| Annotation: | Methods. The study used a comprehensive approach to modeling the trading behavior of participants in the MIRA digital asset market, combining address clustering, a hidden Markov model (HMM), stochastic agent policies, and a recursive model of automated market making with a constant product (CPMM). Based on historical transactions, addresses were aggregated into three behavioral clusters, for which regime-dependent policies of activity, number of transactions, and trading budgets were formed. To determine latent market states, an HMM with filtered probabilities calculated exclusively on the basis of information available before the beginning of the corresponding block was used. The number of transactions was modeled using a negative binomial distribution, trading budgets were modeled using a combination of a truncated Student-t distribution and a generalized Pareto distribution, and their distribution between individual transactions was modeled using a Dirichlet distribution. dynamics without using actual transactions of the simulated interval. Results. The results of the study showed the presence of pronounced behavioral heterogeneity of MIRA market participants, which allowed the formation of three aggregated clusters with distinct characteristics of trading activity. The feasibility of using a three-mode HMM specification to describe latent market states and including regime probabilities in the generative policies of clusters was established. The proposed model provided a consistent formation of activity, number and budgets of operations taking into account the current state of the market and pool liquidity. Recursive CPMM simulation formed a closed loop in which generated operations endogenously change reserves and price, and the obtained state is used to generate the next block. Novelty. The scientific novelty lies in the development of a regime-dependent generative model of trading behavior, which transforms empirically determined behavioral clusters of addresses into aggregated agents with their own stochastic policies. Unlike approaches that consider clustering, regime dynamics and AMM mechanics separately, their integration into a single recursive system is proposed. Practical value. The practical significance of the study lies in the possibility of using the proposed model to analyze the behavior of heterogeneous groups of participants in decentralized financial markets and assess their impact on market dynamics. The developed approach can be used to test counterfactual scenarios, assess liquidity stability and study the consequences of changing the parameters of trading policies and automated market-making mechanisms. |
| Literature: | - 1. Aspembitova, A.T., Feng, L., & Chew, L.Y. (2021). Behavioral structure of users in cryptocurrency market. PLOS ONE, Vol. 16, No. 1. Art. e0242600. DOI: 10.1371/journal.pone.0242600
- 2. Tesfatsion, L. (2006). Agent-Based Computational Economics: A Constructive Approach to Economic Theory. In: Tesfatsion L., Judd K. L. (eds.). Handbook of Computational Economics. Vol. 2: Agent- Based Computational Economics. Amsterdam : Elsevier, 2006. Pp. 831-880. DOI: 10.1016/S1574-0021(05)02016-2.
- 3. Cocco, L., Concas, G., & Marchesi, M. (2017). Using an artificial financial market for studying a cryptocurrency market. Journal of Economic Interaction and Coordination, Vol. 12, No. 2, Pp. 345-365. DOI: 10.1007/s11403-015-0168-2.
- 4. Windrum P., Fagiolo, G., & Moneta, A. (2007). Empirical validation of agent-based models: Alternatives and prospects. Journal of Artificial Societies and Social Simulation, Vol. 10, No. 2. Art. 8.
- 5. Ardia, D., Bluteau, K., & Rüede, M. (2019). Regime changes in Bitcoin GARCH volatility dynamics. Finance Research Letters, Vol. 29, Pp. 266-271. DOI: 10.1016/j.frl.2018.08.009.
- 6. Figà-Talamanca, G., Focardi, S.M., Patacca, M. (2021). Regime switches and commonalities of the cryptocurrencies asset class. Research in International Business and Finance, Vol. 57, Art. 101400. https://doi.org/10.1016/j.najef.2021.101425
- 7. Rabiner, L.R. (1989). A tutorial on hidden Markov models and selected applications in speech recognition. Proceedings of the IEEE, Vol. 77, No. 2. Pp. 257-286. DOI: 10.1109/5.18626.
- 8. Schär, F. (2021). Decentralized Finance: On Blockchain- and Smart Contract-Based Financial Markets. Federal Reserve Bank of St. Louis Review, Vol. 103, No. 2. Pp. 153-174. DOI: 10.20955/r.103.153-74.
- 9. Angeris, G., & Chitra, T. (2020). Improved techniques for evaluating constant function market makers. Proceedings from MIIM ‘20 2nd ACM Conference on Advances in Financial Technologies. (pp. 34-46). New York: ACM. DOI: 10.1145/3419614.3423257.
- 10. Cartea, Á., Drissi, F., & Monga, M. (2023). Decentralised Finance and Automated Market Making: Execution and Speculation. DOI: 10.48550/arXiv.2307.03499.
- 11. Werner, S.M., Perez, D., Gudgeon, L., Klages-Mundt, A., Harz, D., & Knottenbelt, W.J. (2021). SoK: Decentralized Finance (DeFi). DOI: 10.48550/arXiv.2101.08778.
- 12. Tarasova, T., Usatenko, O., Makurin, A., Ivanenko, V. & Cherchata, A. (2020). Accounting and features of mathematical modeling of the system to forecast cryptocurrency exchange rate. Accounting, 6(3), 357-364. DOI: 10.5267/j.ac.2020.1.003
|
Archive