employee
student
The article presents the results of a comparative analysis of three popular artificial intelligence chat interfaces — Deep Seek, Giga Chat, and Chat GPT — in solving the applied problem of forming a balanced investment portfolio for a short-term horizon (3-6 months). The empirical basis of the study was data on the results of securities trading on the Moscow Exchange as of January 5, 2026. During the experiment, each algorithm was given an identical task: based on the provided market data, to form a diversified portfolio with subsequent integration of derivative financial instruments. A comparative analysis of the obtained portfolio structures was carried out according to the criteria of depth of initial data analysis, detail of recommendations, logic of derivatives inclusion, practical feasibility, scientific validity, and presence of factual errors. It was found that Deep Seek demonstrated the best results, providing complete integration of initial data, advanced logic for using derivatives, and high scientific validity of the proposed structure. Giga Chat proposed a correct but template portfolio without reference to a specific market situation. Chat GPT made critical factual errors, including recommending instruments unavailable on the Russian market, and proposed a logically contradictory strategy for including derivatives. It is concluded that there is significant variability in the quality of various AI models when solving financial problems and the need to verify their recommendations taking into account current market conditions. The research results can be used by private investors and financial analysts when choosing AI tools to support investment decisions, as well as by developers when improving algorithms for the financial sector.
artificial intelligence, chat interfaces, Deep Seek, Giga Chat, ChatGPT, investment portfolio, stock market, Moscow Exchange, derivatives, comparative analysis, portfolio theory, hedging, futures, options, diversification, risk management, digital transformation, financial technology, algorithmic trading, investment efficiency
1. Budkina E. S. Transformaciya modeley investicionnogo menedzhmenta v epohu cifrovizacii //Vestnik evraziyskoy nauki. – 2024. – T. 16. – №. 6. – S. 10.
2. Hitrova E. M., Proshutinskaya S. S. Iskusstvennyy intellekt i mashinnoe obuchenie v podderzhke investicionnyh resheniy na osnove ekspertnyh dannyh //Izvestiya Baykal'skogo gosudarstvennogo universiteta. – 2025. – T. 35. – №. 2. – S. 217-227.
3. Shevchenko M. O., Belova E. Yu. Iskusstvennyy intellekt v strategicheskom upravlenii: optimizaciya investicionnyh processov v usloviyah neopredelennosti //Progressivnaya ekonomika. – 2025. – №. 11. – S. 62-89.
4. Itogi torgov cennymi bumagami. Rynochnye sdelki T+ (Osnovnaya torgovaya sessiya)/ (Bezadresnye / Order-driven) 05.01.2026 [Elektronnyy resurs]. – URL: https://www.moex.com/ru/marketdata/?ysclid=mlly0ljqn9769458561#/mode=groups&group=12&collection=72&boardgroup=9&data_type=current&category=main (data obrascheniya 02.02.2026 g.)
5. Chan F. T. i dr. Napravleniya ispol'zovaniya iskusstvennogo intellekta (II) dlya formirovaniya fondovogo portfelya // Promyshlennost'. – 2024. – T. 3. – №. 4. – S. 29-40.
6. Borodavko, L. S. Formirovanie optimal'nogo portfelya chastnogo investora v sovremennyh usloviyah / L. S. Borodavko, V. R. Sorokina // Zhurnal monetarnoy ekonomiki i menedzhmenta. – 2025. – № 10. – S. 433-440. – DOIhttps://doi.org/10.26118/2782-4586.2025.96.64.036. – EDN IIMXWO.
