that investigated profits of Taiwanese individual investors, [17] investigating the average transaction value and average portfolio value of individual investors, [18] that analyzed the detailed trading decisions of single investors acting in the E-mini S&P 500 stock index futures market during the Flash Crash of May 6, 2010, [19] that introduced a network based method to characterize specific trading profiles of clusters of investors trading the Nokia stock at the Nordic Stock Exchange, [20] that investigated the order splitting of large orders of individual investors, [21] presenting a robust measure of the contrarian behavior of retail investors, [22] investigating the relation between investors holds and investors profile for Swedish shareholding, and [23] where authors study the impact of news on the trading behavior of different categories of individual investors.…”