Sonakshi Rohra, Quantitative Researcher Mar 14, 2023
Sector Taxonomies for Digital Assets: Correlation Structures, Diversification Benefits and Quality Assessment
A quantitative and statistical review to compare two taxonomy systems, Digital Asset Taxonomy System (DATS) and Digital Asset Classification Standard (DACS).
Sonakshi Rohra, Quantitative Researcher Feb 07, 2023
Crypto in 2022 - Rising Risks
A review of risks faced by digital assets investors in 2022.
Sonakshi Rohra, Quantitative Researcher Nov 18, 2022
Quantifying Collective Behaviour During the FTX Crash: A Sectoral Study
FTX’s collapse has been a black swan event that came with severe consequences for many investors but also much to learn from. Cloudwall did a review via a sectoral lens of using statistical tools to generate insights t the FTX crash
Boris Skorodumov May 18, 2022
To the Moon and Back: A Factor Lens on the LUNA Crash
Luna’s crash has been a black swan event that came with severe consequences for many investors but also much to learn from. If there is anything we can take away from this, is to ask and answer the question: can we use statistical insights to gain early warnings about such events? Here, we use our risk factor model to decompose risk into factor specific and idiosyncratic risk, as well as study risk contribution amongst individual factors. This gives us insights into how different factors influence the volatility of LUNA, especially right before the crash. We observe a notable divergence in the volatility trend for LUNA at the start of the week of the crash, which is very different from the rest of the assets. More notably, we also observe a regime change for LUNA’s factor risk decomposition, wherein the Momentum factor driving LUNA’s volatility gets replaced by liquidity and volatility factors, serving as an early warning of the liquidity shock.
Sonakshi Rohra, Quantitative Researcher May 16, 2022
Performance Attribution for Crypto Sectoral Indices
In this paper we’ll show how portfolio managers are usually evaluated through performance attribution. As an example, we’ll introduce the Brinson–Fachler model. Instead of evaluating a portfolio manager, we’ll use it to explore DACS, the token taxonomy introduced by CoinDesk. Our objective is to see if using sectoral diversification there’s value to be uncovered. The same methodology can be used to compare e.g., crypto VCs investing in liquid tokens and see if their out- or under-performance is due to asset selection, sector allocation, or the interaction of the two.
Ilya Kulyatin May 06, 2022
A Primer on Factor Investing
While factors have become the foundation of investing, it is increasingly difficult to observe these factors. With a nascent market like crypto, it becomes imperative that we identify and successfully extract targeted factors driving our asset returns. In this article, we come up with a Three Risk Factor Framework, which involves identification: based on economic justification, observable and resistant return patterns, longevity, robustness, and cost-effective implementation. This is followed by a Rationale framework by which asset returns are driven by systematic risk that reflect compensations for providing insurance, exploit market inefficiencies due to persistent investor behavior biases and accommodate structural market imbalances. Lastly, we have a factor extraction framework which includes idiosyncratic factors, macro factors and statistical factors.
Boris Skorodumov Apr 29, 2022
Data-driven Crypto Risk Factors
Usually, risk factor literature assumes that these factors driving returns are observable (at least theoretically) and are usually obtained by a regression model. An alternative approach is to assume that factors are infact unobservable and generate these out of pure statistical methodology. One such technique is Principal Component Analysis (PCA), which can be used to generate orthogonal (or independent) factors that drive variability of returns. In this paper, we apply PCA to extract factors out of a top 1000 (by market cap) crypto asset universe, after data cleaning and processing, leaving a net 103 assets studied. We find that the first 10 factors explain about 60% of variability in returns, which is much lower than the usual explain ability with mature markets like equities, possibly suggesting additional factors in the crypto universe or hinting at a yet amateur market. We also see the effect of survival bias at play by observing a notable increase in total variability when we reduce the data coverage constraints. Lastly, we construct eigen portfolios and compare performance of these with an equally weighted benchmark.
Sonakshi Rohra, Quantitative Researcher Apr 15, 2022
Stylized Facts for Cryptocurrencies: A Sectoral Analysis
Our last paper briefly studied the statistical properties of Bitcoin’s price process. In this paper, we take a step forward and make a thorough study of stylized facts across the crypto market. The paper begins with a statistical analysis of a top 10 token - market cap weighted index. Next, we use the DACS taxonomy to construct sectoral indices and compare statistical behaviour across the 5 sectoral indices. A range of techniques and metrics are utilized such as distribution fitting, Q-Q plots for distribution comparisons, skewness, kurtosis, GARCH fitting and autocorrelation plots to confirm or deny the common stylized facts.