I denna uppsats undersöker vi kryptovalutor ur ett ekonomiskt och finansiellt perspektiv. Vi skapar en optimal portfölj av de 28 största kryptovalutorna enligt marknadsvärdet den 3 mars 2020 genom av att använda oss av Markowitz (1952) portföljvalsteori. Den optimala portföljen jämförs med en marknadsportfölj som är skapad av de 100 största kryptovalutorna enligt marknadsvärdet, för att bilda oss en uppfattning om kryptovalutamarknaden och hur man kan utnyttja kryptovalutor i investeringssyfte. CAPM används för att kunna se relationen mellan risk och avkastning mellan den optimala portföljen och marknadsportföljen. Vi kommer även att undersöka om CAPM ger samma resultat som Markowitz portföljvalsteori. Den optimala portföljen jämförs dessutom med den svenska aktiemarknaden för att undersöka om den optimala kryptovalutaportföljen påverkas av trender på den svenska aktiemarknaden. / This thesis studies cryptocurrencies from an economic and financial perspective. The research is carried out by constructing an optimal portfolio of the 28 biggest cryptocurrencies according to market capital on the 3rd of March 2020 by using Markowitz (1952) portfolio optimization theory. The optimal portfolio is then compared to the market portfolio, which is constructed of the hundred largest cryptocurrencies according to market capital, to study the cryptocurrency market. CAPM is also used to find out the risk-return relationship and to see if CAPM gives us the same optimal portfolio as Markowitz portfolio optimization theory. The optimal portfolio is also compared to the Swedish stock market index, OMXS30, to study if the optimal portfolio is affected by trends in the Swedish stock market.

Various forms of fraud have substantially impacted our lives and caused considerable losses to some people. To reduce these losses, many researchers have devoted themselves to the study of fraud detection. After the development of fraud detection from expert-driven to data-driven systems, the scalability and accuracy of fraud detection have been improved considerably. However, most existing fraud detection methods focus on the feature extraction and classification of a certain transaction, ignoring the temporal and spatial long-term information from accounts. In this work, we propose to address these limitations by employing a lightweight memory network (LiMNet), which is a deep neural network that captures causal relations between temporal interactions. We evaluate our approach on two data sets, the Ether-Fraud dataset, and the Elliptic dataset. The former is a brand new dataset collected from Etherscan with data mining, and the latter is published by the homonymous company. As a set of raw collected data never used before, the Ether-Fraud dataset had some issues, such as huge variation among values and incomplete information. Therefore we have processed Ether-Fraud with data supplementation and normalization, which has solved these problems. A series of experiments were designed based on our analysis of the model and helped us to find the best hyper-parameter setting. Then, we compared the performance of the model with other baselines, and the results showed that Lightweight Memory Network (LiMNet) outperformed traditional algorithms on the Ether-Fraud dataset but was not good as the graph-based method on the Elliptic dataset. Finally, we summarized the experience of applying the model to fraud detection, the strengths and weaknesses of the model, and future directions for improvement. / Olika former av bedrägerier har haft en betydande inverkan på våra liv och har orsakat stora förluster för vissa människor. För att minska dessa förluster har många forskare ägnat sig åt att studera upptäckt av bedrägerier. Efter utvecklingen av bedrägeriutredningen från expertdrivna till datadrivna system har skalbarheten och noggrannheten förbättrats avsevärt. De flesta av de befintliga metoderna för upptäckt av bedrägerier fokuserar dock på utvinning av funktioner och klassificering av en viss transaktion och ignorerar den temporala och spatiala långsiktiga informationen från konton. I det här arbetet föreslår vi att vi tar itu med dessa begränsningar genom att använda ett lättviktigt minnesnätverk (LiMNet), som är ett djupt neuralt nätverk som fångar kausala relationer mellan temporala interaktioner. Vi utvärderar vårt tillvägagångssätt på två datamängder, datamängden Ether-Fraud och Elliptic-datamängden. Det förstnämnda är ett helt nytt dataset som samlats in från Etherscan med hjälp av datautvinning, och det sistnämnda är publicerat av det homonyma företaget. Eftersom det rörde sig om råa insamlade data som aldrig använts tidigare hade Ether-Fraud-datasetet vissa problem, t.ex. en stor variation mellan värdena och ofullständig information. Därför har vi bearbetat Ether-Fraud med datatillägg och normalisering, vilket har löst dessa problem. En serie experiment utformades utifrån vår analys av modellen och hjälpte oss att hitta den bästa inställningen av hyperparametrar. Sedan jämförde vi modellens prestanda med andra baslinjer, resultaten visade att LiMNet överträffade traditionella algoritmer på datasetet Ether-Fraud men var inte lika bra som den grafbaserade metoden på datasetet Elliptic. Slutligen sammanfattade vi erfarenheterna av att tillämpa modellen på bedrägeridetektion, modellens styrkor och svagheter samt framtida riktningar för förbättringar.

The Internet provided humans a new way to exchange information digitally and has changed how we communicate. Blockchain and cryptocurrencies have given humans a new way to exchange value over the internet. With new technology, new possibilities arise, but not always without issues. One problem that has risen with cryptocurrencies is their high volatility, meaning that the currency has big price swings. It has made these currencies objects for speculation and investment almost exclusively, and therefore they have lost their functionality as a currency. For a currency to be viewed as a good means of payment, it cannot be associated with high volatility. This is not only restricted to cryptocurrencies, as for example the Venezuelan Bolivar is a fiat currency with historically high volatility and has been losing its purchasing power due to hyperinflation in the recent years. In regard to this we propose a new cryptocurrency; the Dynamic Network Token, which aims to reduce the volatility in a cryptocurrency by regulating the supply dynamically with burning and minting. The implementation of this functionality will strive to remove the high volatility in the token for the benefits of a more stable and linear growth, and at the same time encourage users to transact with the Dynamic Network Token between each other. / Internet gav människor möjlighet att utbyta information digitalt och har förändrat hur vi kommunicerar. Blockkedjeteknik och kryptovalutor har gett människan ett nytt sätt att utbyta värde på internet. Med ny teknologi kommer möjligheter, men kan även medföra problem. Ett problem som uppstått med kryptovalutor är deras volatilitet, vilket betyder att valutan upplever stora prissvängningar. Detta har gjort dessa valutor till objekt för spekulation och investering, och därmed gått ifrån sin funktion som valuta. För att en valuta ska anses som ett bra betalmedel, bör den inte ha hög volatilitet. Detta är inte bara begränsat till kryptovalutor, då till exempel Venezuelas nationella valuta Bolivar är en fiatvaluta med historiskt hög volatilitet som förlorat sin köpkraft på grund av hyperinflation under de senaste åren. Med detta i åtanke föreslår vi en ny kryptovaluta; Dynamic Network Token, vars uppgift är att reducera volatiliteten i en kryptovaluta genom att reglera utbudet dynamiskt med hjälp av burning och minting. Denna implementeringsuppgift är att minska hög volatilitet till fördel för en mer stabil och linjär tillväxt och samtidigt uppmana användare att använda Dynamic Network Token mellan varandra i nätverket.

The emergence of stablecoins and their current implementations share many similarities to the American free banking era. This was an era with economically inefficient money and payment systems, where banks issued private money that were fully redeemable in theory but not always in practice. As of today, parts of the monetary and payment systems can also be considered inefficient, and a digital currency such as a stablecoin could provide significant improvements in areas such as cross-border payments, financial inclusion, as well as contribute to a more robust monetary system.  This thesis aims to help contribute insights into how the future payment systems may look like, with a specific focus on stablecoins and central bank digital currencies (CBDCs). The emergence of the two technologies are closely interlinked, and are currently in an early state of coexistence. This qualitative study investigates the sentiment regarding the future development of stablecoins and their possible continued coexistence with CBDCs. Semi-structured interviews were held with participants from three categories of the Swedish corporate landscape; cryptocurrency innovators and investors, state representatives, and corporate representatives. The main conclusions were was that in the short term, regulations will be a key enabler for continued stablecoin development. There is currently a lack of clarity and guidelines which is making it hard for generally accepted stablecoins to be established. Furthermore, the general consensus is that stablecoins and central bank digital currencies will co-exist in the future monetary system in the long-term, where each technology will have different use cases. / Framväxten av stablecoins och dess nuvarande implementationer delar många likheter med den amerikanska ”free-banking” eran. Detta var en tidsepok med ekonomiskt ineffektiva pengar och betalsystem, där banker emitterade pengar som skulle vara fullt inlösningsbara teorin, men inte alltid i praktiken. Även idag kan delar av penga- och betalsystemen anses vara ineffektiva, och en digital valuta såsom en stablecoin skulle kunna bidra med betydande förbättringar inom områden såsom utrikesbetalningar, finansiell inkludering, samt bistå i utvecklingen av ett mer robust pengasystem. Denna uppsats ämnar bidra med insikter om hur framtidens betalsystem kan se ut, med ett specifikt fokus på stablecoins och centralbanksvalutor (eng. central bank digital currencies, CBDCs). Framväxten av de två teknologierna är nära relaterade, och för närvarande samexisterar dessa i ett tidigt stadie. Denna kvalitativa studie undersöker sentimentet kring den framtida utvecklingen av stablecoins och dess möjliga fortsatta samexistens med centralbanksvalutor. Semi-strukturerade intervjuer hölls med deltagare från tre kategorier av svenskt näringsliv: innovatörer och investerare inom kryptovalutor, företrädare för staten, samt bolagsrepresentanter. De huvudsakliga slutsatserna var att på kort sikt kommer regleringar vara viktiga för att möjliggöra fortsatt utveckling av stablecoins. I dagsläget saknas det klarhet och riktlinjer vilket gör det svårt för en generellt accepterad stablecoin att utvecklas. Vidare så är den generella uppfattningen att stablecoins och centralbanksvalutor kommer samexistera i framtidens betalsystem, då de båda teknikerna kommer ha olika användningsområden

The purpose of this study was to create understanding about how the reduction of cash affects societal ideas about integrity. We wanted to research human attitudes to the ongoing reduction of cash use in Sweden from a integrity perspective. The method used was a qualitative approach and the empirical data was collected using a focus group with students and three semi structured interviews in different branches of industry based on their way of working with and handling cash. Participating in the study was five students, a store manager in the grocery store business, an economic advisor at a bank and three investigators at the Swedish central bank. The result was analyzed using a theoretical foundation based on Georg Simmel’s work “The philosophy of money” and other socioeconomic research with a focus on the terms value, power and integrity. The results showed that people in the swedish society are aware and partially wary of the development towards a cash free society. Advantages that were illuminated was a reduced risk of robbery towards banks and businesses, less environmental effects due to reduced transportation and handling of cash, easier payments for consumers and increased profitability of card payments for businesses and a increased difficulty of usage of “black” money within the society. Negative aspects that were brought up were the issues of vulnerable groups in society that rely on cash, a fear that the development is moving too fast and that the society will be increasingly vulnerable when the payment infrastucrure relies on a single point of failure. The integrity aspects that the people in the study discussed showed that society is aware of the risks surrounding digital payments, but that the advantages outweigh the potential negatives, and that there is a strong trust in institutions and businesses in Sweden to not exploit their positions when it comes to the integrity of private individuals. / Studiens syfte var att skapa förståelse hur kontantavvecklingen påverkar samhälleliga föreställningar kring integritet. Vi ville undersöka människors inställningar till den rådande kontantavvecklingen i Sverige från ett integritetsperspektiv. Metoden som användes var kvalitativ och det empiriska materialet samlades in genom att genomföra en fokusgrupp med studenter och tre semi-strukturerade intervjuer med branscher som vi valde ut efter deras olika arbetssätt med kontanter och digitala betalningsmedel. I studien deltog fem studenter, en butikschef för dagligvaruhandel, en ekonomisk rådgivare inom bankbranschen samt tre utredare på Riksbanken. Resultatet analyserades med en teoretisk grund baserat på Georg Simmels verk “The philosophy of money” och annan socioekonomisk forskning med fokus på begreppen värde, makt, och integritet. Resultaten visade att människor i det svenska samhället är medvetna och till viss del oroade av utvecklingen mot ett kontantfritt samhälle. Fördelar som uppmärksammades var mindre risk för rån i banker och butiker, mindre miljöpåverkan av transporter av kontanter, enklare betalning för konsumenten och bättre lönsamhet för kortbetalningar för företagen samt att det försvåras att spendera svarta pengar i samhället. De negativa aspekterna av ett kontantfritt samhälle som togs upp var problemen som uppstår för utsatta grupper i samhället som är mer beroende av kontanter, en rädsla för att utvecklingen går för fort och att samhället blir mer sårbart när infrastrukturen förlitar sig på att de digitala betalningssätten fungerar även i krissituation då systemet får en inneboende “single point of failure”. De integritetsaspekter som människorna resonerade kring visade att samhället är medvetet om riskerna med digitala betalningssätt, men att fördelarna med dessa vägde upp de potentiellt negativa, och att det finns ett starkt förtroende för att myndigheter och företag i Sverige inte missbrukar sin ställning när det kommer till privatpersoners integritet.

Background: Within the emerging field of cryptocurrencies, the sub-sector DeFi (decentralized finance) has experienced explosive growth over the last year, and its importance for crypto as a whole has grown with it. The currencies have developed from simple peer-to-peer transactions to complex applications such as lending and exchanges. Several studies have researched determinants of cryptocurrency prices, and a few have focused on metrics central to DeFi, such as total value locked (TVL). However, academia has aimed sparse attention to the relationships between these metrics, which this article seeks to amend.  Aim: The purpose of this essay is to research the relationship between total value locked (TVL) in DeFi, the prices of native tokens on related platforms, and the price of ether, which is the dominant currency across DeFi.  Methodology: This study is deductive and quantitative and categorized as a causal-comparative thesis. The purpose of causal-comparative research is to find relationships between variables, independent and dependent, over a certain period.  The authors used deductive reasoning to form the hypotheses and collect the data necessary to investigate the hypothesis. Additionally, the structure of the paper and the epistemological process is quantitative and based on the scientific method.  The sources used for data gathering have primarily been DefiPulse and their API:s, retrieved using simple python coding and different applications that parse JSON code into the excel format. The transparent nature of blockchain has provided easy access to data needed for this study. Once the data was collected, it was categorized and compiled into an Excel sheet.  Conclusions: It is a considerable result that the ratio of locked ETH to total supply lacks significance for the price of ether, as it is counterintuitive to the macroeconomic theory of demand and supply. Presumably, the locked eth is not to be considered as a corresponding decrease in supply. However, if that was the case, the locked ratio of 10% is considerable and should affect the price as there is less supply available to the market.  In accordance with hypotheses two, three, and four, changes in the price of ether, TVL, and utilization rate affect the price of the native token. A notable distinction between the three different platforms lies in what metrics correlate more strongly with price changes. It for Compound and Aave was TVL, but utilization rate for MakerDAO. What causes these differences between seemingly similar platforms is a subject for further study.

Financial time-series are not uncommon to research in an academic context. This is possibly not only due to its challenging nature with high levels of noise and non-stationary data, but because of the endless possibilities of features and problem formulations it creates. Consequently, problem formulations range from classification and categorical tasks determining directional movements in the market to regression problems forecasting their actual values. These tasks are investigated with features consisting of data extracted from Twitter feeds to movements from external markets and technical indicators developed by investors. Cryptocurrencies are known for being evermore so volatile and unpredictable, resulting in institutional investors avoiding the market. In contrast, research in academia often applies state-of-the-art machine learning models without the industry’s knowledge of pre-processing. This thesis aims to lessen the gap between industry and academia by presenting a process from feature extraction and selection to forecasting through machine learning. The task involves how well the market movements can be forecasted and the individual features’ role in the predictions for a six-hours ahead regression task. To investigate the problem statement, a set of technical indicators and a feature selection algorithm were implemented. The data was collected from the exchange FTX and consisted of hourly data from Solana, Bitcoin, and Ethereum. Then, the features selected from the feature selection were used to train and evaluate an Autoregressive Integrated Moving Average (ARIMA) model, Prophet, a Long Short-Term Memory (LSTM) and a Transformer on the spread between the spot price and three months futures market for Solana. The features’ relevance was evaluated by calculating their permutation importance. It was found that there are indications of short-term predictability of the market through several forecasting models. Furthermore, the LSTM and ARIMA-GARCH performed best in a scenario of low volatility, while the LSTM outperformed the other models in times of higher volatility. Moreover, the investigations show indications of non-stationary. This phenomenon was not only found in the data as sequence but also in the relations between the features. These results show the importance of feature selection for a time frame relevant to the prediction window. Finally, the data displays a strong mean-reverting behaviour and is therefore relatively well-approximated by a naive walk. / Finansiella tidsserier är inte ovanliga att utforska i ett akademiskt sammanhang. Det beror troligen inte bara på dess utmanande karaktär med höga ljudnivåer och icke-stationära data, utan även till följd av de oändliga möjligheter till inmatning och problemformuleringar som det skapar. Följaktligen sträcker sig problemformuleringarna från klassificering och kategoriska uppgifter som bestämmer riktningsrörelser på marknaden till regressionsproblem som förutsäger deras faktiska värden. Dessa uppgifter undersöks med data extraherad från twitterflöden till rörelser från externa marknader och tekniska indikatorer utvecklade av investerare. Kryptovalutor är kända för att vara volatila och oförutsägbara till sin natur, vilket resulterar i att institutionella investerare undviker marknaden. I kontrast tillämpas forskning inom den akademiska världen ofta med avancerade maskininlärningsmodeller utan branschens typiska förbearbetningsarbete. Detta examensarbete syftar till att minska klyftan mellan industri och akademi genom att presentera en process från dataextraktion och urval till prognoser genom maskininlärning. Arbetet undersöker hur väl marknadsrörelserna kan prognostiseras och de enskilda variablernas roll i förutsägelserna för ett regressionsproblem som prognotiserar en sex timmar fram i tiden. Därmed implementerades en uppsättning tekniska indikatorer tillsammans med en algoritm för variabelanvändning. Datan samlades in från börsen FTX och bestod av timdata från Solana, Bitcoin och Ethereum. Sedan användes variablerna som valts för att träna och utvärdera en Autoregressive Integrated Moving Average (ARIMA)-modell, Prophet, en Long Short-Term Memory (LSTM) och en Transformer på skillnaden mellan spotpriset och tre månaders framtidsmarknad för Solana. Variablernas relevans utvärderades genom att beräkna deras vikt vid permutation. Slutsatsen är att det finns indikationer på kortsiktig förutsägbarhet av marknaden genom flera prognosmodeller. Vidare noterades det att LSTM och ARIMA-GARCH presterade bäst i ett scenario med låg volatilitet, medan LSTM överträffade de andra modellerna i vid högre volatilitet. Utöver detta visar undersökningarna indikationer på icke-stationäritet inte bara för datan i sig, utan också för relationerna mellan variablerna. Detta visar vikten av att välja variabler för en tidsram som är relevant för prediktionsfönstret. Slutligen visar tidsserien ett starkt medelåtergående beteende och är därför relativt väl approximerad av en naiv prediktionsmodell.

Bitcoin's decentralized nature enables reasonably anonymous exchange of money outside of the authorities' control. This has led to Bitcoin being popular for various illegal activities, including scams, ransomware attacks, money laundering, black markets, etc.  In this thesis, we characterize this landscape, providing insights into similarities and differences in the use of Bitcoin for such activities.  Our analysis and the derived insights contributes to the understanding of Bitcoin transactions associated with illegal activities through three main aspects. First, it offers a comprehensive characterization of money flows to and from Bitcoin addresses linked to different abuse categories, revealing variations in flow patterns and success rates. Second, a temporal analysis captures long-term trends and weekly patterns across categories. Finally, an analysis of outflow from reported addresses uncovers differences in graph properties and flow patterns among illicit addresses and between abuse categories. These findings provide valuable insights into the distribution, temporal dynamics, and interconnections within various categories of Bitcoin transactions related to illicit activities.

[pt] Com o surgimento dos criptoativos em 2009, iniciado com o Bitcoin, uma nova dinâmica de investimento e de tecnologia emergiu no século XXI com um novo mercado que já chegou a mais de 800 bilhões de dólares em 2018 e conta com mais de 2.000 moedas. Apesar da elevada volatilidade, de vários escândalos de pirâmides, da ausência de regulamentação e da maior utilização como investimento do que em compras de bens e serviços, os criptoativos vêm ganhando seu espaço, em meio às controvérsias, devido a tecnologia disruptiva. Este trabalho tem por objetivo analisar os 50 maiores criptoativos do mercado durante o período de 2015 – 2018 por meio de três ensaios que abordam: (i) a análise e previsão de volatilidade utilizando o MSGARCH (KLAASSEN, 2002), com testes de acurácia (envolvendo funções perda EQM e QLIKE, bem como o MAE, MAPE e o indicador U de Theil); (ii) análise dos fenômenos comportamentais de efeito manada seguindo modificações nas metodologias CSSD (CHRISTIE E HUANG, 1995), CSAD (CHANG, CHENG E KHORANA, 2000) e HS (HWANG E SALMON, 2004), bem como o efeito contágio seguindo modificações nas metodologias do teste FR (FORBES E RIGOBON, 2002) e de testes de comomentos de ordem superior (FRY, MARTIN E TANG, 2010; FRY-MCKIBBIN E HSIAO, 2018); (iii) bem como a análise do fenômeno de feedback trading por meio do modelo seminal de Sentana e Wadhwani (1992). Como principais achados, foi identificado que: (i) há uma forte influência de dois estados de volatilidade; nos criptoativos com maior probabilidade de ocorrência do segundo regime existe uma maior tendência do aparecimento do segundo estado de volatilidade com a subida de preços, onde existe elação ao efeito manada, o modelo CSAD detectou um efeito pouco significativo, e o modelo CSSD detectou um efeito manada forte estatisticamente significativo no movimento de queda de mercado; o modelo HS capturou com sucesso o comportamento de manada e revelou períodos extremos de manada reversa; em relação ao efeito contágio, o teste FR conseguiu captar contágio do Bitcoin em outras moedas em praticamente todos os casos com exceção do Tether Dollar, BITCNY e ECC - que tipicamente possuem controle inflacionário e particularidades das stablecoins; nos modelos de comomentos, os testes indicaram contágio do Bitcoin em relação as moedas analisadas; (iii) em relação ao fenômeno de feedback trading, foi possível captar feedback trade negativo no TETHER e positivo nas moedas BTC, ETH, CSC e ECC, cuja adequação do modelo utilizado foi confirmada posteriormente pelo teste de viés de sinais (ENGLE E NG, 1993), com exceção do TETHER - que contrariou Sentana e Wadhwani (1992) e Shi, Chiang e Liang (2012) ao apontarem que modelos menos parcimoniosos teriam pouca influência na verificação de feedback trading. / [en] With the arise of cryptocurrencies in 2009, started with the Bitcoin, a new dynamic of investment and technology emerged in the 21st century with a new market that has already exceed US 800 billion in 2018 and has more than 2,000 coins. Despite the high volatility, various Ponzi schemes, lack of regulation and the main use as investment than in purchases goods and services, cryptocurrencies have been gaining ground, amid controversy, due to the disruptive technology. The objective of this work is to analyze the 50 largest cryptocurrencies in the market during the period of 2015-2018 by means of three essays that seek to investigate: (i) the volatility analysis and prediction using MSGARCH (KLAASSEN, 2002), with accuracy tests (involving MSE and QLIKE loss functions, as well as MAE, MAPE, and Theil s U indicator); (ii) the analysis of the behavioral phenomena of herd effect following modifications in CSSD (CHRISTIE e HUANG, 1995), CSAD (CHANG, CHENG e KHORANA, 2000) and HS (HWANG e SALMON, 2004) methodologies, as well as the contagion effect following modifications in the methodologies of FR test (FORBES e RIGOBON, 2002) and higher order comoments tests (FRY, MARTIN e TANG, 2010; FRY-MCKIBBIN e HSIAO, 2018); (iii) the analysis of the feedback trading phenomenon through the seminal model of Sentana and Wadhwani (1992). As main findings, it was identified that: (i) there is a strong influence of two volatility states; in the cryptoassets with more probability of occurrence under the second regime, there is a greater tendency of occurrence of the second state of volatility when prices go up, where there is more the volatility - the exception that has been noted only in BTC and ETH, where the first state of volatility is strong when prices go up, with more volatility; there is more accuracy in the forecasting with two volatility states for long term prediction than in short term prediction; (ii) with respect to the herd effect, the CSAD model detected a small herd effect, with little statistical significance, and the CSSD model detected a strong herd effect statistically significant in the down movement of market; the HS model successfully captured herd behavior and revealed extreme periods of reversal in the herd effect; in relation to the contagion effect, the FR test was able to capture Bitcoin s contagion in other currencies in practically all cases except Tether Dollar, BITCNY and ECC - which typically have inflationary control and particularities of stablecoins; in the comoments models, the tests indicated contagion of Bitcoin in relation to the currencies analyzed; (iii) in relation to the feedback trading phenomenon, it was possible to capture negative feedback trading in TETHER and positive in BTC, ETH, CSC and ECC, whose adequacy of the model used was confirmed later by the signal bias test (ENGLE e NG, 1993), with the exception of TETHER - which contradicts Santana and Wadhwani (1992) and Shi, Chiang and Liang (2012) that less parsimonious models would have little influence on feedback trading.

[pt] O Bitcoin surgiu no fim da década passada. Desde então, emergiu uma nova classe de ativos: as criptomoedas. O ecossistema das moedas digitais vem avançando a passos largos, seja pelo surgimento de novas moedas, pelo nível de capitalização, pela escalada de investidores ou pelo expressivo desempenho em 2017. Dado o quadro, as criptomoedas se consolidam a cada dia como uma alternativa de investimento, tornando-se de vez uma rota do mercado financeiro. Por consequência, surge a necessidade de avaliar e estimar medidas de risco para esses ativos. Este estudo estimou os coeficientes Beta das quatorze maiores criptomoedas da economia – de acordo com o nível de capitalização – em relação à índices teóricos, com o fito de auxiliar os gestores de portfólios no apreçamento e na formatação de estratégias. Através de uma regressão de retornos passados destas moedas sobre os retornos dos índices de criptoativos, de ações e de uma cesta de moedas contra o dólar americano, estimou-se o Beta dos ativos. A partir das análises, concluiu-se que o Bitcoin possui elevada sensibilidade aos índices de criptomoedas, mesma condicionante para o Ethereum, porém com correlação mais branda aos referenciais, bem como ao próprio Bitcoin. Quantos às demais moedas, estas não exprimiram fator de risco associado aos índices de criptomoedas, visto os baixos coeficientes. Quando analisados os criptoativos em relação aos índices acionários e de moedas contra o dólar, constatou-se que os coeficientes foram iguais a zero. Portanto, o desempenho das criptomoedas, na janela de tempo estudada, possui relação involuntária às oscilações destes índices. / [en] Bitcoin has risen at the end of the last decade. Since then, a new class of assets emerged: the cryptocurrencies. The cryptocurrencies scenario has been advancing rapidly, by the emergence of new currencies, by the level of capitalization, either by the increase of investors or by its significant performance in 2017. Given the situation, cryptocurrencies keep consolidating itself every day as an investment alternative, becoming a permanent route for the financial market. Consequently, it becomes necessary to estimate risk measures for these assets. This study estimated the Beta coefficients of the largest cryptocurrencies – according to its capitalization level – in relation to theoretical indexes, in order to assist portfolios managers in pricing. Through a regression of past returns of virtual currencies on the returns of the cryptocurrencies indexes, stocks and a portfolio of currencies against the US dollar, the digital assets beta was estimated. From these analyses, it was possible to conclude that Bitcoin is significantly sensitive to cryptocurrencies indexes, the same condition for Ethereum, but with a softer correlation to the references, as well as Bitcoin itself. With respect to the other currencies, they did not express a relevant risk factor associated with cryptocurrencies indexes, due to low coefficient values. When analyzing the cryptocurrency in relation to the stock and currency indexes against the US dollar, it was noted that the coefficients were zero. Therefore, the digital currencies performance of this study, in the given timeframe, has an involuntary relation to the fluctuations of those indexes.