SoftBank Financial Group — A Crypto Assets Manager

CME Group’s Bitcoin futures and NYSE’s Bitcoin futures ETF are milestones in the crypto investing area. The SEC Chairman, Gary Gensler has also stated that he prefers to hold Bitcoin futures funds rather than the cryptocurrency itself, let alone many personal investors. Cryptocurrency investment has a high learning cost, because this is an asset with high volatility, and all participants need to be cautious. For those who are very interested in crypto assets but not interested in buying or holding them, there is still an indirect investment method here, which is to buy crypto asset funds.

SoftBank Financial Group is an investment company that uses big data statistical methods and machine learning as an application system to conduct quantitative investment research on the cryptocurrency market. Headquartered in Canada, it has a complete overseas structure, R&D team, and also has a Canadian MSB license and an American MSB license. Its quantitative investment team was established in 2020, and its core members have working experience in core positions in top hedge funds and first-line Internet companies. The company has built a complete and upgraded industrialized quantitative investment pipeline, has a complete strategic development platform and a quantitative investment system, and has independently developed high-yield, low-drawback, and high-Sharp ratio quantitative investment products.

SoftBank Financial Group continues to upgrade its iterative quantitative strategy, with machine learning methodology as the core, mining multi-factor models from the entire market, and digging deep into big data to make the strategy more adaptable and obtain more stable excess returns under different market conditions.

SoftBank Financial Group has strict risk control, pursues the work style of investment and research first, and rationally controls the scale. In the selection of the target, the correlation and volatility are comprehensively considered to minimize the impact of a single target on the position. The whole process of risk control, risk management runs through the company’s pre-investment, investment, and post-investment links.

SoftBank Financial Group has undergone years of development, back-testing, and real-time operation. It comprehensively uses big data statistical methods, machine learning, and other methods to independently research and develop investment strategies to achieve an annualized rate of return of more than 70 even when the overall market declined by more than 50%. %, and a retracement of no more than 10% of outstanding performance, ranking firmly in the forefront of the industry.

​The emergence of encrypted assets coincides with a series of changes or movements in the traditional investment market, such as machine learning and deep learning technology, making it a perfect asset class for quantitative investment. Quantitative investment in the traditional financial market focuses on predicting market trends, and its data sources mainly rely on third-party data vendors. However, in the field of encrypted assets, the introduction of DeFi has made smart contracts disintermediate, making the functions of smart contracts not only transparent but also accessible through quantitative strategy programming. This makes the quantitative investment of encrypted assets have a new development direction.

​In addition to modeling and trading, DeFi’s quantitative model can also interact with the bottom layer of encrypted assets, such as loan contracts, mortgages, and market-making. These actions will generate an on-chain data footprint, which can be used as unique data in the quantitative model. Source, which does not require data intermediary companies in traditional financial markets. Currently, 21% of hedge funds invest in digital assets, and 86% of them plan to deploy more assets into this category.

​SoftBank Financial Group’s quantitative trading strategy is divided into two parts: one is a price-based strategy, such as a strategy based on momentum or trend tracking; the other is a strategy using non-price signals, such as cryptocurrency network characteristics (number of active addresses, hash rate, etc. ) Or AMM interest rate, etc. We also continue to study the feasibility of traditional financial market strategies in the field of encrypted assets. For example, in many cases, trading strategies that are no longer effective in other asset classes (stocks) can be rejuvenated in encrypted assets.

​The current existing indexes in the cryptocurrency market are lack of consistency and recognition. Some indexes still account for more than 60% of the market value of Bitcoin, and somecompiled large-cap indexes only allocate the top cryptocurrencies according to certain weights. The main gains in those indexes are still the gains of Bitcoin.

​The following figure shows the proportion of mainstream cryptocurrencies in the market. The top 10 currencies account for more than 60% of the overall market value:

​Since the top 30 crypto assets with market capitalization account for more than 80% of the market capitalization, we believe that the Crypto 30 Index can fully reflect the overall situation of the crypto market. In addition, due to the high correlation between various assets and the high volatility of the encrypted assets themselves, our Encryption 30 Index uses these two characteristics to calculate the weight of each asset to ensure that it reflects the market as much as possible, removing high correlation and The impact of high volatility on the index, in order to achieve the stability of the market value and fully reflect the purpose of market information.

​In measuring the many risk factors that cause the asset to lose value, such as macro risk, interest rate risk, asset-specific risk, currency risk, geopolitical risk, liquidity risk, etc., currency price changes are subject to a comprehensive combination of these risk factors in different proportions and degrees. Influence. Correlation includes the fact that these risks change and influence each other over time, leading to an impact on the price behavior of the asset.

​The following figure reflects the correlation heat map between the major cryptocurrencies, indicating that there is a high degree of correlation between the overall cryptocurrencies in the market:

​On the one hand, historical price fluctuations are usually persistent: assets with lower historical volatility are more likely to maintain lower volatility in the short term, while stocks with higher historical volatility are more likely to maintain higher volatility in the short term. On the other hand, the historical price fluctuations of assets with greater risk exposure tend to be greater than those of assets with less risk exposure. Therefore, it can be considered that volatility is an expression of the risks associated with holding a certain currency, and also a comprehensive measure of the risk of that currency.

​Aiming at the two characteristics of the high correlation between the rise and fall of cryptocurrencies and the high volatility of large price changes, Unicorn Park adopted a relatively reasonable basket, added correlation quantification factors, and volatility adjustments, and obtained a low impact on market volatility. To avoid the strong correlation between various currencies to the greatest extent, to achieve a truly stable investment portfolio.

​Due to the high correlation between cryptocurrencies and the active combination of responses to the macro and micro–markets, the crypto 30 index mainly uses principal component analysis when performing correlation analysis.

​Principal component analysis can transform the problems in high-dimensional space into low-dimensional space for processing, making the problem relatively simple and intuitive, and these few comprehensive indicators are not related to each other and can provide most of the original indicators. information. With the process of principal component analysis, the weight of each principal component will be automatically generated, which largely resists the interference of human factors in the evaluation process. Therefore, the encryption index based on the principal component can better guarantee the evaluation. The objectivity of the results faithfully reflects the actual problems of the cryptocurrency market.

​The basic principle of volatility weighting is to give lower weight to index components with higher volatility and higher weight to index components with lower volatility. The goal is to make each component contribute the same amount of expectations to the portfolio. risk. This risk equalization is an intuitive way to strengthen risk diversification. Under ideal circumstances, it can reduce the overall volatility of the investment portfolio, and the volatility level will be lower than the market value weighting method.

​As shown in the figure: BEAN30 Index is affected lightly by market fluctuations, and can truthfully reflect the overall market value changes: