**1. Introduction**

The origin of blockchain tech and cryptocurrencies (cryptos for short) dates back to 2008. That year, Satoshi Nakamoto posted a paper to a cryptography forum entitled "Bitcoin: A Peer-to-Peer Electronic Cash System" [1]. That post described a decentralized peer-to-peer monetary system, whose motivation was that "a purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution" [1]. Thus, in 2009 the first cryptocurrency, so-called bitcoin, was created. Cryptocurrencies are digital currencies based on blockchain technology, which employ cryptographic techniques. They are also non-fiat digital currencies i.e., digital currencies that are not linked to any underlying asset, have no intrinsic value, and do not suppose a liability from any economic agen<sup>t</sup> [2].

Glasser et al. [3] differentiate the following uses of cryptos: as a speculative digital asset and as a currency. They conclude that the use of cryptos is biased to speculation.

**Citation:** Arias-Oliva, M.; de Andrés-Sánchez, J.; Pelegrín-Borondo, J. Fuzzy Set Qualitative Comparative Analysis of Factors Influencing the Use of Cryptocurrencies in Spanish Households. *Mathematics* **2021**, *9*, 324. https://doi.org/10.3390/math9040 324

Academic Editor: Basil Papadopoulos Received: 20 January 2021 Accepted: 3 February 2021 Published: 6 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

So, [4] estimates that above 50% of crypto platforms users have utilized them only in speculative trades. On the other hand, about 46% of cryptocurrency platforms users have employed cryptos as a transactional medium at least once in a year.

Following the statistics by CoinmarketCap [5], the capitalization of crypto markets has grown approximately 100 times from last week of April 2013 (US \$1.37 billions) to last week of 2020 (US \$140 billions). In both dates bitcoin was the most capitalized cryptocurrency but whereas 4th week of April 2013 that value was US \$1.3 billion and concentrated above 94% of whole market, capitalization in last week of 2020 grew to US \$488 billion, but supposed slightly less than 50% of overall market. The magnitude of growth rate is still more impressive for the 10th currency. On April 2013 that currency was Mincoin with a capitalization value US \$118.657. In last week of December 2020 this place was occupied by Polkadlot (US \$4.6 billions). That is to say, the value of Polkadlot at the end of 2020 is 3.5 times overall cryptocurrency market on 2013 April. Likewise, the number of actually negotiated cryptos in market has increased within that period from 10 last week of 2003 April to more than 2000 in last week of 2020.

A simple bibliographic search on Web of Knowledge database shows that from bitcoin creation to middle 2010, papers on this topic were scarce or null. Table 1 shows the results of the simple search "cryptos" AND several terms as e.g., "prices" in Social Sciences publications. Until 2017, research on cryptos grew slowly and likewise at year 2017 it still was scarce [6]. At 2018 there is a breakpoint in the number of published papers that reaches a maximum in 2019.


**Table 1.** Number of papers indexed in Web of Science within 2014–2010 for the search "cryptos AND".

Source: Own elaboration from database Web of Knowledge.

Cryptos generate many opportunities as e.g., fast, efficient, and anonymous transactions and moreover are non-intermediated. However, they also have drawbacks, such as their risk and price volatility, clearly greater than those of a conventional currency; the grea<sup>t</sup> technological and financial knowledge needed for their handle and the fuzzy social perception about holding them. Taking into account these considerations, this paper assesses factors that influence the acceptance of cryptos by households from the framework provided by Technology Acceptance Models (TAMs).

This research has been made with the same sample of adults from Spain as Arias-Oliva et al. [7]. Literature on the application of TAMs on this topic is not so widespread and, due to the reasons above, all very recent. Let us point out apart from [7–22]. Despite all reviewed literature is based in the use of TAMs, the final configuration of hypothesis to test and the use/user of cryptos under consideration have different nuances. They may come from how TAM is applied but also due to the use of cryptocurrency tested: a generic intention to use by individuals [7–20]; as a paymen<sup>t</sup> method in commercial transactions [14–21] or as a purely way of investment [19]. This paper uses the configuration of hypotheses in [7] and also tests the intention to use by households, i.e., their motivation may be either as paymen<sup>t</sup> method or as an investment asset.

With the exception of [19], all reviewed papers on the acceptance of cryptos use Partial Least Squares (PLS) to test the influence of factors. This paper supposes a novelty in this context since uses fuzzy set Qualitative Comparative Analysis (fsQCA) developed by Ragin in [23,24] to extend the results that we reached by using PLS in [7]. This methodology is very used in sociological studies, but also there is a grea<sup>t</sup> deal of applications in managemen<sup>t</sup> and marketing (see [25] for an extended survey). As far as similar questions to ours are concerned, fsQCA has been applied instead PLS in the assessment of new technologies acceptance [26,27] and also as a complementary method to PLS in managemen<sup>t</sup> issues [28,29]. The use of fsQCA provides a complementary approach to correlational methods to deal with causality. Conventional regression is variable-oriented, i.e., it is focused in fitting the mean effect of every variable on the output. On the other hand fsQCA is case-oriented. It is based on measuring the membership degree of each case in the set of attributes and the outcome set [23,28]. Thus, with fsQCA we cannot quantify with a coefficient the influence of a given variable over the output but we can discover how input variables are combined to produce or not produce an output [26].

The rest of the paper is structured as follows. In the second section we built up our hypotheses over the basis of existing literature. Subsequently we present our materials and methods. Fourth section describes results from analytical tools. We finally outline conclusions and future research lines.

#### **2. A Model to Explain the Acceptance of Cryptocurrencies by Households**

The model and hypothesis that we propose to explain the variables influencing crypto acceptance are those we used in [7]. So, our theoretical ground is the Unified Theory of Acceptance and Use of Technology (UTAUT) [30] and its extension UTAUT2 [31]. These models are widely accepted by academic community to explain how an emerging technology is adopted in a society. Both are based on Technology Acceptance Models [32,33], Theory of Reasoned Action [34] and Theory of Planned Behavior [35]. UTAUT models postulate a direct and positive influence of performance expectancy (PE), effort expectation (EE), social norm (SN), and facilitating conditions (FC) on the intention to use (IU) a tech. Likewise, as we do in [7] we include in our model behavioral research findings about how perception of risk and financial literacy affect the IU financial products.

Performance Expectancy (PE) is defined in [30] as the expectation of a person about the influence of using a technology on his/her performance. It is widely accepted that current electronic paymen<sup>t</sup> systems are slow, insecure, inefficient, uncollaborative and nonglobal [36]. So, crypto use has potential to solve these drawbacks [37]. From the emerging of bitcoin, business sphere has integrated progressively cryptos into their activities. The first purchase with bitcoins was done in 2010 to buy two pizzas [38]. Nowadays it is possible to use bitcoins in some 18,500 businesses around the world [39].

Bitcoin is only one of more than 8000 cryptos on the market. That number does not include all cryptos, just those quoted on the market to be traded. The volatility of cryptos opens enormous psychological thresholds in prices [40]. That variety of currencies allows crypto portfolios accessing to wide risk-return configurations. Likewise, as it pointed out by Liu and Tsyvisnsky [41] factors influencing the price of cryptos are different to those linked to the price of conventional financial assets and so, their returns are uncorrelated with those from stock and bonds. For example, cryptos are not influenced by economic cycles. Therefore, they are very suitable to implement so-called alternative portfolio managemen<sup>t</sup> strategies or as shelter investment in recession periods. Moreover, the anonymity provided by the use of cryptos allows public to keep the confidentiality of their savings and movement of funds. The other side of their usefulness is that they make easier criminal acts as e.g., tax evasion, money laundering or contraband transactions [42].

PE is possibly the variable that literature finds as the most relevant in FinTech acceptance. Some evidences in this way are [43] for the use of a paymen<sup>t</sup> authentication system based on biometrics; [44] in the behavioral intention to adopt plastic money; [45] on the use of financial websites; [46] for adopting online banking. In the use of mbanking [26,47–51] obtain similar results. Regarding specific literature on cryptos results in [7–12,14,15,18,20,21] also suggests that PE is the key factor to explain intention to use, so the following hypothesis can be stated:

**Hypothesis 1 (H1).** *Performance expectancy influences positively intention to use cryptos.*

The second variable tested is effort expectancy (EE) that [30] define as the ease extend linked with using a given tech. In this regard [52] states that major part of interviewed people feel that blockchain is not an easy technology to use. In this way [53] outlines that making transaction with bitcoins is a grea<sup>t</sup> challenge for many people. In fact [54] finds that non-users of bitcoin felt incapable of using it and so, it is possibly the greater barrier to the widespread use of cryptos.

There are many evidences of the positive relation EE with the adoption of new financial technologies. Some evidences in this way are [55] for electronic ways of micro crowdfunding and also abovementioned papers [26,43–45,47–50,56]. Within cryptos and blockchain research we find [9–12,15,16,18,20] confirming so. We have to point out that in [20] EE may be assimilated to the construct "web quality" in this paper. In [7] we also found for this relation a positive sign but with a weak statistical significance, so the following hypothesis is proposed:

#### **Hypothesis 2 (H2).** *Effort expectancy is positively linked with the intention of using cryptos.*

Social influence (SI) is the degree to which people feel that close persons think that they have to use a specific technology [30]. At this regard persons from cryptocurrency community participate in collaborative works giving help to the rest [57]. Social support generates trust and commitment and so is linked positively to the intention of using a technology [15]. Despite there are less evidences of the link between SI and IU, there are still a grea<sup>t</sup> number of findings in this way. Within e-banking we can remark [43,48,49,51,55,56]. Regarding cryptos let us outline that [8,10,12,14,17–20,22]. In our paper [7] we also found a positive relation but it was not significant, so the following hypothesis is tested:

#### **Hypothesis 3 (H3).** *Social influence is positively linked with the intention of using cryptos.*

Facilitating conditions (FC) are the degree to which an individual considers that he/she has the necessary infrastructure to run a specific technology [30]. It is clear that operating with cryptos needs being technologically equipped and, likewise, a minimum level of computer comprehension and knowledge is required [58,59]. As far as FindTech is concerned [26,46,47,56] found a positive significant relation. Regarding cryptos and bitcoin [10,16,17,19,21] show that FC influences cryptos IU. In [7] we had also identified FC to be a determinant factor, so the following hypothesis is made:

#### **Hypothesis 4 (H4).** *Facilitating conditions are related positively with the intention of using cryptos.*

Despite not being explicitly considered in UTAUT2, perceived risk (PR) is considered in many papers as a key barrier to using new techs. From a behavioral research perspective, Faqih in [60] defines PR as consumers' belief about the degree of uncertainty and nondesired consequences due to putting to work a product. It has been considered a key variable of consumers' purchase intention [61,62], as well as a predictor of technology adoption [63]. Likewise, standard financial economics state that the risk of an asset is a key variable to make a financial decision. Therefore, market risk, that is, the risk of losses due to the diminution of cryptos price is clearly greater than of conventional currencies [58]. It is well-known that determinants of cryptos return are completely different from those for stocks and bonds [41,64]. Likewise, volatility, deflation and speculative bubbles are more probable in crypto market than in conventional currency market due to cryptos have no supervision from any Central Bank and their intrinsic value is null [65]. These reasons explains why an ING study on bitcoin opinion found that 29% of Europeans had intention of never investing in cryptos since they had the perception that stocks are a less risky investment tool [58]. On the other hand, cryptos are not linked with any country and so, they are not subject to country-risk. Likewise, following [4], national currency-focused fund transfer systems and B2B paymen<sup>t</sup> platforms are more exposed to the risk of exchange

rate than cryptocurrency-focused merchant services. The reason is that the latter often also deal exchanges with cryptos in their paymen<sup>t</sup> activities.

Likewise, cryptos (specially the small ones) are clearly subjected to liquidity and counterparty risk. For example it is often very difficult changing cryptos with local currencies in many countries as e.g., Latin American states [4].

Undoubtedly, a big deal of risk in cryptos use is operational risk that in several papers [14,15,20] is identified as the main determinant of trustiness. Following [4], the greatest risk factor for small exchanges and second one within large exchanges context are security breaks whose consequence may be a permanent loss of funds. These problems may come from possible cyberattacks, the irreversibility of agreements or the impossibility of key recovering [15]. Likewise from large exchanges point of view the lack of regulation is also a source of risk. This question seems to be less relevant in small exchanges. In these agreements a grea<sup>t</sup> drawback comes from the difficulties with maintaining banking relationships. However, for large exchanges this risk seems to be under control [4]. Small trades are more distressed about fraud than large exchanges. The reason could be that they are addressed more often than large exchanges but also that fraud has a greater patrimonial impact due to their constrained budget [4].

Several studies analyze the influence of PR on the IU financial technologies [25,46,48,66]. However, in [66] it is stated that while the direct influence of PR on the IU m-banking is normally small, it supposes a key factor in pre-adoption process. In cryptocurrency context [14,15,17] consider that variable relevant to explain crypto use. Some papers [14,15,18,20,21] reveal that trustiness and perceived security, that are linked with operational risk, are relevant to decide about the use of cryptos. Thus, the following hypothesis is tested:

#### **Hypothesis 5 (H5).** *Perceived risk is related negatively with the intention of using cryptos.*

The last factor tested in our paper is financial literacy (or financial knowledge), FL. Following [67] people's financial literacy consists in the level of their knowledge about financial concepts and in the degree of confidence on their skills to apply that knowledge in real-world situations.

Financial knowledge has been proved to be influent in adopting new financial techs. Whereas [68] conclude that persons with low financial literacy are consistently less likely to trade stocks, the survey by [69] shows that a grea<sup>t</sup> financial knowledge implies a higher propensity to participate in financial markets and investing in shares. In [67] it is pointed out that financial knowledge is associated with more saving planning, active participation in stock markets and rational choices of financial products. On the other hand, lower financial knowledge implies poorer financial decisions as, e.g., more expensive loans. In [70] it is outlined that financial knowledge effectively impacts in financial decisions as those related to credit cards use, mortgage loans, etc. In a cryptocurrency context [13] observes that in Japan the IU cryptos as a paymen<sup>t</sup> method is positively and significantly linked with FL. However, we did not find in [7] that FL to be relevant in the IU. The following hypothesis is put forth:

#### **Hypothesis 6 (H6).** *Financial literacy positively influences the IU cryptos.*

## **3. Materials and Methods**
