<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE root>
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ali="http://www.niso.org/schemas/ali/1.0/" article-type="research-article" dtd-version="1.2" xml:lang="en"><front><journal-meta><journal-id journal-id-type="publisher-id">Economics and Mathematical Methods</journal-id><journal-title-group><journal-title xml:lang="en">Economics and Mathematical Methods</journal-title><trans-title-group xml:lang="ru"><trans-title>Экономика и математические методы</trans-title></trans-title-group></journal-title-group><issn publication-format="print">0424-7388</issn><issn publication-format="electronic">3034-6177</issn><publisher><publisher-name xml:lang="en">The Russian Academy of Sciences</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="publisher-id">697034</article-id><article-id pub-id-type="doi">10.31857/S0424738825040099</article-id><article-categories><subj-group subj-group-type="toc-heading" xml:lang="en"><subject>Mathematical analysis of economic models</subject></subj-group><subj-group subj-group-type="toc-heading" xml:lang="ru"><subject>Математический анализ экономических моделей</subject></subj-group><subj-group subj-group-type="article-type"><subject>Research Article</subject></subj-group></article-categories><title-group><article-title xml:lang="en">Quantile XGBoost and SHAP in Creating and Explaining Forecasting Models for AI Tokens</article-title><trans-title-group xml:lang="ru"><trans-title>Квантильный XGBoost и SHAP в построении и объяснении прогнозных моделей для AI-токенов</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author"><name-alternatives><name xml:lang="en"><surname>Kucherov</surname><given-names>I. I.</given-names></name><name xml:lang="ru"><surname>Кучеров</surname><given-names>И. И.</given-names></name></name-alternatives><email>unequivocally.ivan@gmail.com</email><xref ref-type="aff" rid="aff1"/></contrib></contrib-group><aff-alternatives id="aff1"><aff><institution xml:lang="en">Centre for Financial Research, Data Analytics, HSE University</institution></aff><aff><institution xml:lang="ru">Центр финансовых исследований и анализа данных НИУ ВШЭ</institution></aff></aff-alternatives><pub-date date-type="pub" iso-8601-date="2025-10-15" publication-format="electronic"><day>15</day><month>10</month><year>2025</year></pub-date><volume>61</volume><issue>4</issue><issue-title xml:lang="en">VOL 61, NO4 (2025)</issue-title><issue-title xml:lang="ru">ТОМ 61, №4 (2025)</issue-title><fpage>111</fpage><lpage>125</lpage><history><date date-type="received" iso-8601-date="2025-11-26"><day>26</day><month>11</month><year>2025</year></date></history><permissions><copyright-statement xml:lang="en">Copyright ©; 2025, Russian Academy of Sciences</copyright-statement><copyright-statement xml:lang="ru">Copyright ©; 2025, Российская академия наук</copyright-statement><copyright-year>2025</copyright-year><copyright-holder xml:lang="en">Russian Academy of Sciences</copyright-holder><copyright-holder xml:lang="ru">Российская академия наук</copyright-holder><ali:free_to_read xmlns:ali="http://www.niso.org/schemas/ali/1.0/" start_date="2026-10-15"/></permissions><self-uri xlink:href="https://journals.eco-vector.com/0424-7388/article/view/697034">https://journals.eco-vector.com/0424-7388/article/view/697034</self-uri><abstract xml:lang="en"><p>This research aims to develop a model and methodology for forecasting logarithmic returns (logreturns) for a new asset class of AI (Artificial Intelligence) tokens. To obtain one day ahead forecasts for the 0.9, 0.5, and 0.1 quantiles, the use of XGBoost quantile models is proposed, which represent an ensemble, based on gradient boosted regression trees. Quantile models have an advantage in terms of forecasting over traditionally used regression models because they allow for the estimation of not only point forecasts but also of their confidence intervals, while remaining robust to outliers. This is especially important when forming the forecasts for various cryptocurrencies’ market characteristics, which are known to be highly volatile. In addition to forecasting, the study conducts a post-forecast analysis using the SHAP (Shapley Additive explanations) method, which allows to interpret the XGBoost model, revealing key factors that are important for forecasting AI tokens’ logreturns. Based on the results of feature importance analysis with SHAP, a significant influence of AI stocks’ market characteristics, cryptocurrency market investor sentiment, seasonal fluctuations, as well as features related to the Blockchain ecosystem was identified. The paper also discusses and addresses the shortcomings of modern forecasting and post-forecasting analysis approaches for time series in general. The obtained results, in addition to academic interest, are relevant for private investors, risk managers, firms and regulators.</p></abstract><trans-abstract xml:lang="ru"><p>В статье представлена разработка модели и методологии прогнозирования логарифмических доходностей (логдоходностей) нового класса активов — AI (Artificial Intelligence) токенов. Для получения прогнозов 0,9-, 0,5- и 0,1-квантилей на один день вперед предлагается применять квантильные модели XGBoost, представляющие собой ансамбль, основанный на градиентном бустинге регрессионных деревьев. Квантильные модели имеют преимущество в прогнозировании перед традиционно используемыми регрессионными моделями, так как позволяют давать оценку не только для точечного прогноза, но и для его доверительного интервала, оставаясь при этом устойчивыми к выбросам. Это особенно важно при формировании прогнозов биржевых характеристик криптовалют, которые известны высокой волатильностью. Помимо прогнозирования, в исследовании проводится постпрогнозный анализ с применением метода SHAP (Shapley additive explanations), который позволяет интерпретировать модель XGBoost, раскрывая ключевые факторы, являющиеся важными для формирования прогнозов логдоходностей AI-токенов. По результатам анализа важности признаков c помощью SHAP выявлено значительное влияние биржевых характеристик AI-акций, сентимента инвесторов рынка криптовалют, сезонных колебаний, а также признаков, связанных с экосистемой блокчейн (Blockchain). В работе обсуждаются и корректируются недостатки современных подходов прогнозирования и постпрогнозного анализа временных рядов в целом. Полученные результаты, помимо академического интереса, являются релевантными для частных инвесторов, риск-менеджеров, компаний и регуляторов.</p></trans-abstract><kwd-group xml:lang="en"><kwd>forecasting</kwd><kwd>quantile regression</kwd><kwd>AI tokens</kwd><kwd>XGBoost</kwd><kwd>cryptocurrencies</kwd><kwd>SHAP</kwd></kwd-group><kwd-group xml:lang="ru"><kwd>прогнозирование</kwd><kwd>квантильная регрессия</kwd><kwd>AI-токены</kwd><kwd>XGBoost</kwd><kwd>криптовалюты</kwd><kwd>SHAP</kwd></kwd-group></article-meta></front><body></body><back><ref-list><ref id="B1"><label>1.</label><mixed-citation>Abdullah M., Sarker P. K., Abakah E. J. A., Tiwari A. K., Rehman M. Z. (2024). Tail risk intersection between tech-tokens and tech-stocks. Global Finance Journal, 61, 100989. DOI: 10.1016/j.gfj.2024.100989</mixed-citation></ref><ref id="B2"><label>2.</label><mixed-citation>Adekoya O. B., Oliyide J. A., Saleem O., Adeoye H. A. (2022). Asymmetric connectedness between Google-based investor attention and the fourth industrial revolution assets: The case of FinTech and Robotics, Artificial intelligence stocks. Technology in Society, 68, 101925. DOI: 10.1016/j.techsoc.2022.101925</mixed-citation></ref><ref id="B3"><label>3.</label><mixed-citation>Akpan M. (2024). Attention is all you need: An analysis of the valuation of artificial intelligence tokens. SSRN: 4993784. DOI: 10.2139/ssrn.4993784</mixed-citation></ref><ref id="B4"><label>4.</label><mixed-citation>Ali R., Xu J., Baig M. H., Rehman H. S.U., Waqas Aslam M., Qasim K. U. (2024). From data to decisions: enhancing financial forecasts with LSTM for AI token prices. Journal of Economic Studies, 51 (8), 1677–1693. DOI: 10.1108/JES-01-2024-0022</mixed-citation></ref><ref id="B5"><label>5.</label><mixed-citation>Ali S., Al-Nassar N.S., Khalid A. A., Salloum C. (2024). dynamic tail risk connectedness between artificial intelligence and fintech stocks. Annals of Operations Research, 1–35. DOI: 10.1007/s10479-024-06349-y</mixed-citation></ref><ref id="B6"><label>6.</label><mixed-citation>Ante L., Demir E. (2024). The ChatGPT effect on AI-themed cryptocurrencies. Economics and Business Letters, 13 (1), 29–38. DOI: 10.2139/ssrn.4350557</mixed-citation></ref><ref id="B7"><label>7.</label><mixed-citation>Baklanova V., Kurkin A., Teplova T. (2024). Investor sentiment and the NFT hype index: To buy or not to buy? China Finance Review International, 14 (3), 522–548. DOI: 10.1108/CFRI-06-2023-0175</mixed-citation></ref><ref id="B8"><label>8.</label><mixed-citation>Bao T., Wang X., Mahdavi N., McCarthy C., Rezazadegan D. (2023). Dynamic XGBoost-based quantile predictor for real-time electricity price forecasting. In: 2023 IEEE International Conference on Energy Technologies for Future Grids (ETFG), 1–6). IEEE. DOI: 10.1109/ETFG55873.2023.10407379</mixed-citation></ref><ref id="B9"><label>9.</label><mixed-citation>Blom H. M., Lange P. E. de, Risstad M. (2023). Estimating value-at-risk in the EURUSD currency cross from implied volatilities using machine learning methods and quantile regression. Journal of Risk and Financial Management, 16 (7), 312. DOI: 10.3390/jrfm16070312</mixed-citation></ref><ref id="B10"><label>10.</label><mixed-citation>Bonaparte Y. (2024). Artificial intelligence in finance: Valuations and opportunities. Finance Research Letters, 60, 104851. DOI: 10.1016/j.frl.2023.104851</mixed-citation></ref><ref id="B11"><label>11.</label><mixed-citation>Chen T., Guestrin C. (2016). Xgboost: A scalable tree boosting system. In: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 785–794. DOI: 10.1145/2939672.2939785</mixed-citation></ref><ref id="B12"><label>12.</label><mixed-citation>Hafid A., Ebrahim M., Rahouti M., Oliveira D. (2024). Cryptocurrency price forecasting using XGBoost regressor and technical indicators. In: 2024 IEEE international performance, computing, and communications conference (IPCCC), 1–6. IEEE. DOI: 10.1109/IPCCC59868.2024.10850357</mixed-citation></ref><ref id="B13"><label>13.</label><mixed-citation>Hossain S., Kaur G. (2024). Stock market prediction: XGBoost and LSTM comparative analysis. In: 2024 3rd International conference on artificial intelligence for internet of things (AIIoT), 1–6. IEEE. DOI: 10.1109/AIIoT58432.2024.10574794</mixed-citation></ref><ref id="B14"><label>14.</label><mixed-citation>Jabeur S. B., Mefteh-Wali S., Viviani J. L. (2024). Forecasting gold price with the XGBoost algorithm and SHAP interaction values. Annals of Operations Research, 334 (1), 679–699. DOI: 10.1007/s10479-021-04187-w</mixed-citation></ref><ref id="B15"><label>15.</label><mixed-citation>Jamieson K., Talwalkar A. (2016). Non-stochastic best arm identification and hyperparameter optimization. In: Artificial intelligence and statistics, 240–248. PMLR. DOI: 10.48550/arXiv.1502.07943</mixed-citation></ref><ref id="B16"><label>16.</label><mixed-citation>Jareño F., Yousaf I. (2023). Artificial intelligence-based tokens: Fresh evidence of connectedness with artificial intelligence-based equities. International Review of Financial Analysis, 89, 102826. DOI: 10.1016/j.irfa.2023.102826</mixed-citation></ref><ref id="B17"><label>17.</label><mixed-citation>Kumar K. S., Sree D. I., Devi P. Y., Pujitha M. V. (2024, June). Comparative analysis of LSTM and XGBoost models for short-term bitcoin price prediction. In: 2024 3rd International Conference on Applied Artificial Intelligence and Computing (ICAAIC), 932–939). IEEE. DOI: 10.1109/ICAAIC60222.2024.10575848</mixed-citation></ref><ref id="B18"><label>18.</label><mixed-citation>Lu Y. H., Lin Y. C. (2024). The determinants of voluntary disclosure: Integration of eXtreme gradient boost (XGBoost) and explainable artificial intelligence (XAI) techniques. International Review of Financial Analysis, 96, 103577. DOI: 10.1016/j.irfa.2024.103577</mixed-citation></ref><ref id="B19"><label>19.</label><mixed-citation>Lundberg S. M., Lee S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30. DOI: 10.48550/arXiv.1705.07874</mixed-citation></ref><ref id="B20"><label>20.</label><mixed-citation>Ma C. Q., Liu X., Klein T., Ren Y. S. (2025). Decoding the nexus: How fintech and ai stocks drive the future of sustainable finance. International Review of Economics &amp; Finance, 103877. DOI: 10.1016/j.iref.2025.103877</mixed-citation></ref><ref id="B21"><label>21.</label><mixed-citation>Malik F., Umar Z. (2024). Quantile connectedness of artificial intelligence tokens with the energy sector. Review of Financial Economics. John Wiley &amp; Sons, 43 (2), 135–146. DOI: 10.1002/rfe.1224</mixed-citation></ref><ref id="B22"><label>22.</label><mixed-citation>Oukhouya H., Kadiri H., El Himdi K., Guerbaz R. (2024). Forecasting international stock market trends: XGBoost, LSTM, LSTM–XGBoost, and Backtesting XGBoost models. Statistics, Optimization &amp; Information Computing, 12 (1), 200–209. DOI: 10.19139/soic-2310-5070-1822</mixed-citation></ref><ref id="B23"><label>23.</label><mixed-citation>Pedregosa F., Varoquaux G., Gramfort A., Michel V., Thirion B., Grisel O. et al. (2011). Scikit-learn: Machine learning in Python. The Journal of Machine Learning Research, 12, 2825–2830. DOI: 10.48550/arXiv.1201.0490</mixed-citation></ref><ref id="B24"><label>24.</label><mixed-citation>Shahzad U., Asl M. G., Panait M., Sarker T., Apostu S. A. (2023). Emerging interaction of artificial intelligence with basic materials and oil &amp; gas companies: A comparative look at the Islamic vs. conventional markets. Resources Policy, 80, 103197. DOI: 10.1016/j.resourpol.2022.103197</mixed-citation></ref><ref id="B25"><label>25.</label><mixed-citation>Smelyakov K., Klochko O., Dudar Z. (2023). Building quantile regression models for predicting traffic flow. COLINS, 1, 117–132. DOI: https://ceur-ws.org/Vol-3387/paper10.pdf</mixed-citation></ref><ref id="B26"><label>26.</label><mixed-citation>Somkunwar R. K., Pimpalkar A., Srivastava V. (2024). A novel approach for accurate stock market forecasting by integrating ARIMA and XGBoost. In: 2024 IEEE International students’ conference on electrical, electronics and computer science (SCEECS), 1–6. IEEE. DOI: 10.1109/SCEECS61402.2024.10481891</mixed-citation></ref><ref id="B27"><label>27.</label><mixed-citation>Trabelsi Karoui A., Sayari S., Dammak W., Jeribi A. (2024). Unveiling outperformance: A portfolio analysis of top AI-related stocks against it indices and robotics ETFs. Risks, 12 (3), 52. DOI: 10.3390/risks12030052</mixed-citation></ref><ref id="B28"><label>28.</label><mixed-citation>Vaka S., Reddy M. S., Prabhu S. N. (2024). Hybrid model for cryptocurrency price prediction using Lstm, bidirectional LSTM, and XGBoost. In: 2024 International conference on IoT based control networks and intelligent systems (ICICNIS), 925–932. IEEE. DOI: 10.1109/ICICNIS64247.2024.10823223</mixed-citation></ref><ref id="B29"><label>29.</label><mixed-citation>Xiaoyang X., Ali S., Naveed M. (2024). Artificial intelligence and big data tokens: Where cognition unites, herding patterns take flight. Research in International Business and Finance, 72, 102506. DOI: 10.1016/j.ribaf.2024.102506</mixed-citation></ref><ref id="B30"><label>30.</label><mixed-citation>Yadav M. P., Abedin M. Z., Sinha N., Arya V. (2024). Uncovering dynamic connectedness of Artificial intelligence stocks with agri-commodity market in wake of COVID-19 and Russia-Ukraine Invasion. Research in International Business and Finance, 67, 102146. DOI: 10.1016/j.ribaf.2023.102146</mixed-citation></ref><ref id="B31"><label>31.</label><mixed-citation>Yousaf I., Ijaz M. S., Umar M., Li Y. (2024). Exploring volatility interconnections between AI tokens, AI stocks, and fossil fuel markets: evidence from time and frequency-based connectedness analysis. Energy Economics, 133, 107490. DOI: 10.1016/j.eneco.2024.107490</mixed-citation></ref><ref id="B32"><label>32.</label><mixed-citation>Yousaf I., Youssef M., Goodell J. W. (2024). Tail connectedness between artificial intelligence tokens, artificial intelligence ETFs, and traditional asset classes. Journal of International Financial Markets, Institutions and Money, 91, 101929. DOI: 10.1016/j.intfin.2023.101929</mixed-citation></ref><ref id="B33"><label>33.</label><mixed-citation>Yu H., Wang L., Jiang S., Zhang C., Li J., Hu T. (2023). Ultra-short-term operating reserve requirement assessment of power system based on improved XGboost quantile regression. In: 2023 2nd Asia power and electrical technology conference (APET), 756–760. IEEE. DOI: 10.1109/APET59977.2023.10489014</mixed-citation></ref></ref-list></back></article>
