Micro, Small, and Medium Enterprises (MSMEs) are defined as businesses run by individuals or entities that are not branches of small, medium, and large enterprises (Kilay et al., 2022). Legislation number 20 of 2008 concerning Micro, Small, and Medium Enterprises categorizes Micro, Small, and Medium Enterprises as businesses that have a turnover of between Rp. 50,000,000 equals Rp. 50,000,000,000 (Riani et al., 2019). Whether developed, underdeveloped, or developing countries, the contribution of the MSME sector to a country's development cannot be denied because the contribution of MSMEs can be seen in every aspect of development, be it exports, job creation and Gross Domestic Product (GDP) (Sindhwani et al., 2023). Data from the Ministry of Cooperatives and Small and Medium Enterprises shows that the number of MSMEs in Indonesia in 2022 will reach 64.2 million with a contribution of IDR. 8,573.89 trillion (61.07%) of Gross Domestic Product (GDP) which can absorb 97% of the workforce and can collect up to 60.42% of total investment in Indonesia (Mainak, 2022-02-10). This data shows that MSMEs have a very important role in improving the Indonesian economy.

One of the cities in Indonesia with very rapid development of MSMEs is the city of Surakarta. BPS data shows that the number of MSMEs in Central Java continues to increase from year to year and the city with the largest number of MSMEs is the city of Surakarta. Data from the Central Java Province Small and Medium Enterprises Cooperatives Service shows that the number of Central Java MSMEs as of the first quarter of 2023 is 183,579 units with labor absorption reaching 1,1337,156. Meanwhile, Surakarta City is the city with the largest number of MSMEs in Central Java, reaching 17,263 units. This positive trend cannot be separated from the support of the Surakarta City Government for MSME actors (Primasasti, 2023-07-11). The importance of MSMEs to the Indonesian economy means that the government continues to promote programs that can encourage the development of MSMEs. One of the programs is the digitalization of MSMEs. By 2024, the government is targeting 30 million MSMEs to enter the digital ecosystem. This target can be achieved if MSME players are willing to switch from the conventional economy to the digital economy.

FinTech is a new digital technology that makes payments easier (Kou et al., 2021). Apart from providing convenience and comfort, fintech also creates technological pressure or what is called technostress (Lee, 2021). Technostress is stress resulting from prolonged use of technology (Verkijika, 2019). According to Ragu-Nathan et al. (2008), technostress occurs due to 5 conditions, namely techno-complexity (CPX), techno-uncertainty (UCT), techno-invasion (IVS), techno-insecurity (INS) and techno-overload (OVL). In detail, techno-complexity is a condition where technology users feel they do not have sufficient knowledge about ICT because of its complexity so they are forced to spend more time studying it. Techno-uncertainty is a condition where technology users are forced to continuously learn about ICT because the nature of technology continues to change and update, giving rise to uncertainty. Techno-invasion is a condition where ICT users want to continue to be connected to ICT, causing an imbalance between work and personal life. Techno-insecurity is a condition where users feel worried because their jobs will be replaced by other people who understand ICT better. Techno-overload is a condition where users receive excessive information so it is difficult to differentiate between useful and useless information. This research does not include techno-insecurity like research conducted by (Kim & Lee, 2021) because this research does not compare individual abilities, so techno-insecurity is not suitable to be included in this research.

Several studies link technostress intending to use Fintech which shows that technostress has a negative effect on the intention to use FinTech ((Kim & Lee, 2021); (Liébana-Cabanillas et al., 2020)). Research conducted by (Putriani & Apriani, 2022) shows that techno-complexity and techno-overload can reduce intentions to use fintech. Although there has not been much research linking technostress with fintech, several studies have been conducted in the field of technology. Research conducted by (Fuglseth & Sørebø, 2014) shows that technostress can influence employees' intentions to use ICT. Research conducted by (Maier et al., 2015) shows that technostress has a negative effect on the intention to use Facebook. Research conducted by (Joo et al., 2016) shows that technostress can significantly influence teachers' intentions to use technology. Based on conceptual theory and previous research results, hypotheses 1 – 4 are as follows:

H1: Overloaded has a negative effect on MSMEs' intentions to use FinTech.

H2: The invasion of digital technology has had a negative impact on MSMEs' intentions to use FinTech.

H3: The complexity of digital technology has a negative effect on MSMEs' intentions to use FinTech.

H4: The uncertainty of digital technology has a negative effect on MSMEs' intentions to use FinTech.

One important factor that determines a person's intention to use digital devices is self-efficacy (Kim & Lee, 2021). In social psychology, self-efficacy is defined as an individual's belief in their ability to complete tasks (Bandura, 1978). The greater a person's confidence in digital skills, the less anxious a person feels in using information technology (Filho & Rabaai, 2016). Several studies linking self-efficacy and FinTech have been conducted by several researchers. Research conducted by (Putriani & Apriani, 2022) shows that digital technology self-efficacy can increase intentions to use fintech. Research conducted by (Chung & Lin, 2023) shows that self-efficacy has a positive influence on the intention to use fintech. Research conducted by (Jumardi et al., 2020) shows that self-efficacy significantly influences a person's intention to use digital financial transaction services. Research conducted by (Savitha & Hawaldar, 2022) shows that self-efficacy has a positive influence on the intention to use fintech budgeting applications. Apart from being linked to the use of fintech, several studies also link self-efficacy to the use of digital financial transactions. Several studies show that self-efficacy has a positive influence on the intention to use i-banking banking ((Ariff et al., 2012); (Peral-Peral et al., 2020); (Sharif & Raza, 2017)). Based on conceptual theory and previous research results, the fifth hypothesis is:

H5. Digital Technology Sel-Efficacy has a positive effect on MSMEs' intention to use FinTech

Bandura (1994) classifies self-efficacy into four main sources, namely social persuasion, vicarious experience through social models, mastery experience, and reducing stress. Based on the concept of self-efficacy, it shows that self-efficacy can reduce stress. So it can be said that self-efficacy can also reduce technostress on intentions to use fintech. Several previous studies show that self-efficacy can reduce the impact of technostress in using fintech. Research conducted by (Tarafdar et al., 2014) shows that self-efficacy can reduce the negative impact of technostress. Research conducted by (Lee, 2021) shows that self-efficacy can reduce the impact of technostress on the intention to use fintech. Research conducted by (Kim & Lee, 2021) and (Yener et al., 2021) shows that self-efficacy can moderate the relationship between technostress and intention to use fintech. Research conducted by (Putriani & Apriani, 2022) shows that self-efficacy can reduce technostress in using FinTech. Based on conceptual theory and previous research, the sixth to ninth hypotheses are:

H6. Digital Technology Self-Efficacy reduces the negative impact of Overloaded on MSMEs' intentions to use FinTech

H7. Digital Technology Self-Efficacy reduces the negative impact of digital technology invasion on MSMEs' intentions to use FinTech

H8. Digital Technology Self-Efficacy reduces the negative impact of digital technology complexity on MSMEs' intentions to use FinTech

H9. Digital Technology Self-Efficacy reduces the negative impact of digital technology uncertainty on MSMEs' intentions to use FinTech

This research data was obtained from a questionnaire survey. The questionnaire survey used in this research consists of 5 Likert scales, namely 1-5 (1 = strongly disagree; 5 = strongly agree). The focus of this research is on MSMEs (Micro, Small, and Medium Enterprises) in the city of Surakarta. The city of Surakarta was chosen because according to the Indonesian Central Statistics Agency (BPS), Surakarta is the city with the fastest growth of MSMEs in Central Java. Therefore, the setting in Surakarta City is very suitable for use in this research. The respondent criteria used in this research are MSME actors in any field who provide FinTech options for transactions in the city of Surakarta. FinTech provided by MSME players in transactions includes OVO, Q-RIS, E-Banking, E-Wallet, LinkAja, Dana, etc. All instruments used in this research have passed reliability and validity tests so that they can be directly distributed to real respondents using predetermined criteria. The minimum sample for this study is 70 respondents referring to (Cohen, 1992). After collecting a sample of at least 70 respondents, it was then analyzed according to SEM-PLS with the Warp-PLS analysis tool.

The independent variable of this research is digital technostress. According to (Ragu-Nathan et al., 2008), digital technostress is stress caused by adaptation problems in the use of information technology by individuals. This research uses several digital technostress constructs by referring to the research of (Tarafdar et al., 2007) which consists of 5 dimensions including Overload (OVL), Invasion (IVS), Complexity (CPX), Insecurity (INS), and Uncertainty (UCT). Of the five dimensions, this research only uses 4 dimensions, namely Overload (OVL), Invasion (IVS), Complexity (CPX), and Uncertainty (UCT). Meanwhile, the Insecurity dimension (INS) is not used because the Insecurity dimension (INS) is related to relationships between individuals in the organization, whereas this research only uses a small sample of MSME actors, which means there is no relationship between a person and other people in the organization because the MSME only employs no more than 5 people, even just the owner himself. Therefore, Insecurity (INS) is not suitable for use in this research. Furthermore, the dependent variable of this research is FinTech Usage Intention (FUI). The FinTech Usage Intention (FUI) construct used in this research refers to (Le, 2021) which consists of five measurement items. Apart from that, this research uses a moderating variable, namely self-efficacy. According to (Fiedler, 2018), self-efficacy is an individual's belief in being able to organize and carry out the type of performance and tasks specified. So according to (Lee, 2021), Digital Technology Self Efficacy (DTSE) is the psychological belief that someone uses digital technology effectively. Finally, this research uses control variables consisting of Perceived Ease of Use (PEU) and Perceived Benefits (PU). More clear information about variables and construct measurements can be seen in Table 1.

This research consisted of a total final sample of 138 respondents from MSME actors (both employees and owners). Most MSMEs in this research are small, consisting of less than 50 employees, and many MSMEs are even managed by their owners. There were 51 male respondents (37%) less than 87 female respondents (63%). Respondents who were married consisted of 56 people (40.6%) and the remaining 81 people were single (58.7%). In addition, most respondents had a high school or equivalent educational background (82%) and the remainder had a bachelor's degree (38.4%). Demographic information can be seen in Table 2.

The initial stage before hypothesis testing is the validity and reliability test which aims to ensure that the research instrument is valid and reliable. In more detail, the validity test consists of a convergent validity test and a discriminant validity test. The convergent validity test can be seen in Table 3, namely the factor loading values (bold letters). The requirement for convergent validity is to have a factor loading value of at least 0.4 and preferably above 0.7 (Hair et al., 2014). The results show that all values are greater than 0.70, meaning that all measurement items meet the requirements for convergent validity. Apart from that, the convergent validity test can be seen from the AVE (Average Variance Extracted) value provided the value is greater than 0.5 (Hair et al., 2017). Based on Table 5, shows that all AVEs have values greater than 0.5 so they meet the requirements for convergent validity. Next, the discriminant validity test is seen from the AVE value in the diagonal column of Table 4 (bold letters). If the AVE value is greater than the other numbers in the column, then the construct meets the requirements for discriminant validity. Based on Table 4, shows that all AVE values are greater when compared to other figures, meaning that all constructs meet the requirements for discriminant validity. The second test is reliability which can be seen based on the Composite Reliability (CR) and Cronbach's Alpha (CA) values. According to (Hair et al., 2017), the condition for a construct to meet the reliability requirements is if the CR and CA values are greater than 0.70. Based on Table 5, the CR and CA values are greater than 0.70 so they meet the reliability requirements.

Hypothesis testing in this research uses the Structural Equation Model (SEM). The results of the hypothesis test can be seen in Table 6. The results of the hypothesis test show that all independent variables of digital technostress, namely OVL, IVS, CPX, and UCT, do not have a statistically significant effect on FUI (β = 0.060, p > 0.05; β = -0.025, p > 0.05; β = 0.060, p > 0.05; β = -0.071, p > 0.05), so that H1 to H4 are not supported. Furthermore, the moderating variable Digital Technology Self-Efficacy (DTSE) has a statistically significant positive effect on FUI (β = 0.265, p < 0.01), thus supporting H5. In addition, the DTSE variable moderates all dimensions of Digital Technostress. In more detail, DTSE moderates the relationship between OVL, IVS, CPX, and UCT on FUI (β = -0.518, p < 0.01; β = -0.123, p < 0.1; β = -0.368, p < 0.01; β = -0.329, p < 0.01). Apart from that, the control variables PEU and PU have a statistically significant effect on FUI (β = 0.421, p < 0.01; β = 0.220, p < 0.01).