Determinants of Stock Price Volatility in Indonesian Digital Banks: Long- and Short-Run Evidence from an Error Correction Model

1. Introduction

The banking industry in Indonesia has undergone a significant transformation with the rise of digital banking, which has reshaped financial transactions and investment behavior. Digital banking adoption has accelerated, especially since 2017, when banks prioritized efficiency through technological integration (Asmarani & Wijaya, 2020).

According to EY’s Global Banking Outlook 2018, approximately 82% of retail banks worldwide, including those in Indonesia, have implemented digitalization strategies, with 89% of them focusing on strengthening cybersecurity measures. This transformation has been further accelerated by the COVID-19 pandemic, which has significantly changed consumer behavior, pushing more financial activities into the digital realm. The increasing reliance on fintech and digital banking services has contributed to the rapid growth of the industry, attracting both investors and financial institutions to this emerging sector (Sapulette et al., 2021).

Despite this growth, digital banking stocks have shown significant volatility, raising concerns among investors and financial analysts. Stock price volatility is a crucial indicator of market stability and investment risk, as it reflects fluctuations in stock prices over time. In October 2022, the Jakarta Composite Index (IHSG) experienced a sharp decline, largely driven by the falling stock prices of digital banking companies. Several major digital banks listed on the Indonesia Stock Exchange (IDX) recorded substantial losses. For example, PT Bank Jago Tbk (ARTO) saw a drastic decline of 73.31%, PT Allo Bank Indonesia Tbk (BBHI) fell by 57.64%, and PT Bank Neo Commerce Tbk (BBYB) dropped by 71.48% (Investor.id, 2022). These significant price movements indicate that digital banking stocks are highly sensitive to market changes, raising the question of what factors drive their volatility.

The research gap arises from the lack of consensus on the key factors influencing digital banking stock volatility. While some studies suggest that internal financial factors, such as leverage, firm size, trading volume, and earnings volatility, significantly impact stock price movements, others emphasize the role of external macroeconomic factors, including inflation and interest rates. Previous research on stock price volatility has yielded mixed results, with some studies finding a strong correlation between financial metrics and stock price movements, while others argue that external economic conditions play a more dominant role (Hidayati & Sukmaningrum, 2021; Yulinda et al., 2020). Given these inconsistencies, further research is needed to determine which factors have the most significant impact on stock price volatility in Indonesia’s digital banking sector.

This study aims to analyze the influence of leverage, firm size, trading volume, earnings volatility, inflation, and interest rates on the volatility of digital banking stocks in Indonesia. By employing the Error Correction Model (ECM), this research seeks to distinguish between short-term and long-term effects of these variables, providing a comprehensive understanding of their impact on stock price movements. Understanding these factors is essential for investors, policymakers, and financial managers in developing more effective investment strategies and regulatory frameworks for the digital banking sector.

By addressing the factors influencing digital banking stock price volatility, this study contributes to the existing body of knowledge in financial economics. The findings will offer valuable insights for investors looking to navigate the uncertainties of the digital banking market, while also helping policymakers formulate strategies to stabilize the financial sector in an increasingly digitalized economy.

2. Literature Review

2.1 Digital Bank

Digital banks, based on POJK No.12/POJK.03/2021, are Indonesian legal entities providing banking services primarily through electronic channels without physical branch offices, except for the head office or with limited physical offices. The rise of digital banks is driven by consumer demand for fast and flexible banking services, accessible anytime and anywhere, and the emergence of fintech in the financial sector​.

2.2 Capital Market

The capital market plays a vital role in the financial system by facilitating the trading of long-term financial instruments such as stocks and bonds. It provides essential financing for companies and investment opportunities for investors. The rapid growth of digital banking stocks on the Indonesia Stock Exchange (IDX) has made the market increasingly dynamic and sensitive to economic indicators, company performance, and global trends (Rachmawati, 2019).

The volatility of digital banking stocks, as seen in the sharp declines of PT Bank Jago Tbk (ARTO) and PT Allo Bank Indonesia Tbk (BBHI) in 2022, highlights the market’s susceptibility to external shocks (Rosihan et al., 2022). High trading volumes, often associated with active market participation, contribute to stock price volatility, especially in innovative sectors like digital banking (Dewi & Suaryana, 2016; Septyadi & Bwarleling, 2020).

In digital banking, the capital market serves as both a funding source and a gauge of market confidence. Its dynamic nature underscores the importance of understanding the factors influencing stock price volatility in this sector, forming a key foundation for this study.

2.3 Signalling Theory

Introduced by Bhattacharya (1979), this theory describes how companies signal their financial prospects to investors. Positive signals, such as strong earnings or growth potential, can attract investors, while negative signals can deter them. This theory highlights the importance of transparent financial information in the capital market​.

2.4 Stock Price Volatility

Stock price volatility measures the degree of price fluctuations over time, indicating the level of risk associated with a stock. In the digital banking sector, high volatility is influenced by market sentiment, financial performance, and macroeconomic conditions.

Yulinda et al. (2020) found that firm size negatively affects volatility, with larger firms experiencing more stable stock prices. Rosyida et al. (2020) highlighted that leverage increases volatility due to higher financial risk, while Septyadi & Bwarleling (2020) noted that active trading leads to rapid price changes. Earnings volatility also contributes to price fluctuations, as inconsistent earnings increase market uncertainty.

Macroeconomic factors such as inflation and interest rates further drive stock price volatility. Andayani et al. (2021) emphasized the significant long-term impact of interest rates on stock indices, while inflation affects investor confidence and purchasing power.

The volatility of digital banking stocks in Indonesia, particularly during the COVID-19 pandemic, reflects the sector’s sensitivity to both internal dynamics and external shocks (Rosihan et al., 2022). This study explores these factors to provide insights into the digital banking market and investment strategies.

2.5 Leverage

Leverage reflects the extent to which a company relies on borrowed funds to finance its operations. A higher Debt to Equity Ratio (DER) indicates greater dependence on external loans, increasing financial risk. According to L (2019) high leverage is often associated with higher stock price volatility due to increased liquidity risks. While investors may favor high leverage for potentially higher returns, creditors prefer lower leverage for loan security.

Rosyida et al. (2020) found a positive correlation between leverage and stock price volatility, emphasizing that high debt levels amplify financial instability. However, Dewi & Suaryana (2016) argued that leverage does not significantly affect stock volatility, as investors focus more on a company’s revenue-generating capability than its debt levels.

2.6 Firm Size

Firm size is measured by the total assets of a company and reflects its market presence and operational capacity. Larger firms often have more diversified operations and greater access to market information, leading to lower stock price volatility (Rosyida et al., 2020). Ayuning Putri (2020) found that firm size does not influence stock price volatility, as well-managed companies maintain investor confidence regardless of size. Contrarily, Yulinda et al. (2020) and Hieu Nguyen et al. (2020) reported a negative relationship between firm size and volatility, highlighting that larger firms’ stability reduces price fluctuations.

2.7 Trading Volume

Trading volume represents the number of shares traded in a given period, indicating market activity. High trading volumes often signal active investor participation and can lead to increased stock price volatility (Septyadi & Bwarleling, 2020). The asymmetric information model suggests that investors with private information influence trading volumes and price movements. Dewi & Suaryana (2016) found a positive correlation between trading volume and volatility, as rapid trading responses to new information increase price fluctuations. However, Utami & Purwohandoko (2021) argued that trading volume does not significantly impact stock volatility, as investors may not always consider trading volume when making investment decisions.

2.8 Earnings Volatility

Earnings volatility measures the stability of a company’s profits over time. High earnings volatility indicates inconsistent earnings, which can increase stock price fluctuations due to investor uncertainty (Hidayati & Sukmaningrum, 2021). While Ayuning Putri (2020) found no significant impact of earnings volatility on stock prices, Cahyawati & Miftah (2022) and Mehmood et al. (2019) reported both positive and negative influences, suggesting that fluctuating profits can either deter or attract investors depending on market conditions. Companies with stable earnings are often perceived as low-risk investments, contributing to lower stock volatility.

2.9 Inflation

Inflation, as defined by Bank Indonesia, is the general increase in prices continuously over time, reducing the purchasing power of consumers if income growth does not keep pace. Inflation affects the economy broadly, influencing consumption, production costs, and investment decisions. Aten & Nurdiniah (2020) noted that abnormal inflation disrupts company returns to investors, thereby affecting stock price variability. High inflation increases production costs, prompting companies to raise product prices, which may hurt their financial performance and stock value.

Market participants often react negatively to high inflation due to rising costs and potential economic tightening measures. Octavian et al. (2022) found that inflation positively affects stock price volatility, as investors tend to sell off stocks when inflation spikes. However, Nur Azura et al. (2019) argued that inflation does not significantly influence stock volatility when it remains below 10%, particularly in sectors with high liquidity like digital banking.

2.10 Interest Rates

Interest rates influence borrowing costs and investment returns. Interest rates are tools used by Bank Indonesia to control money circulation and influence investor decisions (Dewi & Suaryana, 2016). Rising interest rates increase borrowing costs, reduce company profitability, and make stocks less attractive compared to low-risk investments like deposits (Siti Hayati, 2018).

While Safitri et al. (2022) found that interest rates did not significantly impact stock volatility, this study showed a positive relationship between interest rates and digital banking stock volatility, as higher rates lead to greater price fluctuations due to changing investor preferences. Nur Azura et al. (2019) highlighted that rising interest rates can drive investors to safer assets, thereby affecting stock prices in digital banking.

3. Material and Methods

3.1 Research Design, Sample, and Data

The study used a quantitative secondary-data design. The population comprised seven digital banks identified as listed on the Indonesia Stock Exchange during the study period. Five banks met the source study criteria of being listed during 2019-2022, publishing the required monthly financial information, and having sufficiently complete data for the variables analysed: PT Bank Jago Tbk (ARTO), PT Allo Bank Indonesia Tbk (BBHI), PT Bank Neo Commerce Tbk (BBYB), PT Bank MNC Internasional Tbk (BABP), and PT Bank Raya Indonesia Tbk (AGRO).

The observation period covered January 2019 through December 2022. Monthly company-level data were obtained from Indonesia Stock Exchange publications and company financial reports, while macroeconomic series were obtained from official Indonesian sources, including Bank Indonesia as documented in the source study. The five banks over 48 months imply 240 potential bank-month observations. The reported estimation output retained 220 observations in the long-run equation and 209 observations in the short-run ECM after source-study adjustments and differencing.

3.2 Variable Measurement

Stock price volatility was the dependent variable and was measured from the monthly highest and lowest share prices, following the extreme-value approach used in the source study. The independent variables were leverage, firm size, trading volume, earning volatility, inflation, and the policy interest rate. Table 1 summarises the operationalisation retained from the source study.

3.3 Econometric Procedure

The analysis was conducted in EViews 10. Stationarity was assessed with the Augmented Dickey-Fuller (ADF) procedure, which extends the unit-root testing framework developed by (Dickey & Fuller, 1979). The source analysis then applied the Johansen cointegration procedure (Johansen, 1988) and an error-correction specification consistent with the long-run or short-run framework associated with cointegration and error correction (Engle & Granger, 1987). Statistical significance was evaluated at α = 0.05.

The reported long-run specification was:

LN_VHSₜ = α + β₁LVRGₜ + β₂FSIZEₜ + β₃TVOLₜ + β₄EVOLₜ + β₅INFₜ + β₆IRₜ + εₜ

The reported short-run ECM was:

ΔLN_VHSₜ = α + β₁ΔLVRGₜ + β₂ΔFSIZEₜ + β₃ΔTVOLₜ + β₄ΔEVOLₜ + β₅ΔINFₜ + β₆ΔIRₜ + γECTₜ₋₁ + εₜ

where VHS denotes stock price volatility; LVRG leverage; FSIZE firm size; TVOL trading volume; EVOL earning volatility; INF inflation; IR the interest rate; Δ the first-difference operator; and ECTₜ₋₁ the lagged error-correction term. Diagnostic checks reported in the source study included residual normality using Jarque-Bera, multicollinearity using variance inflation factors (VIF), autocorrelation using Durbin-Watson statistics, and heteroskedasticity using an ARCH test.

4. Results

4.1 Descriptive Statistics

Table 2 presents the minimum, maximum, and mean values reported in the study dataset for the seven variables analysed.

4.2 Stationarity and Cointegration

The ADF results at level showed that firm size was not stationary (p = 0.3818), whereas the other variables had p-values below 0.05. At first difference, all reported variables had p-values of 0.0000. The Johansen trace output reported four cointegrating equations at the 5% level. Table 3 summarises the reported unit-root results.

4.3 Long-Run and Short-Run Estimates

Table 4 reports the numerical estimates from the source EViews output. In the long-run equation, firm size, trading volume, and the interest rate were statistically significant at the 5% level. Firm size and trading volume had negative coefficients, whereas the interest rate had a positive coefficient. Leverage, earning volatility, and inflation were not statistically significant. The long-run model had R² = 0.176268 and Prob(F-statistic) < 0.001.

In the short-run ECM, only the change in the interest rate was statistically significant among the six explanatory variables. The lagged residual/error-correction term was negative and significant (coefficient = -0.890171; p < 0.001). The short-run model had R² = 0.550791 and Prob(F-statistic) < 0.001.

4.4 Diagnostic Tests

The reported diagnostic statistics indicate that the residual normality test did not reject normality, the VIF values were well below 10, and the ARCH test did not indicate heteroskedasticity at the 5% level. The Durbin-Watson statistics were close to 2 in both equations.

5. Discussion

5.1 Firm Size and Trading Volume

Firm size had a significant negative long-run association with stock price volatility. This direction is consistent with the argument that larger firms tend to have broader operations, greater information availability, and potentially lower information asymmetry, which can dampen extreme price adjustment. Negative firm-size relationships have also been documented in Indonesian and regional stock-volatility studies (Rosyida et al., 2020; Selpiana & Badjra, 2018), although positive or sample-dependent effects are reported elsewhere (Kengatharan & Ford, 2021). The present result is therefore best interpreted as a long-horizon association specific to the digital-bank sample rather than as a universal size effect.

Trading volume was also negative and significant in the long run but insignificant in the short run. Earlier studies provide mixed evidence: Priana and Muliartha reported a negative relationship (Priana, 2017), while other Indonesian studies reported positive or insignificant effects (Hidayati & Sukmaningrum, 2021; Septyadi & Bwarleling, 2020; Utami & Purwohandoko, 2021). One plausible interpretation is that sustained trading activity may improve liquidity and information incorporation over longer horizons, whereas month-to-month changes in volume are too noisy to explain immediate volatility after the other determinants are considered. This horizon-specific pattern helps reconcile apparently conflicting findings in the prior literature.

5.2 Interest Rate

The interest rate was the only determinant that remained statistically significant in both the long-run and short-run specifications, with positive coefficients. Interest-rate changes can affect equity values through discount rates, funding conditions, expected credit demand, bank margins, portfolio rebalancing, and investors’ relative preference for fixed-income or deposit instruments. Broader capital-market research documents substantial equity-price responses to monetary-policy surprises (Bernanke & Kuttner, 2005), while evidence focused on bank equity values shows that the transmission can be particularly relevant for banking-sector valuations and can vary with the interest-rate environment (Ampudia & Heuvel, 2022).

Within this sample, the positive coefficient indicates that higher interest-rate observations were associated with greater digital-bank stock price volatility. The result is consistent with the source study’s view that changes in rates can trigger portfolio shifts and repricing in bank equities, but it should not be read as implying that all interest-rate increases necessarily raise bank equity values or volatility in every regime. The contribution of the present result is its persistence across both temporal horizons.

5.3 Leverage, Earning Volatility, and Inflation

Leverage was not significant in either horizon. Similar findings appear in studies where debt-to-equity ratios did not independently explain stock-price volatility (Hidayati & Sukmaningrum, 2021; Utami & Purwohandoko, 2021). For banks, leverage is also structurally related to financial intermediation, so a simple debt-to-equity ratio may convey different information than it does for non-financial companies. The result should therefore be interpreted narrowly: leverage, as operationalised in this study, did not add statistically significant explanatory power after the other included factors.

Earning volatility was likewise insignificant in the long and short run. Prior Indonesian evidence also reports insignificant earnings-volatility effects in some finance-sector samples (Ayuning Putri, 2020b). In this study, the variable was operationalised as operating profit divided by total assets, following the source thesis. Consequently, the measure captures fluctuations in an operating-profit-to-assets proxy rather than a more elaborate conditional volatility process; this measurement choice may partly explain the weak association with stock-price volatility.

Inflation was not significant in either specification. Previous Indonesian studies likewise report both insignificant and significant inflation effects (Nur Azura et al., 2019; Octavian et al., 2022). The 2019-2022 sample contains a long period of relatively moderate inflation followed by an acceleration in 2022, while policy interest rates changed more directly across the same period. In the estimated model, the policy-rate variable therefore appears to have captured a more immediate monetary-condition channel than inflation itself.

5.4 Error Correction and Practical Implications

The error-correction coefficient of -0.890171 was negative and highly significant. Interpreted within the reported ECM, this implies that approximately 89% of a short-run disequilibrium relative to the estimated long-run relation was corrected in the following observation period. The short-run equation also produced a higher R² than the long-run equation, indicating that monthly changes and adjustment dynamics accounted for a substantial portion of observed volatility even though only the interest rate was individually significant among the differenced regressors.

For investors, the findings suggest separating short-horizon risk monitoring from longer-horizon assessment. Monetary-policy conditions deserve close attention at both horizons, whereas firm size and trading activity may be more informative for persistent volatility patterns. For digital-bank management, the results reinforce that market risk cannot be understood only from internal financial indicators; external monetary conditions can materially affect how the market reprices bank equity risk.

6. Conclusion

This study examined the long-run and short-run determinants of stock price volatility for five Indonesian digital banks over 2019-2022. The reported ECM estimates show a horizon-dependent pattern. Firm size and trading volume were negatively associated with volatility in the long run, while the interest rate had a positive and significant relationship in both the long and short run. Leverage, earning volatility, and inflation were not statistically significant in either horizon. The significant negative error-correction term indicates rapid adjustment toward the reported long-run equilibrium after short-run disturbances.

The findings imply that monetary conditions were the most consistent external correlate of digital-bank equity volatility in this sample, whereas size and trading activity were associated with longer-run rather than immediate volatility. The study is limited by the four-year period, the five-bank sample, and measurement choices inherited from the source dataset. In addition, the reported ADF output shows mixed integration orders at level, while observations from multiple banks were combined in the source estimation. These features warrant caution when generalising the conventional Johansen-ECM estimates. Future research should replicate the analysis with a longer dataset and estimators explicitly suited to mixed-order and/or panel structures, such as ARDL/panel-ARDL, pooled mean group, or panel error-correction approaches, and should consider additional market and macroeconomic controls.