Market Reaction to Earnings Announcements Under Different Volatility Regimes

Abstract

This study investigates the occurrence and persistence of abnormal stock returns surrounding corporate earnings announcements, particularly emphasizing how varying frequencies of financial reporting influence market behavior. Specifically, this research examines the effects of the timing and frequency of disclosures on market reactions and stock price volatility during critical earnings announcement periods. By analyzing firms within the Dow Jones Industrial Average (DJIA) from 2014 to 2024, this study evaluates the interplay between financial reporting schedules and market responses to stock prices. Furthermore, it considers the impact of peer firms’ reporting practices on the assimilation of firm-specific information into stock prices. Using econometric models, including Vector Auto Regression (VAR), Impulse Response Functions (IRFs), and Self-Exciting Threshold Autoregressive (SETAR) models, causal relationships between reporting frequency, stock price volatility, and abnormal return patterns across different volatility regimes are identified. The findings highlight that quarterly reporting practices intensify market responses and contribute to significant variations in stock price behavior in high-volatility periods. These insights provide a deeper understanding of the role of financial disclosure practices and forward-looking guidance in shaping market efficiency. This study contributes to ongoing discussions about balancing the transparency benefits of frequent reporting with its potential to amplify market volatility and sector-specific risks, offering valuable implications for policymakers, investors, and corporate managers.

1. Introduction

For the investment community, publicly reported firm performance information primarily stems from mandatory disclosures, voluntary disclosures, or data furnished by third-party intermediaries. In the aftermath of the recent financial crisis, significant debate has arisen regarding the speed at which information from quarterly financial reports is reflected in stock prices and the length of time that it takes for the market to absorb this information fully (Dow et al., 2024; Fu et al., 2020; Janicka et al., 2020; García Osma et al., 2023; Gigler et al., 2014; Kraft et al., 2018; Plečnik et al., 2022; Roychowdhury et al., 2019).

Critics contend that the mandate for quarterly reporting—reinforced by regulatory bodies and recent legislative changes—places substantial pressure on publicly listed U.S. companies, compelling them to prioritize short-term financial performance at the expense of more sustainable, long-term growth (Brav et al., 2005). Influential media commentators, such as Davis (2022), argue that focusing on short-term profits undermines long-term shareholder value creation. Recent coverage, including an article in The Wall Street Journal (2024), attributes certain managerial decisions at Boeing to the heightened profitability pressures associated with frequent reporting requirements. Moreover, the current shortage of accountants has increased errors and delays in financial statements, exacerbating concerns surrounding quarterly disclosures (Reinstein & Kaszak, 2024; Maurer, 2023; Burke & Polimeni, 2023).

Against this backdrop, questions arise about the interplay between mandated reporting frequency and the market’s processing of new financial data. Critics advocate reevaluating the regulatory framework for financial reporting, suggesting that less frequent disclosure may alleviate immediate pressures, foster more sustainable managerial strategies, and potentially alter how swiftly and thoroughly the market integrates earnings information into stock prices.

This study contributes to the ongoing debate by examining the effects of quarterly financial reporting on stock return patterns using the Dow Jones Industrial Average (DJIA). Rather than focusing solely on synchronicity, as explored by Haga et al. (2022), we investigate the impact of reporting frequency on return dispersion over time. Utilizing Vector Autoregressive (VAR) and Self-Exciting Threshold Autoregressive (SETAR) models, we evaluate how quarterly reporting influences return behavior and volatility. This approach provides insights into the immediate market reactions and temporal patterns of returns that may be shaped by the regulatory and market structures unique to the U.S.

2. Literature Review

2.1. Literature Exploring Reporting Frequencies Across Jurisdictions

Mandatory disclosure frequency varies across countries and over time (Beyer et al., 2010; Blankespoor et al., 2020). In the U.S., public firms have submitted quarterly income statements since 1962, with the SEC mandating more comprehensive quarterly disclosures in 1970. Canada, Japan, China, and Malaysia also require quarterly reports, while the EU and Australia generally employ semi-annual mandates. The UK primarily uses semi-annual reporting, though specific industries must report quarterly. Japan transitioned from semi-annual to quarterly in 2004, and China shifted to voluntary quarterly reporting. Singapore alternates between mandating and retracting quarterly requirements (Ferreira & Morais, 2022; Kajüter et al., 2022; Kubota & Takehara, 2016; Van Buskirk, 2012).

Proponents assert that frequent reporting enhances price efficiency by quickly reflecting firm-specific data in share prices and improving corporate governance. However, some argue that amid today’s abundant information, the impact of quarterly reports may be waning (Kajüter et al., 2022).

2.2. Research Supporting Frequent Financial Reporting

Frequent disclosures are posited to keep investors informed of the most current firm-specific details, potentially producing more efficient share pricing (Abdel-Khalik & Espejo, 1978; Lang & Lundholm, 1993; Tsao et al., 2016). Research consistently shows positive abnormal returns around earnings announcements, termed the earnings announcement premium (Aboody et al., 2010; Ball & Kothari, 1991; Chapman, 2018; Cheng et al., 2021; Patton & Verardo, 2012; Levi & Zhang, 2015; Savor & Wilson, 2016; Barber et al., 2013). More frequent reporting can enhance the predictability of future earnings (D’Adduzio et al., 2024) and improve analysts’ forecasting accuracy (Schipper, 2007). Additional evidence links frequent reporting to reduced cost of capital (Fu et al., 2020) and lower information asymmetry (Guttman et al., 2006). Removing quarterly disclosures can diminish liquidity (Bornemann et al., 2023). These findings align with Fama’s (1970) efficient market hypothesis, suggesting that quarterly updates speed up price formation (McMullin et al., 2019).

2.3. Research Supporting Less Frequent Reporting

Other studies question whether frequent reporting always enhances market efficacy. Nallareddy et al. (2021) find no significant shift in corporate investment following the UK’s mandatory quarterly reporting. This stands in contrast to the findings of Kraft et al. (2018), who noted that the transition in the U.S. from semi-annual to quarterly reporting during the mid-20th century contributed to managerial myopia, characterized by a tendency for management to prioritize short-term gains at the expense of long-term objectives. Managerial myopia has long been a concern in corporate finance. Stein (1989) suggested that the pressure to meet quarterly earnings targets can lead to inefficient investment decisions that undermine future growth prospects. Ernstberger et al. (2017) demonstrated that a higher reporting frequency in the EU had encouraged some firms to engage in manipulative practices in response to investor pressures. Their study is consistent with Healy and Wahlen’s (1999) findings regarding the incentives for quarterly earnings management under performance pressure; in particular, they define earnings management as manipulating financial reports to achieve desired financial outcomes, potentially harming a firm’s long-term performance. Ernstberger et al. (2017) further demonstrated that the higher reporting frequency in the EU had driven some firms toward manipulative practices in response to investor pressures. As summarized by Bui (2024), many studies have explored the motives behind earnings management, and increased reporting frequency creates increased opportunity for earnings management.

Furthermore, García Osma et al. (2023) highlighted that increased guidance associated with quarterly earnings reports may incentivize earnings management, echoing prior studies pointing to the risks of overemphasizing short-term targets (Roychowdhury, 2006). Roychowdhury emphasizes that firms often manipulate actual activities, for example, altering operational decisions, to meet or beat quarterly earnings expectations, which can distort the proper economic health of the company. Similarly, Byun et al. (2024) suggest that companies could manipulate reporting dates to influence investor reactions. Mensah and Werner (2008) studied the relationship between disclosure frequency—quarterly reporting versus semi-annual reporting—and stock price volatility, finding that companies reporting quarterly experienced greater stock price volatility.

Balakrishnan and Ertan (2018) researched the impact of quarterly reporting on the banking sector. Their results indicate that quarterly reporting could reduce risk-taking, suggesting a nuanced landscape where reporting frequency influences managerial behaviors differently across sectors. Quarterly reporting may replicate or pre-empt information already available through alternative communication channels, leading to a static or even reduced volume of public information (Butler et al., 2007; McNichols & Manegold, 1983). Moreover, an increase in interim reporting frequency might result in declining voluntary management disclosures, fostering short-term thinking among management. This finding supports the idea that mandated interim reports may not enhance disclosures’ overall quality or timeliness (Butler et al., 2007; Cho et al., 2023; Fu et al., 2020; Gigler et al., 2014). Increased reporting frequency may also diminish the marginal benefits of new information (Pitre, 2012). Bhandari et al. (2022) indicate that companies demonstrating lower financial reporting aggressiveness are less likely to meet or exceed analyst expectations, which suggests adherence to higher-quality financial reporting standards. Das et al. (2009) discovered that U.S. firms experiencing negative earnings changes in the first three quarters (Q1–Q3) were more likely to report positive changes in the fourth quarter, rather than following a random pattern. This observation implies the potential for earnings management if companies utilize fourth-quarter earnings to “adjust” Q1 to Q3 earnings to meet annual targets, which undermines the reliability of quarterly financial data. Specifically, interim earnings necessitate estimations of annual revenues or expenses and are generally not subject to audit. If analysts neglect this reduced reliability, then they may overreact to interim reports, adversely impacting the accuracy of their forecasts. DeHaan et al. (2015) demonstrate that managers strategically schedule earnings announcements to obscure negative information or highlight positive outcomes. Sengupta (2004) investigated the reporting lag (the interval between the fiscal period’s end and the quarterly earnings release date) and identified firm characteristics contributing to this lag. Byun et al. (2024) analyzed the variability of firms’ annual earnings announcement dates over time and discovered that firms with fewer resources, weaker internal monitoring systems, and increased financial uncertainty were significantly more likely to exhibit greater variability in their announcement dates, resulting in a noticeably weaker capital market response to earnings announcements.

Evidence on whether frequent disclosures invariably lead to myopic decision-making remains mixed (Biehl et al., 2024). The interaction between mandatory and elective disclosures in shaping a firm’s external information environment remains a key focus for scholars in information economics and regulatory organizations (Einhorn, 2005; Gatti et al., 2019; Lennox & Pittman, 2011). Voluntary disclosures are generally perceived as a positive engagement, while the absence of disclosure can carry negative implications. Conversely, in mandatory disclosure settings, silence is often interpreted primarily as a negative signal (Lennox & Pittman, 2011; Polinsky & Shavell, 2012). Voluntary disclosures increase the overall availability of information and generate positive signals, enabling analysts to formulate more accurate forecasts. However, companies can strategically leverage the option of disclosure to shape market perceptions (Castellani et al., 2024).

The pressure to report with increased frequency may lead to short-term thinking among management (Gigler et al., 2014; Kraft et al., 2018), presenting additional challenges for analysts’ predictive efforts. Another area of research investigates the effects of earnings calls and earnings guidance. Atiase et al. (2005) explored the value of current earnings versus future earnings guidance and identified differences in impact based on investors’ relative preferences for reliability.

Smaller firms have lobbied for reduced reporting frequency using the arguments that less frequent reporting would promote sustainable, long-term strategies and reduce the financial burden of producing the financial reports on a quarterly basis (Fink, 2018). Regulators with the Jumpstart Our Business Startups Act of 2012 gave into pressures exerted by smaller firms and reduced the reporting for those firms from quarterly to semi-annual frequency.

Stock price volatility caused by frequent financial reporting has an excessive impact not only on smaller firms but possibly on firms in some sectors. This impact of price volatility by sector has not been explored in prior studies. Our study examined stock price volatility in quarterly reporting environments, with a breakdown for different sectors. The results imply that the reporting frequency should be reduced not only to smaller firms but also to firms from the sectors impacted by the greatest volatility. The results of this study should be considered by policymakers when reducing reporting frequency in sectors impacted excessively during financial reporting announcements.

3. Hypothesis

We hypothesize that by examining the reporting of the Dow Jones Industrial Average (DJIA) components across 40 quarterly cycles, we can better understand the relationship between different volatility regimes and the time that it takes for earnings information to be fully absorbed by the market. We focus on the past ten years because they encompass three distinct volatility regimes, multiple economic cycles, and periods of both low and, later, more pronounced—and ultimately declining—interest rates. These evolving market conditions provide a diverse backdrop against which to assess how corporate earnings disclosures interact with and shape investor behavior, offering valuable insights into price discovery processes across varying levels of market uncertainty.

4. The Data and Research Methods

Our analysis examines the interactions between the log changes in the DJIA (Dow Jones Industrial Average) and those in the VIX (CBOE Volatility Index). The daily data cover a sample period from 3 November 2014 to 11 November 2024, with 2483 observations for each series. The data were obtained from the Federal Reserve Economic Data (FRED), Yahoo Finance, and Compustat. Before conducting empirical modeling, we obtained descriptive statistics and unit root tests for the log changes in DJIA and VIX, as detailed in

Table 1. Descriptive statistics and unit root tests for log changes in DJIA and VIX.

Statistics	Log DJIA	Log VIX
Mean	0.000348	0.000187
Median	0.000661	−0.007243
Maximum	0.107643	0.768245
Minimum	−0.138418	−0.330681
Std. Deviation	0.011070	0.079473
Skewness	−0.216	3.774
Kurtosis	8.208	25.429
Jarque–Bera	55,031.73 ***	6469.26 ***
Probability	0.0000	0.0000
ADF Statistic	−15.779 ***	−20.977 ***
ADF p-value	0.0000	0.0000
KPSS Statistic	0.024	0.012
KPSS p-value	0.1	0.1
Observations	2483	2483

Notes: Daily data from 3 November 2014 to 11 November 2024, with 2483 observations. ADF represents the Augmented Dickey–Fuller test statistic, while KPSS denotes the Kwiatkowski–Phillips–Schmidt–Shin test. *** denotes significance at 1%. Source: FRED and authors’ estimation.

.

Table 1 shows that the distributions of log changes in the DJIA and the VIX are non-normal, based on the Jarque–Bera test, which reports a p-value of 0.000 for each. This means that the data do not follow a normal distribution. The DJIA and the VIX also have kurtosis values above 3 (especially the VIX), indicating they are leptokurtic or “heavy-tailed”, which increases the likelihood of extreme values. Furthermore, the VIX is positively skewed, pointing to more frequent large upward movements, while the DJIA is slightly negatively skewed, indicating a mild tendency toward negative returns.

The Dow Jones Industrial Average (DJIA) demonstrates a historically high correlation with the S&P 500, often exceeding 0.95 over multi-year periods, indicating that the DJIA reliably captures a significant portion of the broader market’s performance (S&P Dow Jones Indices, 2021). This strong relationship underscores the DJIA’s utility as a proxy for studying market trends and dynamics, particularly when a simplified yet representative index is required. Consequently, the DJIA’s selection for this study is justified, given its ability to effectively reflect the movements and patterns of the broader U.S. equity market.

The Augmented Dickey–Fuller (ADF) test strongly suggests that the DJIA and VIX log changes are stationary (with very negative test statistics and p-values close to zero), implying that their statistical characteristics, such as mean and variance, remain relatively consistent over time, making them appropriate for time-series modeling.

Lastly, the KPSS test, which has high p-values (0.1) for the DJIA and the VIX, supports the conclusion that both series are stationary. Therefore, the log changes do not require additional transformations to meet stationarity requirements for further econometric analysis.

4.1. Identification of Low, Intermediate, and High VIX Zones

Our analysis used a SETAR(2,3) model, specifying two thresholds and three lagged terms. This choice was guided by the Akaike Information Criterion (AIC), which evaluates a model’s fit while penalizing unnecessary complexity. A lower AIC value means that the model explains the data well without becoming overly complicated. First, we systematically tested different threshold values that divide the data into distinct volatility zones (low, intermediate, and high). We estimated the model and calculated the AIC for each potential threshold combination. The thresholds that produced the lowest SAC were then selected, ensuring that the final model balanced accuracy and simplicity.

Next, we determined the optimal number of lagged terms by trying various lag lengths and again relying on the AIC to identify the best fit. Ultimately, we settled on three lagged terms (VIX_(t − 1), VIX_(t − 2), and VIX_(t − 3)), capturing how past volatility levels influence current fluctuations. Including these lags allowed the model to account for the persistence and dynamics in the VIX data, improving the regime-switching accuracy.

After finalizing the thresholds and lag length, we validated the model using out-of-sample data. A high R-squared value and statistically significant coefficients indicated that the model was well specified, while the Durbin–Watson statistic of 2.054 suggested no significant autocorrelation in the residuals. These results confirm that the SETAR(2,3) model fits the historical data and remains robust in predictive performance, making it a reliable tool for capturing different volatility regimes as implied by the VIX.

As indicated before, we employed the SETAR(2,p) test to identify three discernible zones of the VIX: low, intermediate, and high. We used the SETAR model as this methodological approach helps to quantify the distinctive zones of changeable interactions between the tested variables quite precisely. Our SETAR testing was rooted in the generalized two-regime, one-threshold SETAR(1,p) model (Tong & Lim, 1980; Tong, 2015), specified as follows:

$$X_t=\left\{\begin{array}{c}{\alpha }_{10}+{\alpha }_{11}{X}_{t-1}+{\epsilon }_{t}if{X}_{t-p}<r\\ {\alpha }_{20}+{\alpha }_{21}{X}_{t-1}+{\epsilon }_{t}if{X}_{t-p}\ge r\end{array}\right\},$$

(1)

where αn is a real constant.

The self-exciting component, the SE of the SETAR, is a lagged value of the dependent variable Xt – p driven by the threshold “r”.

Our extension of Tong’s original methodology aims to identify three volatility zones for the VIX. Our SETAR model, using the same method as Herley et al. (2024), can be expressed as follows:

$${Vix}_t=\left\{\begin{array}{c}{\alpha }_{10}if{Vix}_{t-p}<r_1\\ {\alpha }_{20}if{r_1\le Vix}_{t-p}<r_2\\ {\alpha }_{30}if{Vix}_{t-p}\ge r_3\end{array}\right\}+{\alpha }_{k1}{Vix}_{t-1}+{\alpha }_{k2}{Vix}_{t-2}+{\alpha }_{k3}{Vix}_{t-3}+{\epsilon }_k.$$

(2)

As Equation (2) implies, our SETAR(2,3) model specification was optimized using two thresholds and three lagged terms. The optimization process of the SETAR model involved determining the optimal number of thresholds, lags, and the best fit to capture non-linearity in the data. This optimization was achieved by minimizing information criteria, specifically the Akaike Information Criteria (AIC), which balance the goodness-of-fit with the model complexity.

We systematically tested potential values that segment the data into distinct regimes to identify the optimal thresholds. The AICwas calculated for each potential threshold combination, and the thresholds that produced the lowest values for these criteria were selected. This ensured that the model was both parsimonious and effective in identifying the different volatility zones. We ultimately identified two thresholds, defining three regimes: low, intermediate, and high volatility.

The lag length used in the model was also determined through an iterative process in which different lag lengths were tested, and the model fit was evaluated based on the information criteria mentioned above. Including three lagged terms (VIX t − 1, t−2, and t−3) was optimal, as this minimized the AIC while capturing the persistence and dynamics of the volatility process. The AIC of −2.587 indicated that the model fitted the data well, balancing its ability to explain the variability while avoiding overfitting. This lag length allowed the model to account for the temporal dependencies in the VIX data, making the regime-switching more informative.

After determining the thresholds and lag length, the model was validated using out-of-sample data to assess its predictive accuracy and robustness. The high R-squared value and significant coefficients indicated that the model was well specified. Furthermore, the Durbin–Watson statistic of 2.054 suggested no significant autocorrelation in the residuals, further validating the model’s reliability.

The empirical testing results for Equation (2) are provided in

Table 2. Zones of VIX, estimated with SETAR(2,p).

Volatility Regimes	Condition	Constant Term	No. of Observations
Low VIX Zone	VIX (t − 2) < 16.72%	0.9441 *** (104.69)	1969
Intermediate VIX Zone	16.72% ≤ VIX (t − 2) < 26.53%	1.0000 *** (6361.16)	511
High VIX Zone	VIX (t − 2) ≥ 26.53%0	0.0000 *** (nan)	59

Notes: The table shows the estimated thresholds and zones for VIX based on the SETAR(2,p) model. *** denotes significance at 1%. Source: authors’ estimation.

.

The SETAR estimation identified the following VIX thresholds and zones:

• Low VIX zone, for VIX (t − 2) < 16.72% (1969 observations);
• Intermediate VIX zone, for 16.72% ≤ VIX (t − 2) < 26.53% (511 observations);
• High VIX zone, for VIX (t − 2) ≥ 26.53% (59 observations).

The obtained thresholds and zones were robust and statistically significant at all levels. The estimated constant terms for each regime and the lagged terms were all significant at the 1% level, implying that these thresholds adequately capture changes in market volatility. Including lagged VIX values as non-threshold variables indicates their influence across regimes, providing insights into how past volatility influences current market conditions. The R-squared value of the model was high, suggesting that the model explains most of the variability in the VIX series.

Identifying these zones is critical in understanding volatility behavior in financial markets. Low VIX values typically indicate periods of market stability, whereas high VIX values are associated with heightened market uncertainty or stress. The intermediate zone represents a transition state between calm and turbulent conditions.

4.2. VAR Model

We considered a trivariate Vector Autoregressive (VAR) model including the log changes in DJIA (Dow Jones Industrial Average), the log changes in VIX (CBOE Volatility Index), and the log changes in a stock ticker of interest. The VAR model was specified as follows, in Equation (3):

Let

• $Y_t$ = log changes in the DJIA at time $t$;
• $X_t$ = log changes in the VIX at time $t$;
• $Z_t$ = log changes in the selected stock ticker at time $t$.

Then, the VAR(p) model is given by

$$\left[\begin{array}{c}{Y}_t\\ X_t\\ Z_t\end{array}\right]=\left[\begin{array}{c}{\alpha }_1\\ {\alpha }_2\\ {\alpha }_3\end{array}\right]+\left[\begin{array}{ccc}{\beta }_{11}\left(L\right)& {\beta }_{12}\left(L\right)& {\beta }_{13}\left(L\right)\\ {\beta }_{21}\left(L\right)& {\beta }_{22}\left(L\right)& {\beta }_{23}\left(L\right)\\ {\beta }_{31}\left(L\right)& {\beta }_{32}\left(L\right)& {\beta }_{33}\left(L\right)\end{array}\right]\left[\begin{array}{c}{Y}_{t-1}\\ X_{t-1}\\ Z_{t-1}\end{array}\right]+\left[\begin{array}{c}{ϵ}_{Y,t}\\ {ϵ}_{X,t}\\ {ϵ}_{Z,t}\end{array}\right],$$

(3)

where

• ${\alpha }_i$ (for $i=1,2,3$) are the constant terms;
• ${\beta }_{ij}\left(L\right)$ represents the lagged coefficients of the variables for $i,j=1,2,3$;
• ${ϵ}_{Y,t},{ϵ}_{X,t},{ϵ}_{Z,t}$ are the error terms.

This VAR(p) specification allowed us to model the interdependencies between the DJIA, the VIX, and the selected stock ticker. Through including lagged values for each variable, we captured the dynamic interactions and temporal relationships among them.

The model was estimated using ordinary least squares (OLS) for each equation in the VAR system. The optimal lag length (p) was determined based on information criteria such as the Akaike Information Criterion (AIC).

The VAR model helps us to understand how shocks to the VIX or the DJIA propagate over time to other variables; for instance, a sudden increase in the VIX might affect future values of the DJIA or the stock ticker, allowing us to analyze the volatility spillovers in financial markets and the corresponding impulse response functions (IRFs).

4.3. IRF Model

An impulse response function (IRF) describes the reaction of any dynamic system in response to an external change. Thus, the IRF illustrates how the system responds to a shock over time, capturing a variable’s evolution over a specified time horizon following a shock at a given moment. The IRF represents the response of each variable in the system to a shock, providing insights into the system’s dynamic behavior. As mentioned, Byun et al. (2024) focused on the timing of reporting dates (investor communication) and how investor perception may change based on how managers manipulate reporting dates. Our study takes the reporting dates as given.

The VAR model was transformed into its moving average (MA) representation to derive the impulse response functions. The MA representation expresses each variable as a function of current and past shocks:

• where represents the impulse response coefficients, which describe the response of the variable to a shock in a variable at lag. Intuitively, the first case is the univariate AR(1) process:

$$x_t=\phi x_{t-1}+u_t,$$

(4)

where x_t is a scalar, φ < 1 (which makes the process stationary), and u_t is a (scalar) random disturbance with a mean of 0.

Specifically for this model, we can rewrite

Period $\mathrm{t}$: Initial Impact

For each regime:

Earnings: At time $t$, $E_t$ receives the shock:

$$E_t=c_E^{\left(k\right)}+{\varphi }_{11}^{\left(k\right)}{E}_{t-1}+{\varphi }_{12}^{\left(k\right)}VI{X}_{t-1}+{\varphi }_{13}^{\left(k\right)}DJI{A}_{t-1}+{\sigma }_E,$$

(5)

where $k\in \{1,2,3\}$ indicates the regime.

Period $t+1$: Response of Earnings and DJIA

In period $t+1$, the effect of the shock propagates through the autoregressive structure.

For each subsequent period, previous values are iteratively substituted into the equations to observe how the shock to $E_t$ propagates to both $E_{t+k}$ and $DJI{A}_{t+k}$. The VIX level potentially shifts the model into a new regime through autoregressive terms and across different regimes.

This approach allows us to compute the IRF over time, revealing how a 1 SD shock impacts earnings and the DJIA across low, medium, and high VIX regimes. The results reveal differential impacts depending on the parameters ${\varphi }_{ij}^{\left(k\right)}$ within each regime.

These impulse response functions provide valuable insights into the dynamic interactions between the DJIA, the VIX, and the selected stock ticker, illustrating how shocks to one variable propagate throughout the system and affect the other variables over time. Similarly to Sims (1980), our interest lies in examining the relationships between variables rather than focusing solely on coefficient estimates. Future studies could examine alternative structural breakpoints using the methodology of Bai and Perron (1998).

5. Results

This section summarizes the results of an analysis of Dow 30 stocks over 40 quarters, categorized by distinct volatility regimes. The efficiency of this analysis was constrained by the inherent complexity of market dynamics, including sector-specific factors, variability in stock behavior, and broader economic conditions. Despite these challenges, the findings provide valuable insights into individual stock reporting, time diffusion, and DJIA interactions.

Our analysis as shown in

Table 3. Sector analysis.

Sector/Example Stocks	Typical Initial Reaction	Volatility Profile	Time to Reversion
Technology (MSFT, AAPL)	Highly positive	High	3–5 days
Industrials (HON, CAT)	Moderate to strong	Moderate	~2–4 days
Consumer Staples (KO, PG)	Mild or muted reaction	Low (defensive)	~2–3 days
Financials (JPM, GS)	Strong initial surge; reversion may take several days	Medium–high	~3–5 days
Healthcare and Utilities (Various)	Occasional negative initial responses (sector-specific factors)	Low–moderate	~2–3 days
High-Volatility Stocks (BA, GS)	Substantial fluctuation in IRFs	High	Variable

, confirms the initial hypothesis that the time required for a firm’s earnings or performance news to be fully incorporated into its stock price depends substantially on the prevailing volatility regime, as identified by the SETAR model. Under high-volatility conditions, investors process information more swiftly, leading to shorter impulse response functions (IRFs) being derived from the VAR model. Conversely, information diffuses slower in low-volatility markets, resulting in extended IRFs. While most stocks among the Dow 30 revert to baseline levels within 3–5 days, technology and financial firms often display more drawn-out fluctuations due to their heightened sensitivity to market sentiment and macroeconomic indicators.

An unexpected finding within this framework is the pronounced impact of a company’s beta—its exposure to systematic risk—on the duration and amplitude of these IRFs. Higher-beta stocks, such as Boeing (BA) and Goldman Sachs (GS), exhibited more significant and persistent price swings, even in conditions where rapid market absorption would otherwise be expected. This phenomenon is especially evident during lower-volatility regimes, suggesting that firm-specific risk factors can prolong the market’s adjustment process despite a relatively subdued overall sentiment.

Sectoral characteristics further highlight these dynamics. Technology stocks like Microsoft (MSFT) and Apple (AAPL) tend to experience protracted and more vigorous reactions to new information, mainly when high-impact news on innovation or growth prospects captures investor attention. Industrial names, including Honeywell (HON) and Caterpillar (CAT), stabilize more rapidly as their performance closely tracks macroeconomic drivers like GDP forecasts and interest rates. In contrast, defensive sectors—such as consumer staples (e.g., Procter & Gamble [PG] and Coca-Cola [KO])—exhibit milder initial responses, reflecting their stable demand profiles and weaker correlation with broader market movements. Financial stocks, notably JPMorgan Chase (JPM) and Goldman Sachs (GS), show robust yet sometimes slower-to-dissipate shocks tied to shifts in market sentiment and economic data releases. Once again, the role of the individual stock beta needs to be considered.

These findings underscore the interplay between systemic conditions and firm-level risk characteristics. While broad-market volatility regimes largely govern the speed of information absorption, a company’s beta can unexpectedly prolong or intensify price adjustments—even in calmer markets—thereby serving as a critical determinant of how quickly earnings and performance shocks are fully priced.

6. Conclusions and Future Directions

This study employs comprehensive econometric models to examine how Dow Jones Industrial Average (DJIA) stocks behave under varying volatility regimes, emphasizing the interaction between financial disclosures and market reactions. The findings reveal that quarterly reporting and systemic factors significantly influence the magnitude and duration of shocks at the individual stock and index levels. While many stocks show a positive response to their reporting dates, the intensity and persistence of these reactions differ by sector. High-volatility sectors like technology often display more pronounced and sustained effects, whereas defensive sectors, including utilities, remain relatively stable. Furthermore, most shocks dissipate within three to five days, indicating the rapid market absorption of new information—but this efficiency depends on the prevailing volatility regime: high-volatility periods generally compress reaction times due to heightened investor vigilance.

A critical takeaway is that frequent reporting can enhance market transparency and amplify short-term volatility and sector-specific risks. Policymakers could exempt specific high-volatility sectors from mandatory frequent reporting to balance transparency with market stability. Despite these insights, this study faces inherent complexity due to sector-specific dynamics, variations in stock behavior, and overarching economic conditions. In particular, the absence of beta-adjusted impulse response functions (IRFs) limits the ability to differentiate how individual stocks’ market risk profiles influence their responses under different volatility regimes.

Future research should address these limitations by examining cumulative response functions over successive quarters—comparing, for example, the IRF for Q1 with Q2—to capture the compounded effects of earnings announcements over time. Such an approach would clarify whether reactions become more or less pronounced with repeated disclosures. Incorporating beta-adjusted IRFs, especially under lower-volatility conditions, would offer a more nuanced view of stock price adjustments. Additionally, exploring how growth-orientated versus value-orientated firms respond to varying frequencies and depths of financial reporting could reveal whether reduced disclosure benefits firms with inherently higher beta values. Finally, evaluating the extent of detail provided in quarterly filings may yield strategies to mitigate short-term market pressures and encourage longer-term decision-making. By integrating these enhancements, future studies can refine our understanding of the complex interplay among volatility regimes, financial disclosure practices, and stock price dynamics.