Critique__Evaluating stock price presentation .pptx

stock price evaluation

1. “Evaluating Stock Price Behaviour after Events: An Application of the Self-Exciting Threshold Autoregressive Model" Author-Rakesh Bharati OP5206: Business Analytics & Research Dr. Ankit Mahindroo

2. INDEX 01 PURPOSE OF RESEARCH PAPER 02 PAPER INTRODUCTION 05 OUTCOMES AND INTERPRETATIONS 04 RESULTS 03 RESEARCH METHODOLOGY USED CRITIQUE 06

3. PURPOSE OF RESEARCH PAPER The purpose of this research paper was to evaluate the stock price behaviour after events using a new methodology called the Self-Exciting Threshold Autoregressive (SETAR) model. The study aimed to identify event thresholds and examine the post-event daily stock return behaviour over a period of 20 days. The research focused on determining whether stocks return to their previous levels after negative events and whether there is a pattern of continuation or reversal in stock returns over the long term.

4. PAPER INTRODUCTION The research article focuses on examining stock price behaviour after events using a novel methodology called the Self-Exciting Threshold Autoregressive (SETAR) model. The study aims to identify event thresholds and analyse the post-event daily stock return behaviour over a 20-day period. The research investigates the impact of both positive and negative events on stock returns and explores the presence of overreaction and reversal patterns. Previous studies have primarily used fixed thresholds or dispersion-based norms, while the SETAR model offers a more dynamic approach by endogenously determining event thresholds based on security return characteristics. The article divides the analysis into eight sub periods to account for changes in volatility, liquidity, and institutional trading across different time periods. The research primarily focuses on three major markets: NYSE, Amex, and Nasdaq, providing insights into stock behavior in different market environments.

5. RESEARCH METHODOLOGY USED The research article titled "Evaluating Stock Price Behaviour after Events: An Application of the Self- Exciting Threshold Autoregressive Model" utilises the self-exciting threshold autoregressive (SETAR) model to investigate the behaviour of stock prices after events. Here is an outline of the research methodology used in the study: 2.Data and Sample: The study utilizes the Center for Research in Security Prices (CRSP) database to estimate average abnormal returns for all securities from January 1, 1963, to December 31, 2003. The sample is divided into eight subperiods to account for changes in volatility, liquidity, and institutional trading. 3.Post-event Return Analysis: The study examines the post- event daily stock return behavior over a 20-day period. Abnormal returns are analyzed to identify patterns and trends following positive and negative events . 1. Identification of Event Thresholds: The SETAR model is employed to endogenously determine positive and negative event thresholds based on the characteristics of security returns. This allows for the identification of positive, negative, and non-event regions.

6. RESEARCH METHODOLOGY USED 4.Regression Analysis: Post-event abnormal returns are regressed on event-day abnormal returns to further investigate the behavior of stock prices. The analysis aims to determine the degree of reversal or continuation observed in the short term (days 1 to 3) and longer term (days 4 to 20). 5. Comparison with Other Studies: The research findings are compared with previous studies on stock price behaviour to provide additional evidence on overreaction. The results are evaluated in terms of the magnitude and duration of reversals after positive and negative events.

7. Results 01 Post-event Daily Stock Return Behavior : The study utilized a self-exciting threshold autoregressive (SETAR) model to analyze the behavior of stock returns after events. It found that stocks tend to return to their previous level after negative events, but not after positive events. 03 Market Overreaction : Previous studies have shown that market overreaction can last up to a year following large positive and negative weekly returns. This research focused on daily events and found evidence of overreaction in the short term. 02 Continuation and Reversal Patterns : After events, stock returns exhibit a pattern of continuation in the short term (3 to 5 years) with evidence of reversal. This suggests that there is an initial reaction followed by a correction in stock prices. 04 Impact of Bid-Ask Bounce: When portfolios were formed based on large daily declines or gains, correcting for bid-ask bounce weakened or reversed short-term overreaction. However, in the longer term, there was a general tendency toward continuation.

8. Results 05 Self-Exciting Threshold Autoregressive (SETAR) Model: The SETAR model was used to identify event thresholds and distinguish 07 positive, negative, and non-event regions based on security return characteristics. It accurately captured the interplay between previous day returns and the autoregressive process of stock returns. Mixed Results for Positive Events: The results for positive events were mixed, but there was generally evidence of overreaction in the short term. Positive events showed weaker performance compared to negative events in the short term and weaker reversal overall. 06 Reversal after Negative Events : There was strong evidence of reversal in the short term after a negative event, which was consistent across time and three markets (NYSE, Amex, and Nasdaq). Reversal was also observed in the longer term. 08 Sample and Data: The study utilized the Center for Research in Security Prices (CRSP) database to estimate average abnormal returns for securities from 1963 to 2003. The sample was restricted based on daily returns and the presence of observations in the positive and negative event regions.

9. Results 09 Market Variability: The study found that the Nasdaq market exhibited the widest average bandwidth among the three markets, indicating higher volatility. The SETAR model produced more moderate price changes compared to previous studies. 10 Reversal and Continuation in Subperiods: Analysis of subperiods showed a strong tendency of reversal in the short term after a negative event, but results varied across different subperiods and markets, indicating a combination of reversal and continuation effects.

10. OUTCOMES AND INTERPRETATIONS Reversal after Negative Events: Reversal after Negative Events: The research study reveals that stocks tend to return to their previous level after experiencing negative events. This reversal effect is observed in the short term, spanning from days 1 to 3, as well as in the longer term, ranging from days 4 to 20. The evidence of reversal is strong and consistent across different subperiods and markets (NYSE, Amex, and Nasdaq). Lack of Reversal after Positive Events: In contrast to negative events, the study finds that stocks do not exhibit a significant reversal pattern after positive events. The post-event returns following positive events are mixed, suggesting a weaker performance compared to negative events. However, there is some evidence of overreaction in the short term, indicating that stocks may initially respond excessively to positive events.

11. OUTCOMES AND INTERPRETATIONS c. Continuation of Stock Returns: Following the occurrence of events, the research identifies a decisive pattern of continuation in stock returns over a longer time frame, specifically three to five years. This implies that after the initial reversal period, stocks tend to exhibit a trend of continuing their previous price behavior. d. Comparison of Daily and Weekly Event Horizons: The study highlights a contrast between the post-event performances of daily and weekly event horizons. While large daily returns would be expected to coincide with large weekly returns, the results suggest that the relationship is puzzling. This emphasizes the importance of examining events with a daily horizon and using appropriate models to capture the dynamics of stock returns.

12. OUTCOMES AND INTERPRETATIONS e. Self-Exciting Threshold Autoregressive (SETAR) Model: The research utilizes a SETAR model, which is a non-linear time series model capable of identifying significant events and capturing the interplay between previous day returns and the autoregressive process of stock returns. The SETAR model allows for the endogenous determination of event thresholds, enabling the identification of positive, negative, and non-event regions based on security return characteristics. f. Short-term Overreaction and Bid-Asked Bounce: When portfolios are unconditionally formed based on large daily declines or gains, there is evidence of short-term overreaction. However, the correction for bid-asked bounce weakens or reverses this overreaction in the short term. The bid-ask bounce refers to the price discrepancy between the bid and ask prices, which can influence stock returns.

13. OUTCOMES AND INTERPRETATIONS g. Changes in Volatility, Liquidity, and Institutional Trading: To account for variations in market conditions, the research divides the aggregate period into eight subperiods, considering changes in volatility, liquidity, and institutional trading. This approach acknowledges the dynamic nature of the market and ensures a comprehensive analysis across different timeframes. h. Event Identification and Post-event Analysis: The research employs event identification techniques to scan returns over a two- year period following event thresholds estimation. The analysis of post-event abnormal returns helps in understanding the behavior of stock returns after events, and regression models provide insights into the relationship between event-day abnormal returns and subsequent abnormal returns.

14. CRITIQUE 1. Limited explanation of the methodology: The article introduces the self- exciting threshold autoregressive (SETAR) model without providing a thorough explanation of its workings and assumptions. Readers who are not familiar with this model may find it challenging to understand the analysis and its implications fully. 2. Lack of clarity in event definition: The article mentions that events are identified based on thresholds, but it does not explicitly state how these thresholds are determined. Without a clear explanation of the event definition, it becomes difficult to evaluate the validity of the conclusions drawn from the analysis. 3. Limited sample size: The research utilizes data from 1963 to 2003, which may raise concerns about the relevance of the findings to the current market dynamics. Financial markets have evolved significantly since the cutoff date of the data, and it is unclear whether the observed patterns still hold true.

15. CRITIQUE 4. Potential sample bias: The article mentions that certain criteria, such as a minimum number of daily returns and the presence of observations in both positive and negative event regions, were used to select the sample. This selection process could introduce bias and affect the generalizability of the results. 5.Lack of comparison to previous studies: While the article briefly mentions previous studies on stock market overreaction, there is limited direct comparison or discussion of the differences in findings. A more comprehensive analysis that contrasts the current results with earlier research would enhance the article's contribution to the existing literature. 6.Inconsistent results for positive events: The study reports mixed evidence of overreaction and reversal following positive events. This inconsistency raises questions about the robustness and reliability of the findings. A more in-depth analysis and interpretation of these contradictory results would be beneficial.

16. CRITIQUE 7.Neglect of other potential factors: The article focuses primarily on overreaction and reversal patterns but does not thoroughly consider other factors that could influence stock returns after events. Factors such as market sentiment, industry-specific dynamics, and macroeconomic conditions could contribute to the observed patterns, and their exclusion limits the comprehensiveness of the analysis. Overall, while the article offers insights into stock price behavior after events using the SETAR model, it would benefit from providing more clarity on the methodology, addressing potential sample biases, and offering a more comprehensive discussion of the findings in relation to previous research. Additionally, considering other relevant factors that could influence stock returns would enhance the study's depth and applicability.