Bursa Malaysia IPO data 2004-2021

The dataset covers a timeframe spanning from January 2004 to December 2021, with an initial sample of 493 Initial Public Offerings (IPOs). Several exclusion criteria were applied to ensure the study's focus remains aligned with its objectives. First, since the study exclusively examines IPOs that utilize the fixed-price method, IPOs employing the book-building method were excluded, resulting in the removal of 39 cases from the original sample. Furthermore, the study investigates explicitly investor behavior within the IPO market, focusing on public issues, offer-for-sale transactions, or hybrid forms of these two categories. Other offerings, such as restricted and special issues, were excluded to avoid generating less meaningful results (Albada et al., 2019a). This step led to the removal of an additional subset of IPOs. Additionally, Real Estate Investment Trusts (REITs) were excluded from the final sample due to their distinct financial statement presentation formats, which differ significantly from those of traditional IPOs. This exclusion resulted in the removal of another 104 IPOs. After applying these rigorous selection criteria, the final study sample comprises 350 IPOs. The data were meticulously collected from multiple sources to explore the signaling role of various ex-ante information factors on IPO initial returns, investor demand, and the information gap. This dataset provides valuable insights into how potential investors can enhance their decision-making processes when evaluating IPOs under conditions of pronounced information asymmetry. Additionally, the dataset offers opportunities for future research, including cross-country analyses that explicitly examine the influence of country-specific factors on underpricing outcomes. The data presented in this article are linked to several research studies, including “An Insight into the Signaling Role of Sharia Status: A Case from an Emerging IPO Market” (Albada, 2024), “Machine Learning Insights: Probing the Variable Importance of Ex-Ante Information” (Albada et al., 2025), and “Determinants of Investor Opinion Gap Around IPOs: A Machine Learning Approach” (Albada et al., 2024).

本数据集的时间跨度为2004年1月至2021年12月，初始样本包含493宗首次公开发行（Initial Public Offerings, IPOs）。为确保研究始终契合预设目标，本研究设置了多维度筛选排除标准：首先，由于本研究仅考察采用固定价格发行法（fixed-price method）的IPO，故剔除采用簿记建档法（book-building method）的IPO，由此从初始样本中移除39宗案例。其次，本研究明确聚焦于IPO市场中的投资者行为，仅纳入公开发行、售股发行交易或上述两类的混合形式的发行项目，排除受限发行与特殊发行等其他类型的发行项目以避免产生无实质意义的研究结果（Albada等，2019a），该步骤进一步剔除了部分IPO样本。此外，由于房地产投资信托基金（Real Estate Investment Trusts, REITs）的财务报表披露格式与传统IPO存在显著差异，故将其从最终样本中剔除，该操作额外移除了104宗IPO。经上述严格筛选流程处理后，本研究的最终有效样本共包含350宗IPO。为探究各类事前信息因素在IPO初始收益、投资者需求与信息差距中的信号传导作用，研究人员从多渠道精心采集了本数据集。本数据集可为投资者在信息不对称程度较高的场景下评估IPO并优化决策流程提供重要参考。同时，本数据集也为后续研究提供了支撑空间，例如开展跨国分析以明确国别特定因素对IPO抑价结果的影响。本文所呈现的数据关联了多项已发表研究，包括《伊斯兰教法地位的信号传导作用洞察：来自新兴IPO市场的案例》（Albada，2024）、《机器学习洞察：探究事前信息的变量重要性》（Albada等，2025）以及《IPO市场中投资者意见分歧的决定因素：一种机器学习方法》（Albada等，2024）。