What is the inherent value of the blockchain

Financial applications of blockchain technology

Summary

Blockchain technology was first implemented in 2009 as the technological basis of the cryptocurrency Bitcoin. It is said to have the potential to be a disruptive technology that can lead to lasting changes in many areas of economic life. In this post we give an overview of the technology itself as well as its financial applications. In doing so, we focus in particular on cryptocurrencies, the potential of so-called smart contracts, initial coin offerings, the processing of securities transactions and possible effects on the corporate governance of listed companies.

Abstract

The blockchain technology was first implemented in 2009 as the basis of the cryptocurrency Bitcoin. The technology is said to be a disruptive technology that has the potential to significantly affect many areas of the economy. In this paper we provide a survey of the blockchain technology and its applications in finance. We focus on cryptocurrencies, smart contracts, initial coin offerings, the clearing and settlement of transactions in financial markets, and implications for the governance of exchange-listed firms.

introduction

Large parts of the financial industry are subject to profound change. In addition to regulatory changes as a result of the financial crisis, digitalization in particular contributes to this, changing existing business processes and enabling new competitors to enter the market. Banking transactions are often no longer carried out at the counter, but rather on a smartphone, and investment advice is often provided by computer algorithms. Securities trading is largely electronic and is dominated by high-frequency traders, i. H. of computer programs that make and implement trading decisions in fractions of a second. Crowdfunding platforms appear alongside traditional forms of corporate finance. One aspect of digitization that is potentially important for the financial industry is the development of blockchain technology a good decade ago. In an OECD publication (OECD 2018), various new technologies (such as Big Data, the Internet of Things, Artificial Intelligence and others) are examined to determine to what extent they can perform various functions and services in the financial sector (such as payment transactions, financial investments , Lending and others). The only new technology that has an impact all Functions and services is the blockchain technology that we will deal with in this article.

The blockchain technology is based on the cryptocurrency Bitcoin, the concept of which was presented in 2008 in a pseudonymous contribution (Nakamoto 2008).Footnote 1 The basic idea is to design a decentralized payment system that can do without “trustworthy third parties”, i.e. financial institutions such as central and commercial banks. Since then, cryptocurrencies have not ousted traditional currencies, but they have gained significantly in importance. The market capitalization of all cryptocurrencies is several hundred billion US dollars and there are futures and exchange traded funds on cryptocurrencies. So-called smart contracts allow, among other things, the issue of “tokens”, which are used, for example, in the context of Initial Coin Offerings (ICOs), a new type of corporate financing.

However, the application potential of blockchain technology extends far beyond the design of cryptocurrencies. For example, it is discussed to what extent it could revolutionize the processing of securities transactions. Far-reaching effects on the corporate governance of listed companies were also forecast. In addition, numerous applications in non-financial areas have been discussed, which we will not deal with in this article.

However, there are also voices who see the effects of blockchain technology as less far-reaching. The idea of ​​Bitcoin consisted in doing without financial institutions such as central or commercial banks (Nakamoto 2008). In fact, however, the majority of Bitcoin transactions today take place on exchanges (and thus at financial institutions) where Bitcoin and other cryptocurrencies can be traded with one another and against traditional currencies (see e.g. Brauneis et al 2019), which is obviously the case The basic idea of ​​Nakamoto (2008) contradicts this. Pirrong (2019) goes one step further when he writes (p. 98) "... that the initial soaring hopes have been disappointed because the fascination with a shiny new technology blinded too many to underlying economic realities ...".

In this article we attempt a critical inventory of possible financial applications of blockchain technology.Footnote 2 It is structured as follows. In the second section, we first describe the technological fundamentals of the blockchain, using the Bitcoin blockchain as a starting point, but discussing alternative design options at a suitable point. In the third section, we then present what we consider to be the most important financial applications of the new technology. In particular, we go into cryptocurrencies, smart contracts, initial coin offerings, securities trading and the corporate governance of listed companies. The fourth section closes the article with an outlook.

technology

Bitcoin and other cryptocurrencies are based on blockchain technology - a technology that, as indicated above, is ascribed revolutionary potential in terms of its possible impact on business models worldwide. In a nutshell, the blockchain is a distributed open source database based on state-of-the-art cryptography. Blockchain technology aims to facilitate transactions between untrustworthy actors and eliminate the need for a trusted intermediary. If that were the case, financial transactions would be possible without banks, we would no longer need brokers and rating agencies and other credit agencies would also be superfluous. To understand to what extent the technology can meet these expectations, we explain the basic principles of the technology below.Footnote 3 There are two main structural elements: first, a method of organizing and storing data, and second, a method of building trust in the data.

Organization and storage of data

A blockchain is a kind of cash book that enables information to be recorded and tracked, be it information about financial transactions such as in cryptocurrency applications or information about other things of value. Understanding the key features of this cash book is necessary to understand the potential of blockchain technology. They will therefore be examined in more detail below. First, the blockchain organizes information in the form of an ever-growing list of time-stamped records to which we can add data, but cannot change or delete previous data in it. This does not mean that the system cannot be updated. However, instead of overwriting an existing entry, the change itself is saved, resulting in a new separate time-stamped entry. The data structure therefore always contains the complete history. At first glance, this type of organization and storage of data appears cumbersome and memory-intensive, but, as we will see, it promotes transparency and fraud protection. Technically, a cash book that only allows data to be added but forbids data to be changed or deleted is called an “append-only ledger”. For reasons of efficiency, the data are grouped into blocks. A block contains several entries, e.g. B. multiple financial transactions, and adding a new block adds new information to the system. Each block contains a reference to a previous block (its “identification number”), which leads to a sequential order of the blocks. So we can think of the data structure as a chain of blocks - a metaphor that gave the blockchain its name (see Fig. 1).

Second, a blockchain is a decentralized system. In contrast to a classic cash book, which stores data on a single system and which is maintained by a central entity, the information in a blockchain is distributed over a large number of network participants, so-called nodes. These nodes - think of them as computers - do not trust each other completely. Nevertheless, a clever mechanism enables the cash book to be replicated via these nodes. As a result, there are identical copies of the entire transaction history on a large number of computers worldwide. It seems intuitive that this decentralized type of data organization helps prevent attacks and build trust in the data. The process of organizing data and updating a blockchain can be summarized with the following four steps:

  1. 1.

    A new transaction to be added to the blockchain is entered and then sent to every node on the network. It is now distributed in the network, but not yet part of the blockchain.

  2. 2.

    Nodes bundle these new transactions in a block and a node determined by the so-called consensus mechanism (see below) is given the opportunity to transfer its block.

  3. 3.

    Other nodes evaluate the validity of all transactions in this block.

  4. 4.

    They accept validated blocks by referring to them in their next block.

The result of this process is an ordered set of blocks in which all nodes agree on the order and keep copies of the blockchain.

Promote trust in the data

Of course, the data structure described above is only of value if it is trusted. The exciting question is how it is possible to foster trust between unreliable parties in a decentralized system. In other words, how can we ensure that all parties can trust every single block in the chain? Blockchain technology creates such trust through two essential technical components, through asymmetric cryptography and the so-called consensus mechanism.

Asymmetric cryptography

Asymmetric cryptography is initially used to design new transactions that are to be included in the blockchain (see step 1) so that their legitimacy can be checked. Usually a key pair consisting of a private and a public key is used here, which has a one-way functionFootnote 4 is related to each other. To do this, the sender encrypts their transaction with their private key - one also speaks of signing - and any network participant can decrypt the transaction with the sender's public key and thus verify that the transaction actually originated from the sender. The ultimate goal is to ensure the authenticity and integrity of the transaction, i.e. to make the transaction assignable to its sender at any time and at the same time to ensure that the transaction has not been manipulated unnoticed. This allows honest nodes to identify legitimate transactions and only propagate these on the network. Such cryptographic methods were not specifically developed in the context of blockchain technology. Rather, they have been used for some time in e ‑ mail traffic, electronic banking or when shopping on the Internet.

Consensus mechanism

How do you agree on the transactions carried out in a decentralized system? In other words, how do new transaction blocks get into the blockchain (see step 2)? After all, adding blocks is the only feasible operation and there is no way of changing or even deleting existing information. Therefore, there is basically an incentive to add transactions that favor the originator of the blockchain entry. This is where the main innovation of blockchain technology comes into play, the consensus mechanism. This mechanism regulates which node is allowed to transfer its block to the blockchain next. This step is central, because on the one hand there should be no central authority that controls the addition of new information. On the other hand, if any node can basically attach new information, one must fear that dishonest nodes will try to enrich themselves unjustifiably in the process. The consensus mechanism tries to deal with precisely this trade-off. To do this, you first give as many or even all nodes as possible the general possibility of adding a new block. So that a sufficient number of nodes take part in this step, the nodes that collect new transactions are checked and bundled with the aim of being able to attach the next block to the existing chain (the so-called miners), a reward is promised. Miners are therefore in competition with each other and it is important to make this competition meaningful. Thus, the nodes should have an incentive to only create blocks with legitimate transactions that do not conflict with themselves or with all transactions already contained in the blockchain. In particular, no transactions should be part of the block candidate that spend one and the same asset multiple times (so-called double spend problem).

The most widespread consensus mechanism is the proof-of-work mechanism on which the Bitcoin network is based. The core idea is to make the creation of a block more expensive, but to make the verifiability of the correctness of a new block candidate very cheap for everyone else. This makes "false reports" expensive and easy to unmask.

So-called hash functions are used for this purpose. A hash function maps a large amount of data (e.g. the entire content of a block) to a small amount of data (the hash value) of a fixed size (256 bits in the case of the Bitcoin blockchain). The hash function is designed in such a way that small changes to the input data (e.g. exchanging a letter or a number) lead to a completely different output. The Bitcoin blockchain now specifies a certain maximum value for the hash value, defined by the number of leading zeros. In order to achieve this target value, a number (so-called nonce) is added to a block. This number is now changed by trial and error until the hash value meets the requirements. The number of attempts per unit of time to find a valid solution is called the hash rate. The first node to find such a valid solution is allowed to add the block to the blockchain and receives a rewardFootnote 5 (see step 2). Since the chances of success of mining increase with the computing power used, miners use high-performance computers, which leads to very high energy consumption.Footnote 6 If a miner has successfully added a new block to the blockchain, all other nodes can very easily verify that the task has indeed been solved. All you have to do is enter the block including the nonce as input into the hash function and check whether the resulting hash value meets the requirements. The validity of the transactions contained in a block relative to the history can also be checked. At this point, double spends, i.e. multiple expenditures of the same asset, would be recognized immediately (see step 3). Invalid blocks are simply ignored by honest nodes, whereas a new valid block candidate becomes part of the blockchain in that the miners reference this block when they next create a block (see step 4).

The link that makes a chain (the blockchain) from the individual blocks is achieved in that each block contains the hash value of the previous block. If a manipulative change were made in any block in the chain, this would change the hash value of this block and thus also the hash values ​​of all subsequent blocks, so that such manipulation could easily be discovered. The hash value of a block can be interpreted as the block's fingerprint, so to speak.

It can happen that two correct new blocks are attached to the blockchain at the same time, so that two versions of the blockchain are in circulation at the same time. In this case, it must be regulated which version corresponds to the current status. The rule of consensus is that this is the chain with the highest work performance. As a rule, this is the longest chain in each case. Therefore, the more blocks that follow a particular block, the more certain it is that the collective will consider this block to be part of the blockchain.Footnote 7

Ultimately, this proof-of-work mechanism artificially complicates the creation and addition of new blocks by first having to find a suitable nonce. The difficulty of this probabilistic trial-and-error task is chosen in the Bitcoin network in such a way that a new valid block is created on average every ten minutes.The work done by the miner is eponymous for the proof-of-work mechanism.

As a result, this proof-of-work mechanism regulates that the collective of network participants agrees on the current state of the blockchain and trusts the data it contains. The trust does not arise here through a "trustworthy" intermediary. Rather, the costs for a successful attack on the data - the immense computing power that would be required for this - are set prohibitively high and at the same time incentives for honest behavior are set. A functioning blockchain technology replaces the "trustworthy" intermediary, so to speak, with technological trust.

Depending on the design (see Section 2.3) of the blockchain, other consensus mechanisms may be more suitable. An alternative to the resource-intensive proof-of-work mechanism is, for example, the "proof-of-stake" concept, in which the probability of being able to attach the block candidates you have created depends on the share of assets that you hold on the blockchain or the valuables recorded via it . The main idea is to give those actors greater responsibility and thus control who have a high interest in the success of the respective blockchain. In industrial applications, which as a rule are not primarily designed to establish consensus between unreliable parties, further mechanisms that are based on a certain number of trustworthy nodes (proof-of-authority) come into question.

Design types

The blockchain philosophy presented so far is based on a completely decentralized data structure without any access restrictions with potentially dishonest participants. In fact, a blockchain can be designed in very different ways.Footnote 8 In this way, it is possible to regulate who has access to the information stored in the blockchain. Write permissions can also be regulated in the same way. With a public blockchain like the Bitcoin blockchain, anyone can basically join the network and access the stored information. A public blockchain is therefore highly transparent. With the Bitcoin blockchain, anyone can basically work as a miner and thus make changes in the blockchain - one therefore speaks of a "public unpermissioned" blockchain. In contrast, in industrial applications such as Hyperledger or R3 corda, the group of participants is often limited to certain authorized members. One then speaks of a "private permissioned" blockchain.

Another important characterization concerns the degree of anonymity, i.e. ultimately the question of the extent to which personal data is required for the use of the blockchain. Some cryptocurrencies, such as Monero, operate without any user data (anonymous), other applications require personal data, such as the email address, to be provided directly in order to be able to participate in the blockchain (identified). However, it is standard in the cryptocurrency sector to work under pseudonyms.

Public blockchain technologies based on the proof-of-work concept, such as the Bitcoin blockchain, are characterized by low scalability and high resource consumption per transaction. In addition, various blockchain technologies (such as Bitcoin and Ethereum) are initially incompatible with one another, so that direct transactions between them cannot be made. New technological developments aim to overcome existing restrictions and to take into account the growing importance of interoperability.Footnote 9 There are concepts such as “sidechains” (blockchain running parallel to the primary blockchain), “state channels” (transactions outside the primary blockchain) or the idea of ​​“sharding” (transaction is not processed by all nodes). Compare also Voshmgir (2019).

application areas

The main innovation of blockchain technology is to create a consensus between agents who do not trust each other completely - and that without a central trustworthy intermediary such as a bank, an exchange or a broker. A simple application example is a used car purchase: Many people prefer to do this through a dealer and not buy the car directly from the previous owner. One reason for this is the lack of trust in the previous owner. If all "transactions" relating to the car, such as accidents and repairs, were stored in an immutable database such as a blockchain, trust in the previous owner would no longer be necessary. In place of trust in the car dealer, an intermediary that may be difficult to verify, trust in blockchain technology would move.

The comparative advantages of the new technology can therefore be realized particularly in an environment in which agents who do not completely trust each other conclude transactions with one another. Five such areas of application are presented below.

Cryptocurrencies

According to the classical theory of Menger (1871), goods in which many future potential trading partners are interested are used in barter transactions, although they are not needed for immediate use. The only driver of value for such goods, which are also referred to as “money”, can therefore only be the belief that they can be used again for exchange at a later point in time or with another trading partner (so-called liquidity value). Kocherlakota (1998) sees the main purpose of money in recording the mutual claims of the agents among themselves. In small groups - think of exchanging favors with friends, for example - no formal system is necessary for this. Such informal systems no longer work in large societies. Money acts here as a kind of memory that stores mutual obligations.Footnote 10 Central to such a memory database is, on the one hand, that the ownership structure is clearly defined at all times. Another prerequisite for a functioning system is that each monetary unit can only be spent once (double-spend problem).

Berentsen and Schär (2017) define currencies or more general payment systems using three dimensions. These concern the rules according to which money is created, represented and transferred (see Fig. 2). Money creation can either be monopolized, e.g. B. by a central bank or competitively such as mining for gold. The representation can be virtual as with deposit money in the bank account or physical as with cash. Finally, the transfer between different participants can take place centrally, for example with a transfer via a bank or a connected Gironetz, or decentrally, such as when exchanging cash.

The virtual representation has considerable efficiency advantages over the physical one. At the same time, the dependence on a central instance is viewed negatively for various reasons (lack of trust, securing the system against failure). The combination of virtual representation and decentralized transfer would therefore be optimal, but was not feasible before the development of blockchain technology. However, cryptocurrencies such as Bitcoin or Ethereum combine exactly these properties, which are desirable from an economic point of view. The question is therefore why they have not yet been able to establish themselves on a broad basis.Footnote 11

To answer this question, let us consider the three central functions of monetary units (often referred to in the literature as the "triad of money"):

  1. 1.

    Exchange or means of payment function

  2. 2.

    Store of value function

  3. 3.

    Arithmetic unit function

So that a monetary unit can optimally fulfill the function of exchange or means of payment, there should be as few different currencies as possible in an economy. Formally, every agent will accept a given currency if and only if it assumes that a future exchange partner will accept the currency with a high degree of probability. A new currency is generally not accepted by anyone at first. Due to the dependence of one's own acceptance on acceptance by other market participants, it is therefore very difficult for a new currency to leave this initial equilibrium. This is made even more difficult by the fact that classic currencies such as the euro and US dollar are legal tender in their respective countries, i. H. Seller is obliged to accept it. Further reasons that partially impair the function of cryptocurrencies as a means of payment are firstly the relatively high complexity in handling and secondly the sometimes relatively high transaction costs.Footnote 12 Compared to cash and credit cards, the time typically required for the final (irreversible) confirmation of a transaction also has a negative effect on the acceptance of cryptocurrencies.

The currently (still) very high fluctuations in value of most crypto currencies oppose their function as a store of value and hinder their function as a computing unit. Although the path of money creation is specified in the protocol for most cryptocurrenciesFootnote 13 and therefore there cannot be an inflationary increase in the money supply, its use as an object of speculation, the uncertain future adaptation and the concentration of large parts of the money supply with a few actorsFootnote 14 excessive fluctuations in value. For example, the volatility of the Bitcoin-USD exchange rate has decreased somewhat in recent years, but was still 70.7% p.p. in 2019. a. For comparison, the volatility of the EUR-USD exchange rate was only 5.0% p. a. Gold, which in some respects bears a certain similarity to cryptocurrencies, had a volatility of 11.5% p. a.Footnote 15

In response to the high fluctuations in value, so-called stablecoins were developed. The underlying idea is that the cryptocurrency is backed by a fixed amount of money. An example of this is the cryptocurrency Tether, where each Tether unit is secured by USD 1. Libra, the crypto currency planned by Facebook, is also to be covered by a currency basket.Footnote 16 One problem with stablecoins is that, due to their design, they are dependent on a central authority, namely the company or consortium that carries out the collateralisation. On the one hand, this contradicts the original idea of ​​cryptocurrencies to ensure independence from central intermediaries. On the other hand, it may lead to a systemic relevance of the relevant actors and to the risk of monopolies, which is exacerbated by the large amounts of data arising in connection with the use of a cryptocurrency.

Smart contracts

The term smart contracts is used for contracts that are automatically executed subject to certain conditions. In the blockchain context, the term refers to a program code stored in the blockchain, which is automatically executed depending on predefined conditions, but independent of the original creator of the code.Footnote 17 Smart contracts expand the ability of the blockchain to store information in order to carry out calculations. Due to the technical design of the blockchain, the program code can no longer be changed once it has been stored. It therefore provides the contracting parties with the assurance that the agreed event will actually occur if the conditions are met. If the event is a payment or the transfer of an asset, this is usually also processed directly via the blockchain.

An example of an application is extreme event insurance. If such a contract is implemented in the conventional way with an insurance company as the central intermediary, the processing of claims often takes a relatively long time. Due to the complex payout conditions and contract texts that are difficult to understand, the policyholder has a certain degree of uncertainty about the actual payout. In contrast, processing via a smart contract leads to transparent conditions that are linked to measurable events such as B. wind speeds or earthquake strengths can be linked and lead to an immediate payment if fulfilled. The efficiency in processing is also made possible by the fact that the consensus mechanism inherent in the blockchain can very quickly reach a consensus on the “true” state of the world. Insurance protection can be provided by an insurance company or directly by the community of insured persons and, if necessary, other investors. The implementation as a smart contract thus also offers the possibility of distributing corresponding risks even more efficiently by breaking down entry barriers.Footnote 18

One of the central challenges when creating smart contracts in a blockchain is the integration of external information such as weather data or security prices. Interfaces that collect and verify such information and make it usable for the blockchain are called oracles. Since they are dependent on external conditions and therefore cannot be secured by the cryptographic mechanisms of the blockchain, in many cases they represent a central point of attack and thus a certain security risk.

Simple smart contracts such as linking a payment to complex conditions or the delivery of a crypto asset can also be mapped in the Bitcoin blockchain. However, since the Bitcoin scripting language does not allow loops, many types of calculations cannot be carried out with it - in formal terms: the language is not turing-complete. If more complex smart contracts are to be executed on a public blockchain, Ethereum, for example, can be used. The associated programming language allows the programming of so-called distributed apps, is turing-complete and offers tools for standardizing contracts such as issuing your own crypto tokens.

In order to be able to decide which type of contracts can be sensibly mapped using smart contracts, we discuss the strengths and weaknesses of this type of contract. Smart contracts are reliable and completely transparent. Since the code is machine-readable, unlike the usual forms of contract, they leave no room for interpretation. They offer a high level of security, as contractual conditions cannot be changed afterwards and are usually cryptographically encrypted. By eliminating bureaucratic structures and being independent from third parties, they save time and money in many applications. It should be noted that all contract information must be available in the blockchain or integrated via oracles. The immutability of the code makes subsequent upgrades impossible, even if all parties agree. As a result, errors in the code cannot easily be resolved, i. H. in most cases just fix it with a hard fork.Footnote 19 This problem is made even more difficult by the fact that testing smart contracts is sometimes difficult because there may be interaction with other smart contracts or external services. Interestingly, a study by Nikolić et al (2018) has shown that around 3% of almost a million analyzed smart contracts on the Ethereum blockchain have vulnerabilities.Footnote 20

The processing of a transaction via a smart contract is closely related to the completeness of the underlying contract. Firstly, the implementation via a smart contract requires that all possible future states in which an action is required are recorded in advance and taken into account in the program code. This also applies if the respective probability of occurrence is very low. It is therefore not easily possible to deliberately leave the procedure open for certain situations and only to reach an agreement with the contractual partner if necessary. However, such incomplete contracts are common and optimal in many situations, especially when the number of future states is high and the formulation of a complete contract would be very expensive (cf. e.g. Hart and Moore 1999 and Segal 1999). A typical example of an incomplete contract that cannot be sensibly mapped using a smart contract is an employment contract, as neither the work content nor the working conditions can be precisely defined ex ante (cf. Cartier 1994).

Second, the processing of transactions via a shared blockchain means that future states, which would otherwise not be verifiable for certain actors, can be traced via the decentralized consensus mechanism and can be used directly as conditions in smart contracts.For example, the sales success of an end product can also be traced back to the upstream intermediary. This increases the possibilities of concluding full contracts. The broad information base reduces entry barriers and promotes competition. However, the increased availability of information, which must be accessible to every participant in order to achieve the decentralized consensus, can lead to new possibilities for collusion between competitors. Cong and He (2019) analyze this trade-off and find that the resulting overall economic effect - also depending on regulatory conditions - can be both positive and negative.

Initial Coin Offerings

In an Initial Coin Offering (ICO), a company, in most cases an internet startup, sells tokens, i. H. Tokens. These can either certify payment claims against the company, which may be linked to the company's profit or sales (security tokens) or entitle the holder to use the company's future services (utility tokens). In many cases, utility tokens also focus on the hope of later reselling at a profit. ICOs are therefore typically a mixture of securities issuance and the advance sale of a product (see Fig. 3). This makes them very difficult to evaluate. Existing laws and regulations in the areas of investor and consumer protection do not fit the new concept and the unclear responsibilities make it difficult to develop new regulations.

An ICO is used to raise capital (for the first time) by a (startup) company based on blockchain technology. Often there is no product or service at the time of the ICO, but only a so-called "white paper" that describes the basic product idea and its design. The newly issued tokens are sold without exchanges or other intermediaries, with payment mostly being made in crypto currencies. An ICO is typically processed via a blockchain with the help of smart contracts. However, it is important that the investors do not enter into a direct legal relationship with the company by purchasing the tokens.Footnote 21

The volume of new coins issued by companies rose to USD 21.6 billion in 2018, more than tripling compared to 2017 and an increase by more than a factor of 80 compared to 2016. In 2019 it was only 3.3 Billion USD then led to a sharp drop in volumes.Footnote 22 ICOs compete with classic venture capital financing and crowdfunding or complement them in the form of hybrid models (see strategy & et al 2018). Despite the at times very high volumes, ICOs are primarily associated with scandals in the public perception - and there have actually been a large number of cases in which investors have lost all of their investment as a result of ICOs. ICOs that are advertised by celebrities have fallen into disrepute, so the US regulatory authority SEC explicitly warns against this type of ICO. Since ICOs are often structured over several rounds and the price of the tokens increases over time, they always involve the risk of Ponzi schemes. Some regulators are reacting to the scandals with a complete ban on ICOs (e.g. China). Most western countries, on the other hand, try to protect investors with individual checks and sometimes relatively strict regulations without completely banning ICOs. In response to the scandals, among other things, initial exchange offerings (IEOs) have now established themselves as a less decentralized development alongside classic ICOs. A crypto exchange takes on the role of an intermediary, conducts an initial screening of the company and then handles the sale of the tokens.

To better understand how ICOs should be regulated and in what situations they can be used, let's look at two channels through which ICOs create economic value. Firstly, they help to create platforms (see Li and Mann 2018). With many platform applications, the benefit of all participants increases with the number and activities of other users through network effects, so that joining is only worthwhile if there is a sufficiently large number of other users.Footnote 23 By buying the (utility) tokens, investors send a signal that they will participate in the platform developed by the startup. This reduces the risk for other participants that their benefit will be negative due to fewer further participants. ICOs thus solve the underlying coordination problem. If an ICO takes place over several rounds, an increase in the token price is incentive-compatible and justified by the fact that the project has a higher value due to the existing commitments from previous ICO rounds. A second channel through which ICOs create value is the so-called wisdom-of-the-crowd effect. In contrast to a classic Initial Public Offering (IPO), in which investment banks are responsible for the aggregation of existing information and the evaluation of the company, the evaluation process for ICOs is decentralized and is strongly driven by "online analysts" and early investors (see also Lee et al 2019). Through the recommendations and investment decisions of these groups, combined with the possibility of carrying out ICOs over several rounds, less well-informed participants can observe the investment behavior of the experts and help to achieve the critical mass of participants.

Both effects are closely related to the dual role of ICOs as a right of use and an investment. A high level of interest in the tokens indicates a high level of usage, which leads to a feedback on the value of the shares via network effects. A higher value in turn leads to more attention and thus to a higher adaptation of the platform.Footnote 24

Securities settlement

Securities are traded on stock exchanges or over-the-counter platforms. However, once a transaction is completed, the trading process is not completed. The mutual delivery obligations must be determined (clearing) and the securities must be transferred from the seller to the buyer and the purchase price must be transferred from the buyer to the seller (settlement). Currently, the final settlement of a transaction does not take place until two working days after the trading day. It takes place with the involvement of a number of specialized institutions such as brokers, clearing houses and central securities depositories (central securities depositories). The market is fragmented in the sense that there are several systems (so-called "silos") operating in parallel, such as at Deutsche Börse, where Eurex Clearing carries out the clearing and Clearstream carries out the settlement. This fragmentation in turn requires the existence of interfaces between the systems. This architecture makes the processing of securities transactions in general, and the processing of cross-system transactions in particular, complex and expensive.

A blockchain can be described as a process by which the ownership and transfer of ownership of non-tangible assets can be documented by establishing a consensus on transactions that have been carried out. In this respect, a blockchain should in principle be suitable for processing securities transactions.Footnote 25 However, the requirements for a blockchain in this application are different than for cryptocurrencies, for example, since securities transactions are subject to regulatory and tax rules as well as data protection requirements that must be complied with (cf. In particular, this excludes a completely anonymous system architecture.

Various configurations are conceivable.Footnote 26 For example, the infrastructure within a silo could be transferred to a blockchain. The Australian Stock Exchange is well advanced in implementing such an approach. A private blockchain is planned there with access rights for approved participants and third parties with a legitimate interest (such as regulators).Footnote 27 With such a solution, the business processes and intermediaries would essentially remain unchanged. The currently existing fragmentation would also be preserved if it were not possible to develop and implement common standards.

As an extreme solution that makes full use of the potential advantages of blockchain-supported securities trading, one can imagine a system in which all steps - the transaction itself, clearing and settlement - are carried out through entries in a blockchain, which largely eliminates the need for Would mean intermediaries. In the following we will discuss essential aspects of such a solution.

Fully blockchain-powered securities trading has a number of potential advantages. Security ownership would be easy to track.Footnote 28 This also makes it possible to automate processes such as interest and dividend payments through appropriately designed smart contracts. Processing would be significantly more efficient than in the traditional processing system. The previous settlement two trading days after the trading day could be replaced by a significantly faster settlement or even by real-time settlement. The simultaneous execution of security delivery and payment could reduce or even eliminate settlement risks. A central counterparty would then no longer be necessary.Footnote 29 This is potentially significant as a central counterparty accumulates settlement risks and can thus become a “single point of failure” in the system.

Chiu and Koeppl (2019) develop a theoretical model of a public unpermissioned blockchain based on the proof-of-work process, which carries out the delivery of securities and payment simultaneously. The processing speed is made variable by allowing market participants to pay a fee that makes their transaction more attractive to miners so that it can be executed more quickly. The fee revenue, in turn, is required to create incentives for the miners. This in turn results in the consequence that the capacity of the blockchain must be limited - only if there are bottlenecks will market participants be willing to pay a fee for faster execution. Chiu and Koeppl (2019) suspect that a coordination of market participants on such a blockchain design will be difficult to achieve and see here (p. 1750) a potential need for regulatory intervention.

In addition to the above-mentioned challenges in system design - compliance with regulatory, tax and data protection requirements - blockchain-supported securities trading is also associated with potential risks and disadvantages. The efficiency gains would be maximal if there was only one system. Such a system would then be a monopoly with the resulting problems (similar to Benos et al 2017).Footnote 30 Moving to real-time settlement would also create new problems. First, the number of transactions to be processed would increase dramatically. At the moment, so-called netting takes place after the end of a trading day, in which for each clearing participantFootnote 31 the net delivery obligations and delivery claims are determined from the transactions of one day. Settlement then only takes place for these net claims. Such netting is not possible with real-time processing, so that each transaction has to be processed “gross”. Second, with real-time settlement, it is no longer possible to make short sales within a day and close them out by the end of the day. This option is particularly important for market makers. With real-time settlement, they would instead either have to hold high stocks of securities or ensure the ability to deliver for real-time settlement at all times through appropriate securities lending transactions. This could have a negative impact on market liquidity.

Transactions involving derivative financial instruments may pose additional problems. In such transactions, market participants must regularly provide margins. Losses already incurred from existing positions as well as future loss risks are covered by this collateral. Changes in price lead to changes in the amount of collateral required. If a market participant is unable to meet an increased security requirement (a so-called margin call), his position is forcibly closed. Smart contracts allow these processes to be automated and processed in real time. This has advantages in terms of reducing counterparty risks, but also disadvantages, since market participants have to permanently hold liquidity for possible margin calls. In extreme cases - namely when it comes to compulsory closeouts - adverse price effects can occur.

The above explanations show that the “revolutionary” solution of a completely blockchain-supported trade brings with it numerous problems. This makes it likely that in the foreseeable future more “evolutionary” solutions, in which only the infrastructure within existing institutions is transferred to a blockchain, will be implemented.

Corporate governance

Yermack (2017) discusses the effects of blockchain technology on corporate governance.Footnote 32 His reasoning is based on the assumption that securities trading is completely blockchain-supported. As mentioned above, this makes it easy to keep track of security ownership. This allows issuers to communicate with their shareholders quickly and inexpensively. Lafarre and Van der Elst (2018) argue that blockchain technology can improve the function of the general meeting by reducing the cost of voting by shareholders and strengthening the function of the general meeting as a shareholder forum.

As already mentioned in the previous section, blockchain-supported securities trading will not be anonymous due to the legal framework. It can therefore be assumed that at least the supervisory authorities know the identity of the shareholders. This would make insider trading easier to detect and make it more difficult to stealthily accumulate positions while circumventing regulatory reporting requirements. Information to market participants about reportable securities transactions by company insiders (directors dealings) could take place in real time, which should reduce insider trading profits and increase the information efficiency of the market (see Betzer et al 2015 on the latter point). The re-dating of stock options for managers (so-called backdating, see e.g. Lie 2005) would also be more difficult if the options are also blockchain-based.

Yermack (2017) predicts that less developed economies in particular could be the first to introduce blockchain-based securities trading and take advantage of the resulting governance benefits. His prognosis is based, among other things, on the observation that the existing institutions in these countries are not very trustworthy, which increases the potential advantages of a blockchain solution compared to the existing institutional framework.

outlook

Two central problems of the common public blockchain implementations, as used by Bitcoin, are the high consumption of resources and the low scalability. The high consumption of resources has to do with the incentives set in the system and the proof-of-work consensus mechanism. The remuneration for each newly mined block creates a balance in which the miners have incentives to put new computing power into operation until the marginal costs correspond to the marginal revenue. A high Bitcoin price therefore means that the marginal costs in the form of electricity expenditure are also high and the electricity consumption of the Bitcoin network now corresponds to the electricity consumption of entire countries.Footnote 33

The second problem, the low scalability, is related to the maximum block size specified in the protocol and the time between two blocks. With Bitcoin, a new block with a maximum size of one megabyte is created on average every 10 minutes. This means that around 7 transactions can be carried out per second. If you do that with the requirements of typical payment systems such as B. Comparing Visa, these are several orders of magnitude higher. There are several approaches to solving this problem. On the one hand, the time between two blocks can be reduced (see the crypto currency Litecoin with 2.5 minutes between two blocks) or the size limit per block can be increased (see the crypto currency Bitcoin Cash with currently 32 megabytes per block). However, both options have the consequence that the size and thus the storage requirements of the blockchain, which is kept redundantly on each individual node, grows even more.

Other designs of distributed ledger technologies (DLT) do not lead to such high resource consumption and are more scalable. One example of this is the crypto currency IOTA, which dispenses with a chain structure such as that on which blockchain technology is based.With this cryptocurrency, anyone who wants to carry out a transaction has to confirm other transactions. These are stored in a directed acyclic graph and the core idea is that the higher the proportion of transactions that are later based on them, the more secure they are. The underlying consensus mechanism is, however, much more complex and a consensus in the still small network cannot currently be achieved without a central coordinator.

Cryptocurrencies have desirable properties (especially the combination of decentralized transfer and virtual representation) that cannot be achieved with any other known payment system. Nevertheless, in our opinion, a broad adaptation of "real" cryptocurrencies is not to be expected in the near future. The main reason for this is the loss of efficiency that would result from an additional currency with a value different from that of the main currency (the legal tender). Stablecoins (which also includes the crypto currency Libra initiated by Facebook) do not have this disadvantage. However, stablecoins are backed by a portfolio of assets in traditional currencies (such as government bonds) and therefore recognize a traditional main currency as a benchmark. The adaptation of stablecoins would therefore be a far less revolutionary step than that of a “real” cryptocurrency.

Blockchain technology will show its strengths primarily through smart contracts. These allow detailed regulations to be made based on a broad-based consensus and depending on the states stored in the blockchain and thus now observable. You thus expand the possibilities of concluding contracts. Initial coin offerings enable smaller investors to participate in the company's success at a very early stage, thus benefiting from their expert knowledge and encouraging others to imitate them. Another advantage is the significantly better tradability of crypto coins via crypto exchanges compared to other early forms of participation. Their dual role as right of use and investment helps platform operators to take advantage of the underlying network effects. As explained, securities processing will not be able to be transferred to an anonymous blockchain for regulatory reasons. The implementation of a fully blockchain-based securities trading currently appears unrealistic and is unlikely to be feasible without regulatory intervention. The effects on the corporate governance of listed companies discussed in Sect. 3.5 are largely based on such fully blockchain-based securities trading, so that here, too, revolutionary changes should not be expected in the foreseeable future. In contrast, there will be efficiency gains through blockchain-based processes within the current institutional framework. The same also applies to other fields of application not discussed in detail in this article, such as the issue of promissory note loans.Footnote 34

In summary, it can be said that the disruptive nature of the technology, in addition to the ones discussed, may lead to completely new applications, but their development takes more time than an evolutionary improvement of existing processes. For this reason, it is quite possible that even 10 years after the development of the technology, applications will emerge that have a high potential to fundamentally change economic life in the future.

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Notes

  1. 1.

    The first implementation took place at the beginning of January 2009. It should also be mentioned that the Bitcoin concept was not created in a “vacuum”, but was based on predecessors. An overview is given by Chohan (2017) and Hayes (2019).

  2. 2.

    Current, non-technical reviews of the potential and challenges of blockchain applications in banking and finance can be found in Amin (2020), Chen and Bellavitis (2020) and Tapscott and Tapscott (2017).

  3. 3.

    A more detailed introduction to blockchain technology in the context of Bitcoin application can be found in Berentsen and Schär (2017).

  4. 4.

    A one-way function is easy to compute (i.e., in polynomial time), but difficult to reverse.

  5. 5.

    The reward consists of newly created Bitcoin units and the transaction fees in the respective block. The number of newly created units per block was 50 bitcoins at the beginning. It halves every four years. As of May 2020, it will be 6.25 bitcoins. This mechanism ensures that the number of Bitcoins converges against a fixed value, namely 21 million Bitcoin.

  6. 6.

    At the beginning of 2020, the energy consumption for mining exceeded Austria's consumption, see https://digiconomist.net/bitcoin-energy-consumption (accessed on February 22, 2020). On the same website it is stated that the energy consumption one Bitcoin transaction corresponds to the energy consumption of around 300,000 Visa transactions (verified on May 15, 2020). Even if this number is to be interpreted with caution, it shows that the energy consumption of the Bitcoin network is not only very high in absolute terms, but also in relation to that of other payment systems.

  7. 7.

    In the Bitcoin network, blocks with five or more successor blocks are considered secure components of the blockchain.

  8. 8.

    See also the technical blockchain characteristics in Labazova et al (2019).

  9. 9.

    Strictly speaking, some of the applications discussed here are no longer blockchain technologies but more generally distributed ledger technologies (DLT). Although both terms are often used synonymously, the DLT term is broader and, in addition to the data structure in the form of a blockchain, also includes data structures without a blockchain in the form of an acyclic graph. One example of this is IOTA.

  10. 10.

    For a detailed discussion see e.g. B. Berentsen and Schär (2017).

  11. 11.

    A comparison of the gross domestic product (GDP) of selected currency areas with the transaction volume processed via the Bitcoin blockchain (USD 365 billion in 2019) is a good way of classifying the current degree of diffusion of cryptocurrencies. This corresponds to around 2.7%, 1.8% and 0.4% of the GDP of the euro zone, the United States and the entire world (data sources: blockchain.info, World Bank, data for GDP from 2018). Since many Bitcoin transactions are concluded for purposes other than the purchase of goods and services, the stated percentages should be viewed as an upper limit.

  12. 12.

    Average transaction costs to be paid fluctuate very strongly over time and, for example, amounted to over USD 50 per transaction in the Bitcoin network in times of maximum demand in December 2017 (source: https://blockchain.info).

  13. 13.

    See, for example, the limitation of the Bitcoin circulation to a maximum of 21 million Bitcoin mentioned in footnote 5.

  14. 14.

    For example, a little more than 100 addresses hold over 15% of the assets in circulation in the Bitcoin network, over 21 million addresses - and thus around 73% of all addresses with positive credit - hold a total of only 0.18% of the assets in circulation (data from February 2020 , https://bitinfocharts.com).

  15. 15.

    The volatilities of other exchange rate pairs in the same period are CHF-USD at 5.6% p. a., GBP-USD with 8.3% p. a. and JPY-USD with 5.6% p. a. Calculated with data from Bloomberg, Refinitiv and https://blockchain.info.

  16. 16.

    For a detailed discussion of the Libra concept and the resulting regulatory problems and systemic risks, see Schmeling (2019).

  17. 17.

    See also Narayanan et al (2016) and Berentsen and Schär (2017).

  18. 18.

    See also https://www.skalex.io for further application examples.

  19. 19.

    A hard fork is caused by a change in the consensus protocol that is not compatible with the old protocol, which leads to a split in the blockchain. In contrast, a soft fork does not lead to a split, since the blocks created according to the new protocol still meet the old protocol. Ethereum had a hard fork to reverse the "theft" of tokens, which was made possible by an error in the code (see Securities and Exchange Commission 2017).

  20. 20.