Driven by technological developments and a global decline in cash use, many central Banks are exploring the possibility of issuing digital currencies to replenish cash (PricewaterhouseCoopers, 2019). DCEP is a controllable and anonymous payment instrument issued by the people's Bank of China, which is operated by the designated operating agencies and converted to the public. It is based on the generalized account system, supports the loose coupling function of bank accounts, is equivalent to banknotes and coins, and has value characteristics and legal compensation. Like an entity's fiat currency or cash, DCEP owns all the national currency's features (Cooper et al., 2019), which can be used for daily payment or as a store of value. On the one hand, the China's DCEP is a latest highly decentralized virtual currency. It is a P2P technology that uses a process called cryptography, in which common information is converted into passwords or incomprehensible text to ensure transaction safety, enlarge currency supplement, and avoid fraud (Kim, 2016). The China's DCEP will directly describe an instantaneous picture of monetary activity in a major economy or area, and it can also provide more precise China's GDP estimates' data than those currently available (PricewaterhouseCoopers, 2019).

On the other hand, Korea cross-border business transactions can be conducted by using Klaytn coin without the involvement of the central bank's intermediary (Dahlberg, 2019). The transaction's instantaneous features offered by the Korea's Klaytn such as middlemen removal (Ermakova et al., 2016; Walton and Johnston, 2018), shorter transaction time (Baur et al., 2015; Gunawan and Novendra, 2017), and lower cross-region cost (Baur et al., 2015; Ermakova et al., 2016) made Klaytn acceptable and promised to bring benefits to the users. Besides the representativeness of digital currency (e.g., Klaytn), Korea owns hundreds of sustainable alternatives with dissimilar advantages. Irreplaceable and innovative digital currency characteristics, including but not limited to increase transaction speed (Gao et al., 2016; Gunawan and Novendra, 2017; Mendoza-Tello et al., 2018), reduce business deal costs of remittances (Ermakova et al., 2016; Folkinshteyn and Lennon, 2017), namelessness, and privacy (Shehhi et al., 2014; Seetharaman et al., 2017), and the removal of intermediaries (Gao et al., 2016; Walton and Johnston, 2018).

Overall, China's DCEP and Korea's Klaytn is a potential regulated alternative to private digital currencies. Furthermore, compared with Klaytn, DCEP is considered to be a better supplement or substitute for physical cash (Raskin and Yermack, 2016). For Klaytn in a digital form, DCEP represents a more legal payment method and is generally accepted as legal tender supported and regulated by the global central bank (Cooper et al., 2019).

According to the previous literatures, TAM theory is frequently applied in various studies and can only describe 40% of the usage intentions (Mendoza et al., 2017), while the UTAUT model accounts for over 70% of the usage willingness (Schaper and Pervan, 2007). Obviously, it is clear that UTAUT owns more realistic statistical significance and more explanatory power. Almost at the same time, Venkatesh et al. (2012) expanded UTAUT with price value, hedonic motivation, and habit, and finally proposed UTAUT2, which further improved the interpretation ability of UTAUT model.

UTAUT2 model is widely applied to examine and confirm the factors influencing the intention to accept mobile business (Chopdar et al., 2018; Shaw and Sergueeva, 2019), mobile transactions (Farah et al., 2018; Gyu, 2019), and mobile banking (Khan et al., 2017; Lee and Han, 2019). Obviously, the recent studies incline to integrate various theoretical structures to gain a more comprehensive perspective in researching a new technology users' willingness (Lee and Mu, 2017; Sung, 2019). This research combines UTAUT2 with ITM as a comprehensive integrated framework to assess the elements influencing usage intention to digital currencies between China and Korea. Integrating UTAUT2 with ITM revealed the necessity to fill the theoretical gaps in the area of monetary information technology (Tam and Oliveira, 2017).

Initial trust model (ITM) was termed as “the intent to use trust by customers to meet requirements in the absence of experience or reliable, in-depth information (McKnight et al., 2002; Kim and Prabhakar, 2004).” Thus, Kim et al. (2009) built ITM, in which the initial trust of m-banking could be explained by structural assurance, personal propensity to trust, and firm reputation. ITM theory has been studied in many studies, for instance, m-banking (Martins et al., 2014), m-payment platforms (Chandra et al., 2010), m-shopping (Lu and Su, 2009), online banking (Fisher et al., 2008), and m-healthcare (Fisher et al., 2008).

This research applies the trust theory in e-commerce to digital currency's study: ITM is confirmed to be effective after testing in various business environments; however, the importance of every construct will change differently, which requires some extensions. Particularly, a lot of differences exist in the role of initial trust in digital transactions, which is necessarily integrated into a new transaction trust framework supported by digital currency systems, with a new construct such as UTAUT2 (Zarifis et al., 2014).

In other words, under the broader background of digital currency, it is necessary to discuss the positive correlation between users' willingness to adopt digital currency and the level of digital currency reputation, users' propensity to trust, and the degree of government regulation improvement. Thus, it is ensured that these three supplementary parameters are augmented to ITM to create a user trust model that supports digital currency transactions (Zarifis et al., 2014).

The original data of this study were collected through a questionnaire survey. This survey adopts statistical analysis method and traditional questionnaire survey to collect information of different groups. It considers issues such as demographics, online payment preferences, awareness and satisfaction with security measures, beliefs and attitudes about the services provided by online payment business providers during the COVID-19 pandemic.

In the following empirical analysis, Cronbach's α was used to calculate the reliability of the measurement means through IBM SPSS 24.0, and the validity of each construct was tested and evaluated by examining factor structure and internal correlation. To test the study hypothesis, we used IBM AMOS 24.0, which also verified the causal relationship between endogenous core variables through standard coefficients and significance values. The entire sample is used to analyze the integrated model prior to the hypothesis verification test. For hypothesis testing, we applied the model to the group for a more specific analysis. Structural equation modeling (SEM) is becoming central to the social sciences and arguably the most popular analytical tool.

Considering all the above hypotheses, this research designed a questionnaire to confirm each model and related measurement component to match the final goal. Furthermore, by modifying every ambiguous detail in our questionnaire, this research facilitated that all respondents completely recognize all points of the questions through one-on-one interviews with potential users coming from the Jilin University in China or the Yonsei University in Korea. A survey of 800 questionnaires were issued and 746 copies were collected (response rate 93.25%). After 94 responses were discarded for a lack of critical data, 652 samples (87.40%) were eventually used for deterministic analysis (321 potential Chinese DCEP users and 331 Klaytn Korean users). The sample chrematistics are summarized in Table 1.

Every element was rated on a five-point Likert scale. For example, the scale ranged from 1, “strongly disagreement” to 5, “strongly agreement.” In the subsequent empirical analysis, IBM SPSS 24.0 was used to figure out the measurement reliability with Cronbach's α. Besides, construct validity was assessed by testing the element structure and internal correlation of every construct. IBM Amos 24.0 was used to examine all these hypotheses by determining the causal relationship among core structures through significance value and standard coefficient. For hypothesis testing, all data were used to analyze the integrated model before the hypothesis testing. Middle-aged customers and women are the largest group using the mobile payment system, with a total of 518 respondents (54.47%) being female, of whom 600 (92%) are suitable respondents aged below 40 years. Table 1 provides detailed statistical results on the characteristics of respondents.

Convergent and discriminant validity were applied in the measurement model evaluation, and testing was done by IBM Amos 24.0. According to the results in Table 2, convergent validity refers to the degree to which an evaluation method is associated with others that it should be associated with (Jiang et al., 2013). Three-dimensional indicators were used to measure the convergent validity: (i) the standardized factor loadings, which statistically revealed the correlation between some basic factors and each metric, exceeding the recommended value of 0.50 (Gefen et al., 2003), (ii) the Cronbach's α values were more significant than 0.7 for the reliability of integrated construct (Hair et al., 1998); (iii) the values of AVEs were above 0.50, so the underlying variable explained more than half of the indicator variance (Fornell and Larcker, 1981; Henseler et al., 2009; Hair et al., 2011).

One important criterion was used for the evaluation of discriminant validity, i.e., the Fornell-Larcker criterion. Discriminant validity refers to the degree to which items differ between variables (Thong, 2001). Fornell and Larcker (1981) revealed that each AVE's square root should be bigger than each pair of constructs' related correlation coefficient. According to the results of Table 3, each diagonal value (square root of AVE) was bigger than the non-diagonal values (correlations between structures). The cross-loading criterion recommended that the standardized factor loading of each index should be higher than all cross loadings (Fornell and Larcker, 1981).

Judged by the results of the measurement model, this research revealed that the reliability (construct and indicator) and validity (convergent and discriminant) of the constructs are satisfactory. In other words, our measurement methods were proved to have a satisfactory validity.

The measurement and structural models of this study were evaluated by AMOS 24.0. The χ²/d.f. are 1.112 and 1.205, GFIs are 0.928 and 0.921, AGFIs are 0.918 and 0.911, NFIs are 0.943 and 0.937, CFIs are 0.994 and 0.989, IFIs are 0.994 and 0.989, RFIs are 0.937 and 0.932, PGFIs are 0.806 and 0.818, PCFIs are 0.896 and 0.913, PNFIs are 0.851 and 0.865, RMRs are 0.032 and 0.058, and RMSEAs are 0.013 and 0.018. The abovementioned results support for each structure's association.

Structural equation modeling (SEM) tests were carried out for the proposed hypotheses in this research. For the contracted fit index, it exceeds the acceptable fitness minimum, which is a standard value. Each indicator indicated that the fitting results of the examined samples' fitting results confirmed the combined structures were satisfactory.

Next, to examine the organization of this integrated framework, the structure of DCEP's Chinese sample was tested with the Chinese path coefficient, which is revealed in Figure 2. According to the p-value, 1 of the 14 paths (H2; p-value > 0.05) was rejected, and other 13 paths were proven to be positive.

The DCEP Chinese users' willingness to use generally accounts for 70.8% of the variance usage willingness. The influence on Chinese users (Figure 2) shows that antecedent variables of ITM model and UTAUT2 model account for 71.3 and 78.6% of the variance, respectively, which are related to 70.8% explanatory ability of comprehensive structure on the usage intention.

According to the structural results of Figure 3, this research carries out an analysis of the integrated framework of the Klaytn Korea users' sample. Judged by Figure 3, the Klaytn Korean users' pathway coefficient between the basic hypotheses on this research was well evaluated. According to their respective p-values, one path of these 14 paths (H2; p-value > 0.05) failed, and the rest of the paths were statistically positive.

Klaytn Korea users' willingness to use generally accounts for 77.1% of the variance usage intention. The influence of DCEP China users (Figure 3) shows that prerequisites of ITM and UTAUT2 can explain the variance of 75.4 and 78.9%, respectively, which are related to 77.1% explanatory ability of comprehensive structure on the usage intention.

Table 4 reveals the characteristics of the variables, including the combined framework's coefficients and the hypothesis-analyzed results. According to their respective p-values, one of the 14 pathways (H2; p-value > 0.05) was not supported, and the other 13 pathways were significantly below the 0.05 level.

The comprehensive structure was examined by Korean samples (Table 5), showing that the integrated model was supported. According to the results, one of the 14 paths (H2; p-value > 0.05) was not supported, and the rest of the paths were significantly below the 0.05 level. The comprehensive model was analyzed with the Klaytn Korea users' samples (Table 5).

Table 6 shows normalized coefficients of the pathways, the causal path characteristics, and the confirmed hypothesis results. The empirical analysis results of DCEP China users are shown in Table 6, which confirmed the existence of the comprehensive model. Taking DCEP China users' samples as an example, one of the 14 paths (H2; p-value > 0.05) were not supported, and the rest of the 13 paths were significantly below 0.05.

The empirical results of hypothesis regulation (H11–H20) are shown in Table 7. First of all, judged by the p-values of SI and EE on usage intention, the moderating effect of DCEP and Klaytn groups is not significant. Second, the rest of seven p-values show differences in the regulatory effect between DCEP and Klaytn groups. In the DCEP's Chinese users' group, PE (β = 0.232, p < 0.01), SI (β = 0.208, p < 0.01), HM (β = 0.224, p < 0.01), HA (β = 0.216, p < 0.01), and IT (β = 0.240, p < 0.01) positively influence the usage intention at 5% basic level, unlike in the Klaytn's Korean group. In contrast, SI (β = 0.214, p < 0.01), FC (β = 0.308, p < 0.01), and PV (β = 0.256, p < 0.01) positively affected the basic level of usage intention at 1% in Klaytn's Korean user group, unlike in the Chinese travelers' group. The consequence of this research showed the significance of country moderators in strengthening the explanation power to digital currencies' usage intention in the UTAUT2-based integrated model.

With the rising popularity of DCEP and Klaytn, the digital currencies are expected to gain rapid prominence. In order to confirm the promoters and inhibitors of m-payment, as well as users' response to digital currencies, this study developed a new integrated framework that combines UTAUT2 with ITM, as well as perceived technological security and recommendation intention. The results of this research pointed out that the integrated framework showed better explanatory power in describing consumers' willingness to choose digital currencies as a trustful technology.

On the one hand, from the perspective of ITM model, personal propensity to trust (Gefen et al., 2003) is considered to be the user personality that affects initial trust (Gefen et al., 2003) and is related to user satisfaction, which means to increase the personal trust, tendency can positively influence the initial trust, just like firm reputation and structural assurances (McKnight et al., 2004; Kim et al., 2009). In other words, firm reputation and structural assurances, which refer to company image and security system scale are also the trust's affected indicators, and both have a positive impact on initial trust (Flavian et al., 2005; Chen and Barnes, 2007; Fuller et al., 2007). This research examined personal trust tendency, structural assurances, and firm reputation because all these variables were used by Kim et al. (2009) in their ITM model. According to Gefen et al. (2003), users will obtain positive results in the future through initial trust because initial trust can rise the performance expectancy, increase user satisfaction, and positively influence on buying willingness (Pavlou and David, 2004; Pavlou and Fygenson, 2006; Chang and Chen, 2008).

On the other hand, from the perspective of UTAUT2 model, the findings of both DCEP and Klaytn data revealed that effort expectancy (H2) was not a significant indicator of the willingness of adopting digital currency. However, the results confirmed the significance of PE (H1), SI (H3), FC (H4), HM (H5), PV (H6), and HA (H7) on the willingness of adopting digital currency.

In general, in the context of the exponential growth of digital currency, a comprehensive and comparative perspective is beneficial to other studies, by increasing the possibility of influencing factors to the global digital currencies. Due to the extensive cooperation between China and South Korea in the digital currency area from early 2019, by comparing the differences between users of two countries, the core factors influencing the willingness of consumers to use were stimulated, so as to provide the necessary theoretical basis and practical preparation for the sustainable mobile payment market cooperation between China and South Korea on a larger scale.

Generally, the theoretical contributions of this research are three-fold and are illustrated as follows. First, few scholars have focused on the usage intention of potential users to choose either DCEP or Klaytn. A research blank exists in the digital currency's technology that was not examined at all. This comparison method improved the efficiency of some specific scenarios examining between China and Korea, which also fills the above blank in the digital currency's research. Furthermore, this research also tested the moderating variables and multigroup analysis of the DCEP and Klaytn dissimilarities by improving the integrated multi-model method. The blank of digital currencies' research was increased by testing the adjustment variables under the Chinese and Korean different regulators.

Second, this study proposed a more comprehensive integrated framework that was combined with ITM (Kim and Prabhakar, 2004) and UTAUT2 (Venkatesh et al., 2012) to confirm influencing factors that facilitated the initial trust, while examining the technology of the user terminal characteristics and the initial trust's influence on usage intention. According to the results of this research, it is concluded that the combined model provides a stronger interpretation on adoption willingness than UTAUT2 and ITM, separately, which is consistent with previous studies (Lin et al., 2019, 2022; Lin and Wu, 2021; Lin X. C. et al., 2021; Lin X. et al., 2021). Compared with the single-model analysis, this two-dimensional combined framework can comprehensively confirm the factors that positively influence the usage intention of digital currency's technology.

Third, this research contributed to extend UTAUT2 with ITM, and also explain the moderating power of user satisfaction. As an important crossing-point, the digital currency user's satisfaction can be the essential precondition factor. Inspired by this research's attempt to extend UTAUT2 in a multi-model and multi-group integration viewpoint, studies of future digital currency may consider a more systematical perspective to examine the usage intention of any digital currency.

From a managerial perspective, the managerial implications of this research are proposed as follows. First, this study tested the effect of initial trust and digital technology variables on usage adoption of digital currency users. This evaluation process revealed the effect of several comparison scenarios between DCEP Chinese users and Klaytn Korean users to point out the specific new blanks in the research of digital currency. The final goal is to examine the existing UTAUT2 model with different digital currencies rather than the DCEP and Klaytn of this research to improve the Asia and global monetary markets' development. Thus, the results of this research have the managerial implication to both the academics and the digital currency's providers. The practitioners can plan, improve, and introduce the integrated framework in digital currency's perspective to influence the usage willingness of the consumers.

Second, the findings of this study suggested that the digital currencies' providers should generate the initial trust among users to facilitate their usage intention. If the digital currency turns into real currency, it will consider two major problems, namely, (i) the government policies, the related laws, and regulations to digital currency are important elements positively influencing the future usage intention of digital currency. Selecting the accessible, revolution, and guideline or traditional, limited, and forbidden will strongly influence the acceptance and spread of the digital currency in the global countries. (ii) Even in a moderate market environment, because of the security and convenience of the digital currency business with technical threshold, it is still rather high for some ordinary users. Therefore, the widespread adoption of digital monetary market still requires technological and application innovation, as well as the follow-up of relevant laws and regulations.

Although our research owns a few contributions in theory and management, it also has other limitations and deserves further research; the future research direction remains to be explored. Understanding the relationship between DCEP and other digital currencies in the global different countries is another interesting area to explore. Due to the different maturity and usage of digital currencies, research across multiple countries can offer additional perceptiveness (Dennehy and Sammon, 2015). In addition, security is the last important direction of future study. As so complex production, digital currency transcends the boundary of digital technology, including the dependability of business parties, where dataset is safely settled (device or the digital cloud), proprietorship of the business dataset, and regulatory and administration culture background, including privacy lawmaking. It is worthy of discovering these issues jointly or separately.