Until now, three approaches to asset portfolio management have been used. The first approach is the classic one, based on the "Efficient Market Hypothesis" (EMH). The second and more modern approach is related to the "Fractal Market Hypothesis" (FMH). Modern economic practice is characterized by the presence of structurally unstable markets, included as nodes in the network of the world economy, which functions in real time. The structure of the available financial instruments is heterogeneous and non-Markovian processes arise in them. The third approach is the formation of a dynamic strategy of investment management of the asset portfolio. Due to the complex structure of the modern global financial market, the heterogeneous structure of available financial instruments and traders using different approaches and time horizons, forecasts, as a rule, require a large number of observations, work poorly at the edges of bifurcations, and do not have a computer model that could build forecasts in real time. In these structures, slow diffusion-type processes with a memory phenomenon occur, i.e., non-Markovian processes. Therefore, the formation of a dynamic strategy using modern methods of mathematical and computer modeling is very promising.
 Real financial data is expressed in rational numbers. Computational models were developed on the basis of classical substance analysis. Therefore, p-adic analysis methods are increasingly used in financial mathematical modeling. These methods are used, in particular, in the construction of neural networks, cellular automata, and percolation models. The paper seems to have taken the first step towards building a "synthetic" model of dynamic asset portfolio management. The model has the form of a differential equation in fractional derivatives obtained using the so-called interbasin kinetics method. A general form of the energy market model is also offered, as one of the specific, especially nowadays, markets.