Using Two Cryptocurrencies Enhances Volatility Forecasting
Researchers from the HSE Faculty of Economic Sciences have found that Bitcoin price volatility can be effectively predicted using Ethereum, the second-most popular cryptocurrency. Incorporating Ethereum into a predictive model reduces the forecast error to 23%, outperforming neural networks and other complex algorithms. The article has been published in Applied Econometrics.
Cryptocurrencies are among the most unpredictable financial instruments. Their prices can swing sharply and frequently—a characteristic known as volatility, meaning that an asset's value can change dramatically even over a short period. In contrast, stocks and bonds tend to be much more stable.
The cryptocurrency market responds sharply to external events. Its total market capitalisation—the combined value of all issued coins—has already surpassed $3 trillion. Bitcoin and Ethereum make up over 60% of this market, trading $18 billion and $8 billion respectively in a single day.
This is why the ability to predict sudden price fluctuations has become crucial for risk management and investment strategies. Until now, researchers have primarily studied the volatility of each cryptocurrency separately. Anatoly Peresetsky, Research Professor at the HSE Faculty of Economic Sciences, and Maksim Teterin, a doctoral student, have proposed a new approach that examines the interdependence between the two cryptocurrencies.
To achieve this, the authors analysed high-frequency trading data for Bitcoin and Ethereum at five-minute intervals from January 1, 2018, to June 23, 2024. After filtering out unreliable data, they obtained over 2,300 observations for each cryptocurrency, and then calculated realised volatility, which measures how much an asset's price fluctuated.
Anatoly Peresetsky
'For example, Ethereum prices show slightly higher volatility. Interestingly, both cryptocurrencies experience pronounced extreme swings, underscoring the high-risk nature of the market,' explained Prof. Peresetsky. The authors also calculated realised covariances and correlations between the assets, reflecting the degree of interdependence and the relationship between Bitcoin’s and Ethereum’s price movements.
The researchers then constructed 11 variations of econometric models, each with different parameter combinations. Each model was trained using a 300-day observation window and then used to generate day-ahead forecasts. This approach enabled them to account for structural shifts in the market. Altogether, the procedure produced over 2,000 day-ahead forecasts per model.
To evaluate the accuracy of the models’ forecasts, standard metrics were used: root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). Each metric measures the extent to which the model’s predictions deviate from the actual values on the trading day.
Ultimately, the baseline model produced a MAPE of 24.65% for Bitcoin and 20.98% for Ethereum. The best-performing model, which incorporated parameters for both cryptocurrencies, reduced the error to 23.38%. These results outperform previous studies using neural networks and other advanced models, which reported a MAPE of 25.6%.
'Compared to traditional financial assets, Bitcoin and Ethereum remain the most volatile instruments in the market. This high volatility presents opportunities, but it also entails significant risks. As a result, developing more accurate volatility forecasts is of great interest to both investors and regulators,' commented Prof. Peresetsky.
The study’s findings can help optimise investment portfolios and shape financial strategies in response to cryptocurrency market volatility, while emphasising the importance of monitoring the network of interdependencies rather than focusing solely on individual tokens.
See also:
Larger Groups of Students Use AI More Effectively in Learning
Researchers at the Institute of Education and the Faculty of Economic Sciences at HSE University have studied what factors determine the success of student group projects when they are completed with the help of artificial intelligence (AI). Their findings suggest that, in addition to the knowledge level of the team members, the size of the group also plays a significant role—the larger it is, the more efficient the process becomes. The study was published in Innovations in Education and Teaching International.
New Models for Studying Diseases: From Petri Dishes to Organs-on-a-Chip
Biologists from HSE University, in collaboration with researchers from the Kulakov National Medical Research Centre for Obstetrics, Gynecology, and Perinatology, have used advanced microfluidic technologies to study preeclampsia—one of the most dangerous pregnancy complications, posing serious risks to the life and health of both mother and child. In a paper published in BioChip Journal, the researchers review modern cellular models—including advanced placenta-on-a-chip technologies—that offer deeper insights into the mechanisms of the disorder and support the development of effective treatments.
Administrative Staff Are Crucial to University Efficiency—But Only in Teaching-Oriented Institutions
An international team of researchers, including scholars from HSE University, has analysed how the number of non-academic staff affects a university’s performance. The study found that the outcome depends on the institution’s profile: in research universities, the share of administrative and support staff has no effect on efficiency, whereas in teaching-oriented universities, there is a positive correlation. The findings have been published in Applied Economics.
Physicists at HSE University Reveal How Vortices Behave in Two-Dimensional Turbulence
Researchers from the Landau Institute for Theoretical Physics of the Russian Academy of Sciences and the HSE University's Faculty of Physics have discovered how external forces affect the behaviour of turbulent flows. The scientists showed that even a small external torque can stabilise the system and extend the lifetime of large vortices. These findings may improve the accuracy of models of atmospheric and oceanic circulation. The paper has been published in Physics of Fluids.
Solvent Instead of Toxic Reagents: Chemists Develop Environmentally Friendly Method for Synthesising Aniline Derivatives
An international team of researchers, including chemists from HSE University and the A.N. Nesmeyanov Institute of Organoelement Compounds of the Russian Academy of Sciences (INEOS RAS), has developed a new method for synthesising aniline derivatives—compounds widely used in the production of medicines, dyes, and electronic materials. Instead of relying on toxic and expensive reagents, they proposed using tetrahydrofuran, which can be derived from renewable raw materials. The reaction was carried out in the presence of readily available cobalt salts and syngas. This approach reduces hazardous waste and simplifies the production process, making it more environmentally friendly. The study has been published in ChemSusChem.
How Colour Affects Pricing: Why Art Collectors Pay More for Blue
Economists from HSE University, St Petersburg State University, and the University of Florida have found which colours in abstract paintings increase their market value. An analysis of thousands of canvases sold at auctions revealed that buyers place a higher value on blue and favour bright, saturated palettes, while showing less appreciation for traditional colour schemes. The article has been published in Information Systems Frontiers.
New Method for Describing Graphene Simplifies Analysis of Nanomaterials
An international team, including scientists from HSE University, has proposed a new mathematical method to analyse the structure of graphene. The scientists demonstrated that the characteristics of a graphene lattice can be represented using a three-step random walk model of a particle. This approach allows the lattice to be described more quickly and without cumbersome calculations. The study has been published in Journal of Physics A: Mathematical and Theoretical.
Scientists Have Modelled Supercapacitor Operation at Molecular and Ionic Level
HSE scientists used supercomputer simulations to study the behaviour of ions and water molecules inside the nanopores of a supercapacitor. The results showed that even a very small amount of water alters the charge distribution inside the nanopores and influences the device’s energy storage capacity. This approach makes it possible to predict how supercapacitors behave under different electrolyte compositions and humidity conditions. The paper has been published in Electrochimica Acta. The study was supported by a grant from the Russian Science Foundation (RSF).
Designing an Accurate Reading Skills Test: Why Parallel Texts are Important in Dyslexia Diagnosis
Researchers from the HSE Centre for Language and Brain have developed a tool for accurately assessing reading skills in adults with reading impairments. It can be used, for instance, before and after sessions with a language therapist. The tool includes two texts that differ in content but are equal in complexity: participants were observed to read them at the same speed, make a similar number of errors, and understand the content to the same degree. Such parallel texts will enable more accurate diagnosis of dyslexia and better monitoring of the effectiveness of interventions aimed at addressing it. The paper has been published in Educational Studies.
Internal Clock: How Heart Rate and Emotions Shape Our Perception of Time
Our perception of time depends on heart rate—this is the conclusion reached by neuroscientists at HSE University. In their experiment, volunteers watched short videos designed to evoke specific emotions and estimated each video's duration, while researchers recorded their heart activity using ECG. The study found that the slower a participant's heart rate, the shorter they perceived the video to be—especially when watching unpleasant content. The study has been published in Frontiers in Psychology.