Market Reactions to Fintech M&A: Evidence from Event Study Analysis of Financial Institutions

Abstract

The rise of fintech has disrupted traditional financial services, prompting banks and asset managers to respond strategically, often through mergers and acquisitions. This study investigates the short-term market reaction to M&A announcements involving fintech targets by incumbent financial institutions. Using an event study methodology centered on different event windows and cumulative abnormal returns computed via the market model, the analysis incorporates regression models with bidder-, deal-, and target-level variables to identify the drivers of performance. The results show that, on average, financial institutions experience negative abnormal returns around announcement dates, suggesting limited short-term value creation. Higher market-to-book ratios and tax rates are positively associated with CARs, while lower profit margins are linked to better market reactions. Subsample comparisons reveal that U.S. acquirers underperform their European peers, commercial banks fare worse than asset managers and investment banks, and pre-COVID-19 deals yield more favorable returns than post-COVID-19 ones. Robustness checks using different market benchmarks demonstrate that key patterns—especially those related to geography and timing—are sensitive to benchmark selection. Overall, this study highlights market skepticism toward fintech acquisitions by traditional financial institutions, particularly in specific contexts, and emphasizes the importance of controlling for structural factors when interpreting abnormal returns.

1. Introduction

The increased digitization and digitalization of the financial industry, i.e., the transformation of information from a physical to a digital format and the adoption of digital solutions in business models, has had significant effects on the affairs of traditional financial institutions like banks and asset management companies (Li et al., 2017). Responsible for these disruptions are a multitude of factors, the most important ones being advances in technology brought forward by large tech companies and new financial technology (fintech) firms (Ochirova & Miriakov, 2025), radical changes in customers’ behaviors (Collevecchio et al., 2024), and strict regulatory hurdles that traditional financial institutions have been subjected to as a response to the 2008 Global Financial Crisis (Li et al., 2017). These changes have posed challenges for banks, which suffer from being heavily regulated entities, often of significant size and having to rely on dated legacy information systems making it difficult for them to respond to the innovations in products and services introduced by fintech firms (Collevecchio et al., 2024). The advantages these firms may have over traditional financial institutions are a consequence of their digital business models, which leverage automated processes and require a limited amount of infrastructure, leading to more convenient, accessible, and user-friendly products and services (Ochirova & Miriakov, 2025). One answer to the threat that these companies can pose to incumbent financial institutions can be M&A: traditional financial institutions can acquire the capabilities, talent, and intellectual property (IP) of fintech firms while leveraging their existing business and networks to offer better products without having to develop the IP in-house, a process which can take a long time and does not guarantee results (Collevecchio et al., 2024). In recent years, global fintech investment has surged: in 2021, total fintech funding (including M&A, venture capital, and private equity) reached approximately USD 238.9 billion across about 7321 deals, with the fintech-M&A component alone valued at roughly USD 105.1 billion. In 2022, despite a slowdown, fintech investment still amounted to about USD 164.1 billion across 6006 deals, including USD 73.9 billion in M&A transactions (KPMG, 2023). By 2024, however, global fintech investment had fallen further to USD 95.6 billion across 4639 deals, marking the lowest level since 2017 (KPMG, 2025). These figures illustrate the growing but volatile strategic importance of fintech acquisitions for incumbent financial institutions. Accordingly, there are several studies in the literature that investigate the M&A activity of banks and other financial institutions bidding for fintech targets. However, when it comes to studies that assess what is the short-term market performance of the bidders around announcement of the transaction, most analyses either focus on the entire group of financial institutions, or they limit their investigation to banks, often defined broadly. It is in light of these considerations that this study has been conducted to assess whether the market has different views on the potential outcome of an M&A deal depending on the specific type of financial institution involved, be it a commercial (or mostly commercial) bank or an asset manager or investment bank, and whether the market reaction to fintech M&A announcement has changed since after the 2020 COVID-19 pandemic. Methodologically, this study applies an event study approach (MacKinlay, 1997) to measure cumulative abnormal returns (CARs) of acquiring firms around fintech M&A announcements. The event study is complemented by a set of OLS regressions to identify deal- and firm-specific drivers of CARs, and by two-sample tests to validate differences across subgroups. The sample is built from S&P Capital IQ data and comprises 90 qualifying transactions announced between 2008 and 2024, in which publicly traded banks, asset managers, or investment banks acquired fintech firms. This paper contributes by providing one of the first comparative analyses of short-term market reactions to fintech acquisitions across different types of financial institutions, thereby extending the literature beyond bank-only samples.

2. Literature Review

The review of the existing literature on the topic of M&A between fintech firms and traditional financial institutions has been structured into three sections: an examination of the literature on mergers and acquisitions, followed by an assessment of the literature on financial institutions’ M&A, and lastly a review of the studies available on mergers and acquisitions involving high tech and fintech firms.

2.1. Mergers and Acquisitions

The literature on M&A consists of a robust body of work that includes several studies analyzing the rationale for pursuing M&A as well as the performance of both the target companies and acquirers in relation to an acquisition or a merger. Berkovitch and Narayanan (1993) distinguish between different motives that can drive takeovers, namely, synergies, agency, and management hubris. The authors find that in takeovers that generate positive gains for the combined entity, there is a positive correlation between these gains and those of the target, suggesting that synergies may be the drivers of these acquisitions. They also find that the hubris motive can be associated with takeovers where the correlation between target and acquirer gains is not statistically significant while still being positive for target and total gains. Lastly, in deals where the combined entity reports negative gains, the agency motive could be the driver of M&A as there is a negative correlation between target and total gains, consistent with the idea that agency-driven takeovers are pursued by acquirers’ managers for their own self-interests and at the expense of acquirers’ shareholders. The hubris hypothesis was introduced by Roll (1986) and suggests that the value of M&A targets should increase, while the value of bidders should decrease, resulting in a slightly negative performance for the combined entity, so that in hubris-driven M&A, there is effectively a transfer of wealth from bidder shareholders to target shareholders caused by the overconfidence of the bidder’s management. Analyzing synergies, Alhenawi and Krishnaswami (2015) found that in the long run they tend to materialize, generating excess value, and that they tend to be more relevant in diversifying mergers than in deals between companies in the same industry, suggesting that the former takeovers may indeed be driven by synergies while for the latter different considerations may be more relevant. Looking at the evidence for value creation in M&A, Agrawal and Jaffe (2000) show in their review of the literature that mergers tend to generate significant positive returns for target firms, but the long-term performance of acquirers is generally negative or insignificant. Kellner (2023) finds that European target companies still report positive returns, while bidders show no meaningful stock price reaction. Rau and Vermaelen (1998) find that long-term abnormal performance of “glamour acquirers”, high market-to-book-value companies, is significantly negative, suggesting that the market is too optimistic about their future performance because of successful past results, and is backed by Sudarsanam and Mahate (2003), which draw similar conclusions arguing that glamour acquirers underperform post-acquisition and that cash deals outperform stock transactions as the latter suggest to the market that the stock is overvalued. Ang and Cheng (2006) use again the market-to-book value of bidders to estimate overvaluation, finding that overvalued firms tend to pursue more stock transactions and that it also has a positive impact on the likelihood of a successful combination. They argue that this is evidence of the validity of the market-driven acquisition theory proposed by Shleifer and Vishny (2003). Lastly, Rosen (2006) finds that market reaction to an acquisition is positively correlated with the reactions to other recent M&A deals. He also estimates that short-term returns in periods featuring large M&A activity tend to be more positive than in other periods.

2.2. Financial Institutions M&A

When focusing specifically on the acquisitions carried out by financial institutions, Cybo-Ottone and Murgia (2000) find evidence in European banking M&A suggesting that banks report significant positive announcement returns, and that while combinations with insurance players contribute to this result, acquisitions of foreign targets do not generate the same benefits. The author finds that these results are contrary to the evidence relative to banking M&A in the US and attributes this contrast in results to the different regulatory markets and levels of competition that characterize the European and the US markets. In analyzing whether diversification, in terms of both geography and activity, affects value in banking M&A, DeLong (2001) proves that transactions that concentrate banks’ activities and geographies tend to be viewed more positively by the market, generating positive abnormal returns, while diversification in general does not help banks create value in financial markets. Still on the topic of European banking M&A, Beitel et al. (2004) find results that are in contrast with those obtained by Cybo-Ottone and Murgia (2000). Specifically, the authors find that M&A does not generate positive and significant cumulative abnormal returns and that, if anything, these seem to be slightly negative. However, they also find evidence supporting DeLong (2001), proving that acquisitions where the target shares the same activity or geography with the bidder outperform transactions focused on diversification. Another study on European banking M&A is Leledakis and Pyrgiotakis (2019), which investigates the acquisitions that take place after the financial crisis. The authors find that banking M&A generates positive returns which are significantly higher than pre-2009 returns and attribute these positive results to the increase in market concentration following the 2008 financial crisis, as well as to greater market power.

2.3. Technology and Fintech M&A

A solid body of literature has been developed to study the combinations between firms for purposes of technological development. One important topic in this area of study is post-merger integration. Puranam et al. (2006) find that integration decisions affect the outcome of technology acquisitions, arguing that large firms buying smaller technology companies should align their integration model with the maturity of the target as younger target firms require a greater degree of flexibility compared to more mature ones. Ranft and Lord (2002) argue that one of the issues to consider during the integration of technology companies is knowledge transfer. They find several factors that affect this knowledge transfer, such as granting a degree of autonomy to the acquired firm (a factor that Puranam et al. (2006) have also considered) and retaining key employers. In terms of value creation, Sears and Hoetker (2014) find that the degree of technological overlap between target and bidder negatively affects the creation of value in technological acquisitions computed using cumulative abnormal returns. Focusing specifically on fintech M&A, Hornuf et al. (2021) analyze the diverse types of alliances that banks can form with financial technology firms and find that large banks invest in smaller fintech firms while they decide to collaborate with larger ones. Additionally, they find that the announcement of fintech alliances is associated with negative announcement returns, but the opposite happens for digital banks, which benefit from a strategic alliance. Focusing on the Italian market, Faes et al. (2022) find that out of 90 public partnerships established between banks and fintech firms, 84 did not involve any equity investment. The existing literature on the link between value creation and fintech M&A provides some insights on the main factors that influence announcement performance. Kueschnig and Schertler (2023) look at US financial firms stock performance around announcement of both fintech and non-fintech deals and find that the former generate higher cumulative average abnormal returns compared to the latter. Cappa et al. (2022) focus instead on the impact of bank-fintech M&A on banks’ future profits, which are proxied by banks’ abnormal returns around announcement, assuming Fama’s (1970) market efficiency theory holds. The authors find that the type of acquisition has an impact on abnormal returns, with full acquisitions resulting in negative abnormal returns, and that the acquisition of players in the payment sector results in significantly negative future profits. Carlini et al. (2022) look at all investments made by banks in fintech firms, including equity investments in startups’ capital raising rounds, finding evidence that fintech investments generate negative abnormal returns for bidders. Collevecchio et al. (2024) explores original variables to find that the best outcomes in bank-fintech M&A arise from minority acquisitions carried out by sustainable banks (high ESG score) and great institutional distance between the bank and the fintech firm’s country. More recently, Zheng and Mao (2024) provide a comprehensive global analysis of fintech M&A and confirm that market reactions remain highly heterogeneous across regions and business models, with fintech acquisitions generally generating short-term value destruction for traditional financial institutions. Their findings reinforce the view that integration and valuation uncertainty remain critical drivers of announcement-period underperformance. Similarly, Alfhaili et al. (2025) investigate how banks’ business models affect the outcomes of fintech acquisitions. They show that banks with more diversified business models tend to engage in fintech M&A for strategic transformation rather than efficiency gains, often facing greater market skepticism reflected in weaker abnormal returns. This study highlights the importance of accounting for heterogeneity among financial acquirers when assessing fintech deal performance. Complementing these insights, Maurici et al. (2025) analyze the post-acquisition effects of bank–fintech deals on lending, deposit growth, and risk-taking. Their evidence suggests that fintech acquisitions can enhance banks’ technological capabilities and deposit growth, but also increase risk-taking, especially when integration is rapid. This recent research underscores that fintech M&A outcomes depend strongly on the acquirer’s strategic motives and integration approach, providing additional context for the mixed performance patterns observed in event studies. Lastly, Wang (2024) argues that fintech M&A increases firm value, measured by Tobin’s Q and market-to-book ratio, for financial bidders.

2.4. Hypotheses

Despite extensive research, there is no consensus on whether fintech M&A creates value for bidders. In this study, we argue that traditional players in the financial sector and fintech firms generally cooperate with each other through alliances and partnerships (Hornuf et al., 2021; Faes et al., 2022). It is not often the case that fintech firms target the same customers of other traditional financial institutions: while digital banks and neobanks compete directly with traditional banks, other fintech firms generally do not enter into direct competition with financial institutions but rather opt for collaborations that can be mutually beneficial (Murinde et al., 2022). Accordingly, financial firms buying fintech firms would be falling under the category of diversifying mergers and as such they should destroy value for bidder shareholders, as existing literature suggests (DeLong, 2001; Beitel et al., 2004; Alhenawi & Krishnaswami, 2015). Additionally, acquisitions would expose buyers to post-merger integration risks, further suggesting that fintech deals should destroy value for buying financial institutions.

Hypothesis 1.

Financial institutions acquiring fintech targets report negative statistically significant cumulative average abnormal returns around deal announcement.

Reviewing the literature, we recognized that most studies focus on banks’ announcement returns (Cappa et al., 2022; Carlini et al., 2022; Collevecchio et al., 2024) or on the returns of financial institutions in general, (Kueschnig & Schertler, 2023; Wang, 2024). To the best of this author’s knowledge, there are no studies that identify the performance of asset management companies and investment banks separate from that of commercial banks to assess whether they create value for bidder shareholders. However, it seems evident that for asset management companies, business combinations with fintech players active in the digital assets and digital investment segments have the strongest strategic rationales: these fintech firms can provide complementary services and often offer customers low prices as they face lower costs by leveraging digital technologies, providing services on user-friendly platforms, often including automated services. Since fintech firms seem to be a good fit for synergy-driven M&A for asset management companies and investment banks, we hypothesize that these companies’ announcement returns should outperform those of commercial banks.

Hypothesis 2.

Asset management companies and investment banks generate cumulative average abnormal returns that are significantly higher than those reported by commercial banks.

The outbreak of the COVID-19 pandemic and its related restrictions have had a significant impact on the financial system, with changing demands driving innovation. The Financial Stability Board (2019) reports that payment habits of customers have changed, with demand moving towards contactless and digital payments; digital banking has increased during the years of the pandemic, driven by customers’ inability to access bank branches and safety considerations; customers increasingly seek convenience in accessing financial services, and savings have gone up during COVID-19 lockdowns, increasing demand for equity securities, and thus for investment services. These changes suggest that demand is driving the push towards the digitalization of financial services, and fintech firms should be the ultimate beneficiaries of these trends in demand. However, while it is true that fintech could reap the benefits of these changes, banks have rapidly repositioned themselves and have managed to respond quickly to changing customer demands, for example offering online banking services, showing that external investments in fintech may not be necessary. Therefore, the thesis that investments in financial technologies would be necessary for traditional financial institutions to rapidly face competition from fintech firms proved incorrect, negatively impacting the market’s response to fintech M&A announcements. Recent research (Maurici et al., 2025) further supports this interpretation by showing that the pandemic accelerated banks’ digital transformation internally, reducing their dependence on external fintech acquisitions for innovation.

Hypothesis 3.

Cumulative average abnormal returns for fintech deals announced by financial institutions after the outbreak of the COVID-19 pandemic are significantly more negative compared to deals announced before the pandemic.

3. Materials and Methods

This research leverages an event study to analyze cumulative abnormal returns of bidding firms around M&A announcements. The event study is followed by the investigation of different OLS regression models to identify the drivers of anormal returns and by an analysis of the differences in abnormal returns between subsamples that uses a two-sample paired test to validate the significance of results. Lastly, the sample selection process that has been adopted in this study is presented.

3.1. Event Study

To investigate whether the acquisitions of fintech firms conducted by traditional financial institutions destroy value for the buyers’ shareholders in the short term, an event study approach was used. The event study approach follows MacKinlay’s (1997). Accordingly, two different time periods, namely the estimation window and the event window, have been identified around the announcement of the M&A deals for each security included in the sample. The estimation window consists of 200 days prior to the event window, which in turn consists of 41 days, including 20 days before announcement date and 20 days after announcement of the transaction. In those situations where markets were closed on announcement date, the following day has been considered as announcement day. The time periods have been specified in such a way as to avoid any overlap between the estimation window and the event window (

Table 1. Time windows for the event study analysis.

Windows	Time Periods	Number of Days
Announcement date	0	1	0
Estimation window	[−220, −21]	200	From T0 to T1 − 1
Event window	[−20, 20]	41	From T1 to T2

).

To compute the abnormal returns of securities, their daily prices have been collected alongside the daily prices of the indices selected as benchmarks (the market portfolios). The expected returns have been computed using the market model and the market-adjusted model.

3.2. Cumulative Abnormal Returns

Once the expected returns for the securities have been computed, the abnormal returns can be calculated as simply the difference between the actual returns registered by each security in each day and the expected returns of that security on that day. These abnormal returns represent the excess portion of returns that cannot be explained by normal movements in the market but rather can be attributed to the event that is being analyzed, which in this case means that abnormal returns reflect the reaction of the market to the announcement of an M&A transaction.

Once the abnormal returns have been computed for each security in the sample, the cumulative abnormal returns (CAR) could be calculated as the sum of the abnormal returns of a security within the event window. The test of the null hypothesis of ${CAR}_i$ equal to zero can be conducted to check where these abnormal returns are statistically significant; however, this information is not very useful as it does not allow us to draw any conclusion or generalizations on the entirety of our sample of M&A transactions. To assess whether fintech M&A involving financial buyers generates significant abnormal returns for the buyers’ shareholders in the short-term, we have to investigate the statistical significance of the aggregated cumulative abnormal returns around announcement. At this point of the analysis, we can test the statistical significance of results and find out whether financial institutions that acquire fintech firms see abnormal returns in the value of their securities around the announcement of the transaction or not.

3.3. OLS Regression Models

The next step of this study after investigating the cumulative abnormal returns registered around fintech M&A announcement is the identification of the possible drivers of these cumulative abnormal returns. These drivers can depend on either the characteristics of the acquiring firms, deal-specific factors, or target-specific factors, but considering that many fintech firms are private companies, target-specific factors have not been considered in the regression models bar for the fintech sector to which the targets belong. In line with prior event-study research on financial institutions and M&A (e.g., Houston et al., 2001; Hagendorff et al., 2008; Collevecchio et al., 2024), the explanatory variables included in the regression models are selected because they have been shown to capture firm size, financial health, growth potential, and deal characteristics that are relevant to the market’s assessment of value creation. In terms of buyer-specific characteristics, the variable $\mathrm{L}\mathrm{n}\left(TA\right)$, representing the natural logarithm of the total assets of the buyer before deal announcement, has been included in the regression model as a proxy of the size of the buyer, as larger firms may have more resources to integrate acquisitions but also face higher integration risks (Houston et al., 2001). The variable $Market-to-Book$ has also been included in the regression models. This variable has been used as a proxy for the solidity of the financial acquirer, as a higher price to book ratio indicates a solid balance sheet, especially for financial institutions, consistent with studies that find that firms with stronger balance sheets are better positioned to benefit from acquisitions (Hagendorff et al., 2008). One of the main rationales often presented to justify financial institutions and fintech M&A transactions is the fact that such acquisitions can boost a financial institution’s revenue growth and profitability. As such, the variable $Revenue\text{\_}growth$ was included in the OLS regression model. Under similar assumptions, the regression models also include the $ProfitMargin$ explanatory variable. Accordingly, Revenue_Growth and Profit Margin are included as proxies for the acquiring firm’s operational performance, consistent with Collevecchio et al. (2024), who highlight the importance of financial performance in determining market reactions to fintech acquisitions. The $TaxRate$ variable represents the acquirer’s effective tax rate and has been included as a proxy for tax synergies, including tax credits and potential tax rate cuts implemented by regulators to support the development of the financial technology industry, in line with prior work linking tax efficiency and M&A outcomes (De Mooij & Ederveen, 2008). The variable $Relativevalue$ is equal to the ratio between deal value and buyer’s market cap, a factor widely used in M&A studies (Moeller et al., 2004). This variable should capture the response of the market to the size of the fintech deal. Additionally, the $CashConsideration$, $FintechExperience$, and $Covid$ variables are dummy variables included to reflect any significant variability in cumulative abnormal returns that can be explained by the type of consideration paid by the buyer (the $CashConsideration$ variable takes the value 1 for an all-cash deal and 0 otherwise), by any previous experience of the buyer with fintech M&A in the five years prior to the announcement of the transaction being considered ($FintechExperience$ takes the value of 1 if the buyer has made such acquisitions and 0 otherwise), and by whether the deal was announced before or after the outbreak of the COVID-19 pandemic (the $Covid$ variable is equal to 1 if it has been announced after 1 March 2020 and 0 otherwise). Cash vs. stock payment has long been shown to affect announcement returns in M&A (Travlos, 1987), while prior acquisition experience and the pandemic environment represent important contextual factors. Lastly, the dummy variables $FinancialMedia$, $Payments$, $InvestmentTech$, $BankingTech$, $BusinessProcess$, and $DigitalLending$ were included in a single regression model (Model 5) to check if the target fintech sector can explain any variability in cumulative abnormal returns, as suggested by recent evidence on sector-specific effects in fintech M&A (Ochirova & Miriakov, 2025). These variables take the value of 1 if the target belongs to that sector and 0 otherwise. In all models except for Model 1, we included year-fixed effects.

The models are specified as follows:

Model 1

$$\begin{array}{l}{CAR}_i=\alpha +{\beta }_1\mathrm{Ln}\left(TA\right)+{\beta }_{2}Market\text{\_}to\text{\_}Book+{\beta }_{3}RevenueGrowth\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{4}ProfitMargin+{\beta }_{5}TaxRate+{\beta }_{6}RelativeValue+{\epsilon }_i\end{array}$$

(1)

Model 2

$$\begin{array}{l}{CAR}_i=\alpha +{\beta }_1\mathrm{Ln}\left(TA\right)+{\beta }_{2}Market\text{\_}to\text{\_}Book+{\beta }_{3}RevenueGrowth\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{4}ProfitMargin+{\beta }_{5}TaxRate+{\beta }_{6}RelativeValue\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+\sum Year\text{\_}fixed\text{\_}effects+{\epsilon }_i\end{array}$$

(2)

Model 3

$$\begin{array}{l}{CAR}_i=\alpha +{\beta }_1\mathrm{Ln}\left(TA\right)+{\beta }_{2}Market\text{\_}to\text{\_}Book+{\beta }_{3}RevenueGrowth\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{4}ProfitMargin+{\beta }_{5}TaxRate+{\beta }_{6}RelativeValue\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{7}CashConsideration+{\beta }_{8}FintechExperience\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+\sum Year\text{\_}fixed\text{\_}effects+{\epsilon }_i\end{array}$$

(3)

Model 4

$$\begin{array}{l}{CAR}_i=\alpha +{\beta }_1\mathrm{Ln}\left(TA\right)+{\beta }_{2}Market\text{\_}to\text{\_}Book+{\beta }_{3}RevenueGrowth\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{4}ProfitMargin+{\beta }_{5}TaxRate+{\beta }_{6}RelativeValue\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{7}CashConsideration+{\beta }_{8}Covid\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+\sum Year\text{\_}fixed\text{\_}effects+{\epsilon }_i\end{array}$$

(4)

Model 5

$$\begin{array}{l}{CAR}_i=\alpha +{\beta }_{1}FinancialMedia+{\beta }_{2}Payments+{\beta }_{3}InvestmentTech\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{4}BankingTech+{\beta }_{5}BusinessProcess\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_{6}DigitalLending+\sum Yea{r}_{fixe{d}_{effects}}+{\epsilon }_i\end{array}$$

(5)

To assess whether subsamples varying with respect to a specific variable show different announcement performance, the cumulative average abnormal returns of the subsamples have been compared to estimate any significant differences. The significance of the results was assessed through a two-sample t-test.

3.4. Sample Selection Process

The sample has been constructed by referring to the S&P Global Universal database provided by S&P Capital IQ. To identify the transactions relevant for this study, only those deals that have been announced between 1 January 2008 and 1 July 2024 have been considered. Moreover, only deals with a minimum transaction value greater than USD 1 million have been included in the sample. Since we are analyzing fintech M&A, the deals included in the sample are those that S&P Capital IQ considers “Financial Technology” deals. These are transactions where, according to S&P Capital IQ’s definition, the target belongs to the financial technology industry group. S&P Capital IQ defines companies in the “Financial Technology” sector as “establishments that provide a service or technology to the financial services industry”. Both deals that are pending and completed have been considered. Additionally, only deals where the buyer obtained control over the target have been included, thus eliminating minority acquisitions. This led to a sample of 1659 deals. Of this sample, only the acquisitions where the target fintech sector was one of the following, namely banking technology, business process outsourcing, digital lending, financial media and data solutions, investment and capital markets tech, payments, and security technology, have been considered. Since this study is based on the returns of buyers’ stock around announcement of a fintech M&A deal, only those transactions where the buyer was a publicly traded financial institution are relevant. Specifically, the type of buyers considered relevant were banks, asset management companies, and broker-dealers. These are defined by Capital IQ as follows:

▪

Banks: commercial banks whose businesses are derived primarily from conventional banking operations, such as retail banking and small and medium corporate lending.

▪

Asset management companies: financial institutions primarily engaged in investment management and/or related custody and securities fee-based services (includes companies operating mutual funds, closed-end funds, and unit investment trusts, but excluding banks and other financial institutions primarily involved in commercial lending, investment banking, brokerage, and other specialized finance activities).

▪

Broker-dealers: financial institutions primarily engaged in investment banking and brokerage services, including equity and debt underwriting, mergers and acquisitions, and securities lending and advisory services (excluding banks and other financial institutions primarily involved in commercial lending, asset management, and specialized financial activities).

Exchanges are considered broker-dealers, but they have been manually excluded from the sample to make sure that the broker-dealers group is representative of investment banks, further reducing the sample to a total of 99 transactions. Asset managers and investment banks are grouped together since many firms that offer investment banking services and are classified as investment banks also provide asset and/or wealth management services to their clients. The following step of the sample selection process was checking that all the relevant data was available to carry out the analysis of cumulative abnormal returns and that for every single stock, there had not been other fintech M&A announcements in the 200-day estimation window prior to the event window as well as in the event window itself. This led to the exclusion of another 7 transactions, reducing the sample to 92 deals. The final adjustments consisted of the exclusion of deals where the beta of the buyer’s stock estimated was not statistically significant. The result was a final sample of 90 transactions, if considering each stock’s local primary index as benchmark for expected returns, and two samples consisting of 86 transactions each if considering the MSCI World Banks index and the MSCI World Financials index as benchmarks. While 90 transactions may appear relatively small in absolute terms, this sample size is consistent with prior event-study research on specialized or emerging M&A segments (e.g., Dranev et al., 2019; Cappa et al., 2022; Carlini et al., 2022). Fintech M&A involving publicly listed financial acquirers remains a niche subset of global M&A activity, and the filtering criteria applied—public status, full acquisition, industry classification, data availability, and statistical validity of estimated betas—ensure both precision and internal consistency. Furthermore, the final dataset captures transactions spanning 16 years and multiple geographies, providing sufficient cross-sectional variation to identify statistically significant relationships. The standard errors are robust and the key coefficients reach conventional levels of significance (5% and 1%), suggesting that the sample, while focused, is sufficiently powered for the empirical design and allows for meaningful inference. Thus, this study prioritizes quality and comparability of observations over raw sample size, in line with best practices in event-study methodology.

3.5. Data

Data relative to the stock prices of acquirers and the daily prices of national market indices as well as global indices was retrieved from the FactSet database (

Table 2. Samples used in this study; the samples vary based on the indices that have been used as market portfolios to estimate expected returns.

	Local Indices	MSCI World Banks	MSCI World Financials
Market portfolio	Primary local equity index	MSCI World Banks	MSCI Financials
Number of deals	90	86	86
Number of acquirers	62	60	59

). The analysis involves three different samples, varying based on the indices being used as market portfolios. In the first sample, the “Local indices” sample, expected returns for every security are computed using the primary equity index of the country where the acquirer issuing the security is incorporated. For two securities1 trading on foreign exchanges, the indices used as benchmarks for the computation of the expected returns were the primary equity indices of the countries where these two bidders’ stocks were trading. The second sample, the “MSCI World Banks” sample, uses the MSCI World Banks index as market portfolio, while the benchmark used in the third and last sample, the “MSCI World Financials” sample, is the MSCI World Financials index. The MSCI World Banks and the MSCI World Financials samples are subgroups of the local indices sample. This complication allows us to check the robustness of results.

The country where the most fintech M&A deals have been announced is by far the United States, with a total of 50 transactions between 1 January 2008 and 1 July 2024, followed by Brazil with 7 and Japan with 4 (Figure 1). Altogether, the number of deals announced by European buyers reaches 19 transactions. Moreover, in 2024, the total value of deals spikes to an extremely high level as the Local indices sample includes the acquisition of Discover Financial Services by Capital One Financial Corporation, whose deal value alone is equal to ~US$USD 35.3 billion (Figure 2). Finally, descriptive statistics of the deal values of the transactions are described in

Table 3. Descriptive statistics of the deal values of the transactions included in the Local indices sample.

(USD mn)	Mean	Median	Max	Min	Standard Deviation
Deal Value	773.78	63.82	35,337.92	1.60	3804.26

.

To compute the OLS regression, the variables required for the analysis have been retrieved from the FactSet database as well as from the S&P Capital IQ database. Specifically, all data relative to the buyers’ financial fundamentals and buyers’ market data have been retrieved from FactSet, while deal-specific information and information on the target fintech sector has been collected using the S&P Capital IQ database.

4. Results

The following section investigates the results of the event study and discusses the outcome of the cumulative abnormal returns analysis for the different event windows. This analysis is followed by the discussion of the OLS regression outputs, considering as a dependent variable the CARs for the [0, 1] event window. The section ends with the presentation and discussion of the results of the differences in CAARs between subsamples.

4.1. CAR Analysis

The analysis of cumulative returns computed on event windows of different length provides significant results. Specifically, as

Table 4. Results of the CAR analysis across different event windows with the two models.

Event Window	Market Model			Market-Adjusted Model
	Local Indices	MSCI World Banks	MSCI World Financials	Local Indices	MSCI World Banks	MSCI World Financials
CAR [−20, 20]	−0.58% [−0.482]	−0.63% [−0.511]	−1.02% [−0.846]	−1.06% [−0.872]	−1.43% [−1.152]	−1.93% [−1.607]
CAR [−10, 10]	−1.46% * [−1.680]	−1.30% [−1.467]	−1.24% [−1.449]	−1.84% ** [−2.118]	−1.76% ** [−1.985]	−1.93% ** [−2.252]
CAR [−5, 5]	−1.57% ** [−2.505]	−1.63% ** [−2.535]	−1.56% ** [−2.514]	−1.94% *** [−3.084]	−1.97% *** [−3.068]	−2.05% *** [−3.307]
CAR [−3, 3]	−1.11% ** [−2.210]	−1.06% ** [−2.064]	−1.06% ** [−2.130]	−1.33% *** [−2.664]	−1.21% ** [−2.364]	−1.30% *** [−2.630]
CAR [−1, 1]	−0.50% [−1.526]	−1.17% *** [−3.500]	−1.12% *** [−3.464]	−0.64% * [−1.939]	−1.19% *** [−3.544]	−1.18% *** [−3.636]
CAR [0, 1]	−0.49% * [−1.833]	−0.98% *** [−3.594]	−0.93% *** [−3.521]	−0.64% ** [−2.408]	−1.03% *** [−3.757]	−1.01% *** [−3.794]

Significance at the 90%, 95%, and 99% confidence levels is denoted by *, **, and ***, respectively; values for the t-tests are reported in squared brackets. The sample size for the local indices sample is 90 observations; the sample size for both MSCI World Banks and MSCI World Financials samples is 86 observations. The relevant information on the indices used can be found in the Appendix A.

shows, transactions where financial institutions acquire fintech companies tend to destroy value for the shareholders of the acquirers in the short term. Considering here only CARs computed using the market model where local indices are used as market portfolios for expected returns, one can see that fintech M&As tend to result in a drop in the acquirer’s stock price in the two-day window including the announcement date and the next day of around 0.5%. This result is statistically significant at the 10% level. By expanding the event window of interest, we can see that CARs become more negative except for the CAR [−1, 1], which are not statistically significant. The largest drop is observed in the [−5, 5] window, where acquirers’ stocks tend to fall by approximately 1.6%. Here, results are statistically significant as well. As we can see from the table below, all the robustness tests support the hypotheses that fintech M&As indeed destroy value for financial acquirers.

4.2. OLS Regression

Under the assumption of semi-strong market efficiency, the new information that the market receives, namely the M&A deal announcement, should be swiftly reflected into prices in the [0, 1] window. Moreover, changes in acquirers’ stock prices should directly reflect the impact of M&A deal announcement without there being other variables diluting or intensifying it. Before performing the regression analysis, the variables have been tested for multicollinearity issues. The OLS regression analysis indicates that the variable $Market-to-Book$ has a positive coefficient and is statistically significant at either the 10% level or the 5% level, depending on the model that we take into consideration (

Table 5. Cumulative abnormal returns in the event window [0, 1] computed using local indices as market portfolios; CARs are computed using the market model.

CAR (0, 1) Local Indices	1	2	3	4	5
Constant	−0.009 [0.018]	0.005 [0.026]	0.006 [0.026]	0.005 [0.026]	0.006 [0.023]
Ln(TA)	0.000 [0.001]	−0.001 [0.001]	−0.001 [0.002]	−0.001 [0.001]
Market-to-Book	0.002 * [0.001]	0.003 ** [0.001]	0.003 ** [0.001]	0.003 ** [0.001]
Revenue_Growth	0.002 [0.002]	−0.002 [0.003]	−0.002 [0.003]	−0.002 [0.003]
Profit_Margin	−0.009 [0.015]	−0.011 [0.015]	−0.011 [0.015]	−0.010 [0.015]
Tax_Rate	0.012 * [0.006]	0.016 ** [0.007]	0.015 ** [0.007]	0.015 ** [0.007]
Relative_Value	−0.002 [0.039]	0.008 [0.043]	0.000 [0.045]	0.002 [0.044]
Cash_Consideration			−0.005 [0.007]	−0.005 [0.007]
Fintech_Experience			−0.002 [0.007]
COVID				−0.010 [0.025]
Financial_Media					0.000 [0.014]
Payments					0.012 [0.012]
Investment_Tech					0.003 [0.013]
Banking_Tech					-
Business_Process					0.005 [0.016]
Digital_Lending					−0.015 [0.023]
Year-Fixed Effects	No	Yes	Yes	Yes	Yes
Number of Observations	90	90	90	90	90
Df	83	67	65	66	68
R-Squared	0.107	0.344	0.351	0.350	0.311

Significance at the 90% and 95%, confidence levels is denoted by * and **, respectively; standard errors are reported in squared brackets. The sample size for the local indices sample is 90 observations. The outputs of the other regressions in the [0, 1] event window are reported in the Appendix A.

). The $TaxRate$ variable has a positive coefficient in all the models where it has been included and is always statistically significant. Both the $RevenueGrowth$ and the $ProfitMargin$ variables are not statistically significant, although, as one could initially suspect, their coefficients generally happen to be negative. It is worth noting that the Market-to-Book and Tax Rate variables consistently appear as significant predictors of abnormal returns in the [0, 1] window, suggesting that firms with higher market valuations relative to book value and more favorable tax conditions may experience more positive investor reactions to M&A announcements. However, the overall robustness of the OLS results is limited. When using the MSCI World Banks index instead of local indices as the market benchmark, the statistical significance of variables shifts substantially: for instance, Profit Margin becomes the only statistically significant variable, and it shows a negative sign. This shift highlights that CARs are sensitive to the choice of benchmark, and such sensitivity should be taken into account when interpreting the regressions. However, the overall explanatory power of the regressions is limited, as shown by the relatively low R-squared values (ranging from 0.107 to 0.351). This suggests that a substantial portion of variation in CARs remains unexplained, which is not unusual in event studies where firm-specific and market-wide factors not captured in the model may influence short-term stock price reactions.

The variables representing the targets’ fintech sector are not statistically significant. Lastly, the variable $FintechExperience$, which is a dummy variable set equal to one for acquirers that have carried out a fintech deal in the five years prior to the M&A deal that is being considered, lacks statistical significance.

4.3. Differences in CAARs

The comparative analysis of CAARs across different subsamples reveals several statistically significant variations, offering a deeper understanding of how market participants perceive M&A announcements. The results reported in the table below reveal that in the [0, 1] window, there is a meaningful, though not robust, negative difference between the cumulative abnormal returns around announcement date reported by US-based acquirers compared to the CAARs of European fintech buyers (

Table 6. Differences between the CAARs of separate subsamples; the CAARs have been estimated using local indices as market portfolios. CARs are computed using the market model. The table reports the values of the differences, the results of the t-tests, the number of degrees of freedom, and the estimated p-values. CAAR_AM represents the cumulative average abnormal returns of asset management companies and investment banks.

Event Window [0, 1]
Local Indices	Difference	t-Test	df	p-Value
CAARUS − CAAREU	−0.27% **	−2.334	28.480	0.027
CAARAM − CAARBanks	0.21%	1.598	22.400	0.124
CAARPre-COVID − CAARPost-COVID	0.30% ***	3.279	53.572	0.002
CAARAdvanced − CAAREmerging	0.09%	0.452	15.830	0.658

Significance at the 95%, and 99% confidence levels is denoted by **, and ***, respectively.

). This difference of −0.27% is statistically significant at the 5% level. However, when changing the benchmark to the MSCI World Banks index, the direction of this effect reverses: US acquirers outperform European ones by 0.71%, a difference that is highly significant (

Table 7. Differences between the CAARs of separate subsamples; the CAARs have been estimated using the MSCI World Banks index as market portfolio. CARs are computed using the market model. The table reports the values of the differences, the results of the t-tests, the number of degrees of freedom, and the estimated p-values. CAAR_AM represents the cumulative average abnormal returns of asset management companies and investment banks.

Event Window [0, 1]
MSCI World Banks Index	Difference	t-Test	df	p-Value
CAARUS − CAAREU	0.71% ***	5.352	24.924	0.000
CAARAM − CAARBanks	0.74% ***	5.790	21.752	0.000
CAARPre-COVID − CAARPost-COVID	1.19% ***	12.781	51.732	0.000
CAARAdvanced − CAAREmerging	−0.56% **	−2.774	13.779	0.016

Significance at the 95%, and 99% confidence levels is denoted by **, and ***, respectively.

). This benchmark-dependent reversal highlights that the estimation of abnormal returns is sensitive to the reference market used to compute expected returns. Local indices tend to capture region-specific dynamics and investor expectations, while global benchmarks such as the MSCI World Banks index reflect broader international market trends and global investor sentiment. Therefore, when the market model is anchored to a global index, the relative performance of acquirers may appear stronger or weaker depending on whether their domestic market outperformed or underperformed global peers during the event period. In practical terms, this means that US-based institutions—whose stocks are more closely aligned with global financial trends—benefit more when global benchmarks are used, while European acquirers’ performance appears relatively weaker under such a benchmark. Conversely, using local indices isolates region-specific market movements, potentially dampening global effects. Hence, the observed variation across benchmarks does not necessarily indicate instability in the results but rather reflects differences in the underlying market representation and investor base captured by each model. In terms of differences in market expectations of acquirers’ performance post-M&A based on the degree of development of their countries’ economies, results are not clear: by considering the local indices sample, the two cumulative abnormal returns seem comparable with no statistically significant difference between the two subsamples, whereas by considering the MSCI World Banks sample, the difference is negative and statistically significant at the 1% level. This suggests that emerging market acquirers may face higher skepticism or lower investor confidence when assessed from a global perspective. When it comes to the difference in CAARs between asset management companies and investment banks compared to commercial banks, we see that the difference is significantly positive when considering the MSCI World Banks sample, while for the local indices sample, the difference is not as significant, with a p-value of 0.124. Thus, the market seems to reward M&A activity by non-traditional banking institutions more favorably in a global context, possibly due to perceived synergies or innovation-related expectations. Lastly, as the table below shows, deals that were announced before the outbreak of the COVID-19 pandemic generate higher cumulative average abnormal returns compared to deals announced after its outbreak, with results that are statistically significant at the 1% level. The effect of the pandemic appears to have had a significant impact on investor sentiment surrounding M&A deals, leading to lower CARs post-COVID-19 across all specifications and benchmarks. The difference reaches 1.19% with high statistical confidence in the MSCI World Banks sample, indicating a clear structural shift in market reactions post-2020.

5. Robustness Checks

To ensure the consistency and reliability of the results, a series of robustness checks have been carried out by extending the event window to [−5, 5], allowing for a broader view of market reactions around M&A announcements. These checks aim to verify whether the previously observed patterns hold under different specifications and whether they are sensitive to the choice of market benchmark. OLS regressions using CARs in the [−5, 5] window (

Table 8. Cumulative abnormal returns in the event window [−5, 5] computed using local indices as market portfolios; CARs are computed using the market model.

CAR (−5, 5) Local Indices	1	2	3	4	5
Constant	0.010	−3.119	−3.577	−5.052	−4.038
	[0.0334]	[2.648]	[2.6577]	[4.1009]	[2.7152]
Ln(TA)	−0.003	−0.003	−0.003	−0.003
	[0.0026]	[0.0026]	[0.0026]	[0.0026]
Market-to-Book	0.003	0.003	0.003	0.002
	[0.0021]	[0.0022]	[0.0022]	[0.0022]
Revenue_Growth	−0.001	−0.000	0.000	0.001
	[0.0042]	[0.0043]	[0.0043]	[0.0043]
Profit_Margin	0.006	0.001	−0.003	0.001
	[0.0287]	[0.029]	[0.0291]	[0.0293]
Tax_Rate	−0.007	−0.008	−0.008	−0.007
	[0.012]	[0.0119]	[0.0119]	[0.012]
Relative_Value	0.004	−0.011	−0.035	−0.020
	[0.0733]	[0.0741]	[0.0749]	[0.0746]
Cash_Consideration			−0.015	−0.015
			[0.012]	[0.0122]
Fintech_Experience			−0.016
			[0.0121]
COVID				−0.011
				[0.0178]
Financial_Media					0.002
					[0.0241]
Payments					0.016
					[0.0217]
Investment_Tech					0.011
					[0.022]
Banking_Tech					-
Business_Process					0.003
					[0.0272]
Digital_Lending					0.020
					[0.0416]
Year-Fixed Effects	No	Yes	Yes	Yes	Yes
Number of Observations	90	90	90	90	90
Df	83	67	65	66	68
R-Squared	0.266	0.294	0.348	0.325	0.206

Standard errors are reported in squared brackets. The sample size for the local indices sample is 90 observations. The outputs of the other regressions in the [−5, 5] event window are reported in the Appendix A.

) confirm the earlier findings. The Market-to-Book ratio remains positively related to abnormal returns, though its significance weakens slightly. The Tax Rate variable turns negative, suggesting that tax effects may fade or even reverse as information diffuses beyond the immediate announcement window. Revenue Growth, Profit Margin, and Relative Value remain insignificant, indicating that firm-level fundamentals are not central drivers of market reaction. Introducing additional controls (Cash Consideration, COVID-19, and fintech subsector dummies) does not materially alter results, and the Fintech Experience variable remains negative and insignificant. To reduce repetition, detailed coefficients for both benchmarks are reported in the Appendix A, while discussion here focuses on benchmark contrasts.

The comparison between local indices and the MSCI World Banks index (

Table 9. Differences between the CAARs of separate subsamples; the CAARs were estimated using local indices as market portfolios. CARs are computed using the market model. The table reports the values of the differences, the results of the t-tests, the number of degrees of freedom, and the estimated p-values. CAAR_AM represents the cumulative average abnormal returns of asset management companies and investment banks.

Event Window [−5, 5]
Local Indices	Difference	t-Test	df	p-Value
CAARUS − CAAREU	−1.86% ***	−6.822	28.480	-
CAARAM − CAARBanks	1.66% ***	5.483	22.400	0.000
CAARPre-COVID − CAARPost-COVID	−0.88% ***	−4.151	53.572	0.000
CAARAdvanced − CAAREmerging	0.99% **	2.133	15.830	0.050

Significance at the 95%, and 99% confidence levels is denoted by **, and ***, respectively.

and

Table 10. Differences between the CAARs of separate subsamples; the CAARs have been estimated using the MSCI World Banks index as market portfolio. CARs are computed using the market model. The table reports the values of the differences, the results of the t-tests, the number of degrees of freedom, and the estimated p-values. CAAR_AM represents the cumulative average abnormal returns of asset management companies and investment banks.

Event Window [−5, 5]
MSCI World Banks Index	Difference	t-Test	df	p-Value
CAARUS − CAAREU	−0.14%	−0.438	24.924	0.666
CAARAM − CAARBanks	2.76% ***	9.185	21.752	-
CAARPre-COVID − CAARPost-COVID	0.60% ***	2.746	51.732	0.008
CAARAdvanced − CAAREmerging	3.77% ***	7.899	13.779	-

Significance at the 99% confidence levels is denoted by ***.

) reveals that results vary somewhat depending on the benchmark, underscoring the methodological relevance of this choice. U.S. acquirers underperform European ones by −1.86% under local indices, but this difference disappears (−0.14%) under the MSCI benchmark, reflecting regional biases in local indices versus broader investor sentiment captured by global benchmarks. Institutional-type patterns remain consistent: asset management firms and investment banks show higher CAARs than commercial banks under both benchmarks, with the advantage growing from 1.66% to 2.76% under the global benchmark—consistent with stronger perceived innovation synergies. When comparing pre- and post-COVID-19 periods, the sign of the effect varies by benchmark. Local indices show higher pre-COVID-19 returns (−0.88%), while the MSCI benchmark suggests stronger post-COVID-19 performance (+0.60%), implying divergent adjustments between local and global investor expectations after 2020. Lastly, when contrasting advanced with emerging markets, the results show statistical significance under both benchmarks, though the effect is substantially larger when global returns are used. The CAAR difference rises from 0.99% with local indices to 3.77% with the MSCI benchmark. This suggests that fintech M&A transactions involving acquirers from advanced economies are more positively perceived by the market, particularly in a global investment context.

In sum, the robustness analysis based on the [−5, 5] event window supports the main conclusions of the primary [0, 1] analysis, though it also reveals that some relationships—especially those concerning geographic origin and timing—are sensitive to the benchmark specification. These findings highlight the importance of carefully selecting the market portfolio in event studies, as the choice can materially affect both the magnitude and the interpretation of abnormal returns.

To address potential endogeneity in acquirer choice of fintech targets, the regressions are interpreted as identifying associations consistent with the prior literature rather than strict causality. Firm-level controls and year-fixed effects mitigate omitted-variable bias, while the event-study design minimizes reverse causality since market reactions cannot affect pre-deal fundamentals. Short windows, standardized variables, and subsample comparisons further reduce heteroskedasticity risks, suggesting that these factors are unlikely to bias the results materially.

6. Discussion

The results from the CAR analysis show that fintech M&A causes a downward turn in financial performance for the bidding financial institutions’ stocks. This indicates that the market does not see fintech acquisitions as drivers of either growth or profitability for financial firms, showing that the potential risks broadly associated with these acquisitions outweigh the potential benefits. This supports the agency theory perspective that managers may pursue fintech acquisitions for strategic prestige or overconfidence rather than shareholder value maximization (Jensen & Ruback, 1983). From a market efficiency standpoint (Fama, 1970), prices immediately incorporate these skeptical expectations, leading to negative abnormal returns. These results support the findings of Carlini et al. (2022) that fintech investments indeed generate negative abnormal returns for bidders. Analyzing the variables that drive market performance around M&A announcement date, this study finds that the $Market-to-Book$ variable has a significantly positive coefficient (+0.2% to +0.3%), consistent with information asymmetry and financial constraint theories (Myers & Majluf, 1984), indicating that in its assessment of M&A transactions, the market prefers deals where the buyer has a solid balance sheet, showing stability and reducing uncertainty about future performance which may instead be an issue for less solid financial institutions, especially considering the risks involved in the integration of a fintech firm with a traditional institution. The statistical significance of the $TaxRate$ explanatory variable, which also displays a positive coefficient (+0.12% to +0.16%), supports the findings of Dranev et al. (2019), which argue that the variable can be considered a proxy for the regulation of the financial technology sector, and it is under this assumption that they conclude that the market may be expecting tax rate cuts for companies acquiring fintech firms. Another perspective, although generally considered less relevant, may point to the fact that banks and other highly regulated financial institutions could benefit from acquiring less regulated businesses that can enable them to offer capital-light services. Another meaningful result is the lack of significance of the $\mathrm{P}\mathrm{r}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{x}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}$ dummy variable, in contrast with the results obtained by Carlini et al. (2022).

Innovation diffusion theory would suggest that fintech acquisitions provide access to novel technologies and markets, thereby improving efficiency and revenue potential (Rogers, 2003). However, this study finds no evidence supporting this assumption, as both Revenue Growth and Profit Margin are not statistically significant. In fact, while one would suppose that the market expects financial institutions that struggle to grow in revenues or improve margins to benefit from fintech M&A, both $RevenueGrowth$ and $ProfitMargin$ are not statistically significant, although, as one could initially suspect, their coefficients generally happen to be negative. However, analyzing the sample with the MSCI World Banks as market portfolio, one finding is that the variable $ProfitMargin$ is the only statistically significant one, with its significance being at the 5% level over all the regression models. In that case, the coefficients for $ProfitMargin$ are always negative, supporting the hypotheses that if financial acquirers already have a strong business model with high marginality, the market does not appear to be in favor of fintech M&A, which would introduce new risks and costs. In contrast with the results discovered by Cappa et al. (2022), which found that the acquisition of fintech targets that operate in the payments subsector have a negative effect on banks’ performance, in this study, we find no significant evidence suggesting that there is a meaningful difference in announcement returns that can be explained by the difference in the fintech sector in which the target operates. Lastly, the results of this study suggest that any previous experience in fintech M&A the bidder may have does not affect its market performance around deal announcement, an outcome that is in contrast with the results found by Dranev in their 2019 research, specifically the fact that the first fintech deal would have a more positive effect on stock returns in the short term compared to following fintech deals.

The results of the two-sample paired tests indicate a new meaningful contribution to the fintech M&A field of research. Indeed, this study finds that US bidders display a significantly worse financial performance in the short term compared to their European counterparts. This result points to the fact that, while the financial technology industry is more developed in the US than it is in Europe (in the US, there is the largest number of fintech companies anywhere in the world, according to numbers reported in 2024 by Statista (2024)), investors have better expectations about European financial institutions after their acquisition of fintech firms compared to their US counterparts. This may be consequence of the lower number of fintech players in Europe, meaning lower competition and greater benefits from fintech acquisitions as they can be among the few offering innovative digital services. Another significant contribution of this study to the existing literature is the identification of the different market response associated with fintech M&A of asset management companies and investment banks compared to commercial banks, with the former generating significantly more positive announcement returns compared to the latter, indicating that the market sentiment around fintech M&A involving financial bidders varies based on the specific segment the bidder operates in and requiring future research to take a more granular approach to defining bidding financial firms when studying fintech M&A as results may significantly vary based on this definition. Lastly, the fact that deals that were announced before the outbreak of the COVID-19 pandemic generate higher cumulative average abnormal returns compared to deals announced after its outbreak proves that market expectations around fintech acquisitions by financial institutions have changed dramatically since February 2020. While this result may also be influenced by the fact that macroeconomic conditions have changed significantly since after the pandemic, it still shows that demand for fintech has drastically reduced, with the market preferring to follow different trends such as artificial intelligence. This pattern reflects adaptive market theory (Lo, 2004), suggesting that investor preferences evolve dynamically: post-pandemic, capital and attention have shifted from fintech toward emergent technologies such as AI, reshaping perceived innovation value.

7. Conclusions

Financial technology innovation has significantly impacted the financial services industry, with even legacy players being challenged by the new products and services that fintech has introduced. Many financial institutions have decided to react to these changes by investing directly in fintech firms. However, this study shows that fintech M&A has a negative effect on buyers’ stock market returns in the short-term, destroying value for bidders’ shareholders. While this value destruction generally holds, it can vary based on bidder- and deal-specific factors. Indeed, European financial institutions record higher returns around fintech M&A announcements than US ones. Additionally, asset management companies and broker-dealers tend to gain higher returns compared to banks, and deals that took place before the COVID-19 pandemic report higher results, indicating a negative shift in market sentiment somewhere between the pandemic and the summer of 2024, when this study was performed. In conclusion, while in general fintech M&A undertaken by financial acquirers tends to destroy value for the buyers, at least in the short term, there are some indicators that can explain with a high degree of confidence the variance in cumulative abnormal returns.

This study expands the current literature on financial institutions and fintech M&A by showing that European financial institutions generally outperform their US counterparts around fintech M&A announcement, that asset management companies and investment banks gain better results with fintech M&A compared to commercial banks, and that market expectations around fintech and its impact on the financial industry have changed since COVID-19, with the market showing a more negative reaction to fintech M&A in recent deals. Practitioners can leverage this study to understand what the market sentiment around fintech M&A is and, more specifically, what are the key aspects that affect the market response to fintech M&A.

Unfortunately, the size of the samples used in this study is small, meaning that the results of this investigation may be strongly dependent on the specific samples used and may be difficult to replicate on larger, more comprehensive datasets. One way to limit this issue could be relaxing the criteria for sample selection, for example by removing the requirement for a minimum deal or transaction value. Additionally, while there are some transactions involving emerging markets acquirers, data on these types of deals are scarce, meaning that while this study attempts to analyze abnormal announcement returns of global acquirers, almost 80% of the sample is represented by buyers in advanced economies.

Building on these discoveries, new analyses could be performed to assess whether the better performance of asset management companies and investment banks in the short-term lasts for longer periods, indicating that there are indeed strong strategic fits between these buyers and fintech firms, for example by looking at buy-and-hold abnormal returns (BHAR).