Deep Learning for the Russian Equities Portfolio Management: Case-Study On Natural Resources Sector

Russian public equity markets underwent singnificant structural changes since 2022 as exodus of foreign investors opened-up floor for domestic investors. Geopolitical turbulence, more complicated information flow and other reasons might have been behind worse performance of domestic institutional investors vs. index. These developmnents coinside with rapid expansion and increased afordability of AI in finance. The structure of the Russian equities market and key market indices are heavily skewed towards natural resources equities (natural resources equities occupy around 50% of total mcap of free-float). Additionally, the Russian natural resources companies remain integrated into global commodity markets thanks to their export prices. The paper attempts to combine state-of-the-art LSTM workflow in forecasting optimal portfolio weights for Russian natural resources stocks. Our findings and framework might be instrumental for further research in the field.