A risk measurement method and device of an asset allocation model in an uncertain environment

Abstract

The application provides a risk measurement method and device for an asset allocation model, the method comprising: obtaining asset historical data and generating multi-asset yield data; constructing a multivariate stochastic process model to obtain a yield uncertainty parameter set; determining a market state set based on the historical data and constructing a hidden Markov model to generate a state random transition process; performing Monte Carlo simulation under the constraints of the parameter set and the state transition process to obtain multi-scenario multi-asset yield rate samples, and generating a multi-scenario portfolio yield sequence by weighting and summarizing based on candidate asset weights; performing risk and return quantitative evaluation based on the Kelly formula, and constructing a multi-objective optimization criterion combined with a risk preference parameter, and outputting a risk measurement result by iteratively updating asset weights through an adaptive optimization algorithm. The technical scheme of the application realizes risk measurement of asset allocation in an uncertain environment, and improves the robustness of risk and return evaluation and the adaptability of weight updating.

Description

Technical Field

[0001] This invention belongs to the fields of asset management and artificial intelligence technology, and specifically relates to a risk measurement method and device for an asset allocation model. Background Technology

[0002] Asset allocation is a fundamental aspect of financial investment management. A common practice is to determine the allocation ratio of various assets based on historical returns and risk information to achieve a balance between returns and risks. However, with the increasing complexity of the macroeconomic environment and market structures, asset returns are exhibiting characteristics such as increased volatility, changing correlations over time, and frequent extreme events. This makes accurately measuring portfolio risk under uncertain market conditions a key issue in asset allocation modeling and strategy evaluation.

[0003] In existing technologies, typical risk measurement and allocation frameworks include mean-variance models, risk budgeting, and minimum variance portfolios. These methods typically use historical statistics (such as mean, variance, and covariance) as the main inputs to construct optimization objectives and constraints. While these methods are simple to implement, when market volatility intensifies or related structures change rapidly, fixed statistics often fail to stably characterize the return distribution and correlations, leading to risk measurement bias and unstable allocation results.

[0004] To improve the ability to characterize uncertainty, asset allocation strategies incorporate approaches such as the Black-Litterman model, Copula-related modeling, scenario analysis, and simulation evaluation. These methods use richer prior information or distribution structures to describe the dependencies between assets and conduct scenario-based tests on portfolio performance. However, these models are primarily based on deterministic assumptions, relying on historical data and fixed parameters for asset allocation. These traditional models cannot dynamically and promptly respond to market changes. Asset allocation becomes unstable when facing market fluctuations and various uncertainties (such as market sentiment, policy changes, and economic cycles), resulting in insufficient adaptability and flexibility in practical applications. Consequently, they fail to achieve the expected return and risk control objectives, leading to reduced returns.

[0005] Therefore, there is an urgent need for a risk measurement method for asset allocation models to improve the robustness of risk measurement results and the consistency of asset weight update processes in uncertain market environments. Summary of the Invention

[0006] This invention provides a risk measurement method for an asset allocation model. The scheme of this invention constructs a risk measurement link that includes multi-asset return uncertainty modeling and random market state transitions. In the multi-scenario simulation process, Kelly quantification assessment and risk preference multi-objective optimization criteria are introduced to iteratively update the weights of candidate assets and output the risk measurement results. This solves the technical problems that the distribution of asset returns and state transitions are difficult to uniformly characterize in an uncertain market environment, that risk-return assessment and weight updates are difficult to form a consistent calculation chain, and that risk preferences are difficult to incorporate into the weight update process.

[0007] A first aspect of the present invention provides a risk measurement method for an asset allocation model, the method comprising: Obtain historical data of the asset to be configured, and generate multi-asset return data based on the historical data of the asset. Based on the multi-asset return data, a multivariate stochastic process model for characterizing return uncertainty is constructed, and a set of return uncertainty parameters is obtained. Based on the historical asset data, a set of market states is determined, a hidden Markov model is constructed to describe the evolution of market states, and a random transition process of market states is generated. Based on the set of return uncertainty parameters output by the multivariate stochastic process model and the stochastic transition process of the market state, a multi-scenario multi-asset return sample sequence is generated by combining Monte Carlo simulation; the multi-scenario multi-asset return sample sequence is weighted and summarized based on the candidate asset weights to generate a multi-scenario portfolio return sequence; the risk and return of the multi-scenario portfolio return sequence is quantitatively evaluated based on the Kelly formula to obtain the portfolio risk and return quantitative evaluation result. Obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, use an adaptive optimization algorithm to iteratively update the weights of candidate assets under the constraints of the multi-objective optimization criterion, generate updated asset weights, and generate risk measurement results corresponding to the updated asset weights based on the updated asset weights.

[0008] The risk measurement results may include: updated asset weights, quantitative assessment results of portfolio risk and return, and a set of risk indicators (such as maximum drawdown, Sharpe ratio, Kalmar ratio, etc.) obtained based on the updated asset weights. In other embodiments, the risk measurement results may also include outputs expressed in the form of risk level, risk score, or risk exposure vector to meet the interface requirements of different asset allocation application scenarios for risk measurement output.

[0009] By adopting the above scheme, the risk measurement method of the asset allocation model of the present invention generates multi-asset return data based on historical asset data and constructs a multivariate stochastic process model to form a set of return uncertainty parameters, thereby achieving a quantitative characterization of the uncertainty of multi-asset returns; it determines a set of market states based on historical asset data and constructs a hidden Markov model to generate a stochastic transition process of market states, thereby achieving a stochastic evolution characterization of market state switching; it performs Monte Carlo simulation under the constraints of the return uncertainty parameter set and the stochastic transition process of market states to generate a multi-scenario portfolio return sequence, thereby achieving unified sampling and aggregation of multi-scenario return distributions; it performs a quantitative risk-return assessment of the multi-scenario portfolio return sequence based on the Kelly formula, thereby achieving a comparable quantitative output of risk and return; furthermore, it introduces a risk preference parameter to construct a multi-objective optimization criterion that includes return and risk objectives, and uses an adaptive optimization algorithm to iteratively update the weights of candidate assets, thereby achieving adaptive weight updates under risk preference constraints, significantly improving the consistency and robustness of risk measurement results and asset weight updates under uncertain environments.

[0010] In some embodiments of the present invention, generating multi-asset return data based on the historical asset data includes: performing serialization processing on the historical asset data to obtain multi-asset return data, wherein the serialization processing includes time series analysis.

[0011] In some embodiments of the present invention, constructing a multivariate stochastic process model for characterizing return uncertainty includes: estimating the parameters of the multivariate stochastic process model based on the multi-asset return data to obtain the return uncertainty parameter set, wherein the return uncertainty parameter set includes at least expected return parameters and volatility parameters.

[0012] In some embodiments of the present invention, the step of determining a set of market states based on the historical asset data, constructing a hidden Markov model to describe the evolution of market states, and generating a random transition process of market states includes: the set of market states includes at least two market states; estimating the parameters of the hidden Markov model based on the historical asset data; and generating the random transition process of market states based on the parameters of the hidden Markov model.

[0013] In some embodiments of the present invention, the risk-return quantification assessment includes: calculating the Kelly evaluation value corresponding to the candidate asset weights using the Kelly formula based on the multi-scenario portfolio return sequence, and using the Kelly evaluation value as part of the portfolio risk-return quantification assessment result.

[0014] In some embodiments of the present invention, the risk-return quantification assessment further includes: calculating one or more of the maximum drawdown, Sharpe ratio, and Kalmar ratio based on the multi-scenario portfolio return sequence, and combining the calculation results with the Kelly criterion to form the portfolio risk-return quantification assessment result.

[0015] In some embodiments of the present invention, generating a multi-scenario combined benefit sequence includes: repeatedly generating a multi-scenario combined benefit sequence under different random states, and performing parameter sensitivity analysis based on the repeatedly generated multi-scenario combined benefit sequence.

[0016] In some embodiments of the present invention, the multi-objective optimization criterion includes at least a return objective, a risk objective, a liquidity risk objective, and / or a tail risk objective.

[0017] In some embodiments of the present invention, the iterative update of candidate asset weights includes: determining fitness evaluation results according to the multi-objective optimization criterion, and performing selection, crossover, and mutation operations based on the fitness evaluation results, wherein the crossover probability and / or mutation probability are adaptively adjusted according to the convergence state of the iterative update process.

[0018] Compared with existing technologies, the advantages of this invention are as follows: It obtains multi-asset return data by performing serialization processing on historical asset data, thus achieving a unified data entry point for the risk measurement chain; it achieves parameterizable expression of return uncertainty by estimating parameters of a multivariate stochastic process model and obtaining a set of return uncertainty parameters, including at least expected return and volatility; it achieves state switching modeling under a set of market states by constructing a hidden Markov model and generating a stochastic transition process of market states; and it generates multi-scenario multi-asset return samples through Monte Carlo simulation and forms a multi-scenario portfolio return sequence based on the weighted aggregation of candidate asset weights. This approach enables the reproducible construction of multi-scenario portfolio returns; it achieves a multi-dimensional characterization of risk-return assessment results through the combination of Kelly criterion with indicators such as maximum drawdown, Sharpe ratio, and Kalmar ratio; it verifies the stability of assessment results under model parameter perturbations by repeatedly generating multi-scenario portfolio return sequences and performing parameter sensitivity analysis; furthermore, by introducing liquidity risk and / or tail risk objectives into the multi-objective optimization criteria and using an adaptive optimization algorithm to update weights, it achieves a comprehensive trade-off of multiple risk constraints, significantly improving the robustness and adaptability of asset allocation risk measurement and weight updates in uncertain market environments.

[0019] A second aspect of the present invention provides a risk measurement system for an asset allocation model, comprising: The data processing module is used to acquire historical data of the assets to be configured and generate multi-asset return data based on the historical data of the assets. The return uncertainty modeling module is used to construct a multivariate stochastic process model to characterize return uncertainty based on the multi-asset return data, and output a set of return uncertainty parameters. The market state modeling module is used to determine the set of market states based on the historical asset data, construct a hidden Markov model to describe the evolution of market states, and generate a stochastic transition process of market states. The simulation evaluation module is used to perform Monte Carlo simulation based on the set of return uncertainty parameters and the random transition process of market state to generate a multi-scenario multi-asset return sample sequence; and to perform weighted summation of the multi-scenario multi-asset return sample sequence based on the candidate asset weights to generate a multi-scenario portfolio return sequence; and to perform a risk-return quantitative evaluation of the multi-scenario portfolio return sequence based on the Kelly formula to obtain the portfolio risk-return quantitative evaluation result. The constraint optimization module is used to obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, and use an adaptive optimization algorithm to iteratively update the weights of the candidate assets under the constraints of the multi-objective optimization criterion to generate updated asset weights; and output the corresponding risk measurement results based on the updated asset weights.

[0020] A third aspect of the present invention provides a risk measurement device for an asset allocation model, characterized in that the device includes a computer device, the computer device including a processor and a memory, the processor storing computer instructions, and when the computer instructions are executed, the device implements the risk measurement method of the asset allocation model.

[0021] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the text, or may be learned by practice of the invention. The objects and other advantages of the invention will become apparent from the description and the accompanying drawings.

[0022] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0024] In the attached diagram: Figure 1This is a flowchart illustrating a risk measurement method for an asset allocation model provided in an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram of a risk measurement system for an asset allocation model provided in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0027] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0028] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0029] Figure 1 This is a flowchart illustrating a risk measurement method for an asset allocation model provided in an embodiment of the present invention.

[0030] Example 1, as Figure 1 As shown, the present invention provides a risk measurement method for an asset allocation model, the method comprising the following steps: S1. Obtain historical data of the asset to be configured, and generate multi-asset return data based on the historical data of the asset. Historical asset data can come from market data databases, trading systems, or data service interfaces, and the data granularity can be daily, weekly, or higher. Multi-asset return data is generated based on this historical asset data. This multi-asset return data can be represented as a set of return sequences sorted by time, where each asset corresponds to one return sequence. All asset return sequences are aligned on the same time axis to form the multi-asset return data.

[0031] S2. Based on the multi-asset return data, construct a multivariate stochastic process model to characterize the uncertainty of returns, and obtain a set of return uncertainty parameters; Furthermore, parameter estimation is performed on the multivariate stochastic process model to obtain a set of return uncertainty parameters. This set of return uncertainty parameters may include: parameterized descriptions of the distribution patterns of returns for each asset, and parameterized descriptions of the dependencies between multiple assets, enabling the generation of random samples under the constraints of uncertainty parameters.

[0032] S3. Based on the historical asset data, determine the market state set, construct a hidden Markov model to describe the evolution of market states, and generate a random transition process of market states. The set of market states may include at least two market states (e.g., states under different volatility levels or different trend characteristics). Based on this, a Hidden Markov Model (HMM) is constructed to describe the evolution of market states, and the HMM parameters are estimated. Based on the HMM parameters, a stochastic transition process of market states is generated, thereby characterizing the stochastic switching pattern of market states over time.

[0033] S4. Based on the set of return uncertainty parameters output by the multivariate stochastic process model and the stochastic transition process of the market state, a multi-scenario multi-asset return sample sequence is generated by combining Monte Carlo simulation; the multi-scenario multi-asset return sample sequence is weighted and summarized based on the candidate asset weights to generate a multi-scenario portfolio return sequence; the risk and return of the multi-scenario portfolio return sequence is quantitatively evaluated based on the Kelly formula to obtain the portfolio risk and return quantitative evaluation result. The quantitative assessment results of portfolio risk and return can characterize the quantitative relationship between portfolio growth returns and risk exposure under different scenarios, and serve as one of the evaluation criteria for subsequent weight iteration updates.

[0034] S5. Obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, use an adaptive optimization algorithm to iteratively update the candidate asset weights under the constraints of the multi-objective optimization criterion, generate updated asset weights, and generate risk measurement results corresponding to the updated asset weights based on the updated asset weights.

[0035] The risk preference parameter can reflect the tendency to trade off between return and risk objectives. After obtaining the updated asset weights, the risk measurement results corresponding to the updated asset weights are generated, such as outputting the updated combination of risk indicators, risk exposure level or comprehensive risk score in multiple scenarios.

[0036] The risk measurement results may include: updated asset weights, quantitative assessment results of portfolio risk and return, and a set of risk indicators (such as maximum drawdown, Sharpe ratio, Kalmar ratio, etc.) obtained based on the updated asset weights. In other embodiments, the risk measurement results may also include outputs expressed in the form of risk level, risk score, or risk exposure vector to meet the interface requirements of different asset allocation application scenarios for risk measurement output.

[0037] By adopting the above scheme, the risk measurement method of the asset allocation model of the present invention generates multi-asset return data based on historical asset data and constructs a multivariate stochastic process model to form a set of return uncertainty parameters, thereby achieving a quantitative characterization of the uncertainty of multi-asset returns; it determines a set of market states based on historical asset data and constructs a hidden Markov model to generate a stochastic transition process of market states, thereby achieving a stochastic evolution characterization of market state switching; it performs Monte Carlo simulation under the constraints of the return uncertainty parameter set and the stochastic transition process of market states to generate a multi-scenario portfolio return sequence, thereby achieving unified sampling and aggregation of multi-scenario return distributions; it performs a quantitative risk-return assessment of the multi-scenario portfolio return sequence based on the Kelly formula, thereby achieving a comparable quantitative output of risk and return; furthermore, it introduces a risk preference parameter to construct a multi-objective optimization criterion that includes return and risk objectives, and uses an adaptive optimization algorithm to iteratively update the weights of candidate assets, thereby achieving adaptive weight updates under risk preference constraints, significantly improving the consistency and robustness of risk measurement results and asset weight updates under uncertain environments.

[0038] In some embodiments of the present invention, generating multi-asset return data based on the historical asset data includes: performing serialization processing on the historical asset data to obtain multi-asset return data, wherein the serialization processing includes time series analysis.

[0039] In this embodiment, the "generating multi-asset return data based on historical asset data" includes serializing the historical asset data. Serialization processing may include time series analysis, such as performing stationarity tests on historical price series, handling missing values, handling outliers, time alignment, and resampling, to form a return data structure that can be used for subsequent modeling. Through the above serialization processing, multi-asset return data is obtained.

[0040] In some embodiments of the present invention, constructing a multivariate stochastic process model for characterizing return uncertainty includes: estimating the parameters of the multivariate stochastic process model based on the multi-asset return data to obtain the return uncertainty parameter set, wherein the return uncertainty parameter set includes at least expected return parameters and volatility parameters.

[0041] In this embodiment, the parameter estimation of the multivariate stochastic process model may include calculating model parameters using methods such as maximum likelihood estimation, moment estimation, or Bayesian estimation to obtain a set of return uncertainty parameters. This set of return uncertainty parameters includes at least expected return parameters and volatility parameters. The expected return parameter characterizes the central tendency of the return level, while the volatility parameter characterizes the intensity of return volatility, thus providing parameter constraints for scenario sampling.

[0042] In this embodiment, a stochastic model of asset returns is constructed based on historical data, combining time series analysis and multivariate stochastic processes, to estimate the return distribution of different asset classes under different market conditions, thereby describing the uncertain financial market environment. The asset return rate is set as... Then the multivariate stochastic model is expressed as:

[0043] in, The expected rate of return on an asset. For asset volatility, This is standard Brownian motion.

[0044] The method in this embodiment simulates stochastic processes in the market, more realistically reflecting the return performance of assets under different market conditions. This method can not only handle general fluctuations, but also cope with situations with large or extreme market fluctuations. It solves the problem of overly idealistic assumptions in traditional models, enabling asset allocation models to cope with complex and dynamic market environments and improving the risk prediction capability in uncertain markets.

[0045] In some embodiments of the present invention, the step of determining a set of market states based on the historical asset data, constructing a hidden Markov model to describe the evolution of market states, and generating a random transition process of market states includes: the set of market states includes at least two market states; estimating the parameters of the hidden Markov model based on the historical asset data; and generating the random transition process of market states based on the parameters of the hidden Markov model.

[0046] In this embodiment, the market state set includes at least two market states. The state division can be obtained by clustering or segmenting based on the statistical characteristics or market characteristics (such as volatility level, trend strength, or drawdown characteristics) of historical asset data.

[0047] Based on this, the parameters of the Hidden Markov Model are estimated using historical asset data. These parameters may include parameters related to state transition probabilities and distribution parameters associated with state observations. A stochastic transition process of market states is then generated based on these Hidden Markov Model parameters to drive the state evolution over time during scenario generation.

[0048] In this embodiment, a Hidden Markov Model is used to construct a stochastic transition process of market state based on the above results. This process monitors the changes in the state of different assets in the market, such as changes in rise, fall, and oscillation, in order to prepare for the next step of return-risk assessment.

[0049] The technical solution of this invention introduces a Hidden Markov Model (HMM) to identify the transition process between different market states, track market trends in real time, and provide more accurate input data for risk and return assessment. The HMM can predict future market transition states, allowing investors to adjust their strategies in advance based on market state predictions, thereby avoiding losses due to sudden market changes. This is more adaptable and forward-looking than existing static models. This model solves the static problem of traditional models, enabling asset allocation strategies to dynamically respond to market changes and improving the stability and profitability of the investment portfolio during market fluctuations.

[0050] In some embodiments of the present invention, the risk-return quantification assessment includes: calculating the Kelly evaluation value corresponding to the candidate asset weights using the Kelly formula based on the multi-scenario portfolio return sequence, and using the Kelly evaluation value as part of the portfolio risk-return quantification assessment result.

[0051] In this embodiment, the Kelly criterion can be used to characterize the quantitative performance of portfolio growth returns under multi-scenario return-driven conditions, and the Kelly criterion is used as part of the portfolio risk-return quantitative assessment results to provide calculable input for subsequent multi-objective optimization criteria.

[0052] In some embodiments of the present invention, the risk-return quantification assessment further includes: calculating one or more of the maximum drawdown, Sharpe ratio, and Kalmar ratio based on the multi-scenario portfolio return sequence, and combining the calculation results with the Kelly criterion to form the portfolio risk-return quantification assessment result.

[0053] In this embodiment, based on the Kelly rating, the calculation results of the multi-scenario combined return sequence are combined with the Kelly rating to form a quantitative assessment result of combined risk and return, so as to simultaneously characterize the growth return performance and risk exposure characteristics under the same evaluation framework.

[0054] Based on the constructed stochastic transfer process environment, the Kelly criterion is used to evaluate the risk-reward ratio of different asset classes, and the Sharpe ratio, Kalmar ratio, maximum drawdown and other indicators of the strategy are calculated to quantify the risk performance of the investment strategy in an uncertain market.

[0055] In some embodiments of the present invention, generating a multi-scenario combined benefit sequence includes: repeatedly generating a multi-scenario combined benefit sequence under different random states, and performing parameter sensitivity analysis based on the repeatedly generated multi-scenario combined benefit sequence.

[0056] In this embodiment, parameter sensitivity analysis may include: changing the values ​​of some parameters in the set of return uncertainty parameters, changing the values ​​of some parameters in the state transition process, or changing the settings such as the number of scenarios / time span, and observing the changing patterns of the portfolio risk-return quantitative assessment results to test the stability of the risk measurement output on the disturbance of key parameters.

[0057] Specifically, in the process of quantifying the risk performance of investment strategies in uncertain markets, Monte Carlo simulation methods are incorporated to repeatedly test and evaluate various stochastic states, ensuring the maintenance of an optimal risk-reward ratio in uncertain market environments. Furthermore, the parameter sensitivity of the strategy is analyzed, and the effectiveness and robustness of the model are further verified by simulating asset returns and risks under different market scenarios. When using Monte Carlo simulation methods to test and evaluate the risk of the portfolio, the total return of the portfolio is set as... The evaluation method can be expressed as:

[0058] in, For the first The weight of each asset in the portfolio, This represents the number of asset classes.

[0059] Finally, based on the calculated risk-reward ratio and the total return of the portfolio, and combined with a custom risk preference parameter, the asset weights are adjusted in real time for estimation, optimizing the portfolio to obtain the best risk-reward ratio. The objective optimization function can be expressed as:

[0060] in, This represents the variance of the portfolio's returns.

[0061] The technical solution in this embodiment uses the Kelly criterion to calculate the optimal investment ratio in an uncertain market, balancing returns and risks. Combined with Monte Carlo simulation, it repeatedly tests various future market conditions, providing more reliable risk assessment results.

[0062] Monte Carlo simulations can predict asset returns and risks under different future market scenarios through random sampling. Compared with existing methods that rely on single-market forecasts, this invention can provide a more robust risk assessment based on a large amount of simulation data.

[0063] By combining the Kelly Criterion with Monte Carlo simulation, this invention achieves greater accuracy in risk measurement and solves the problem that traditional models struggle to handle extreme risks and sudden market events. This approach significantly enhances the ability of asset allocation models to cope with complex market conditions.

[0064] In some embodiments of the present invention, the multi-objective optimization criterion includes at least a return objective, a risk objective, a liquidity risk objective, and / or a tail risk objective.

[0065] Specifically, liquidity risk can be characterized by quantifying factors such as transaction costs, trading constraints, or turnover restrictions; tail risk can be characterized by measuring the risk of extreme return ranges. By incorporating these objectives into a multi-objective optimization criterion, multiple risk dimensions can be comprehensively weighed during the iterative update of weights.

[0066] In this embodiment, by introducing a multi-objective optimization strategy, a comprehensive asset allocation strategy optimization model is formed by measuring risks in multiple dimensions such as return volatility, liquidity risk, and tail risk.

[0067] By comprehensively considering different risk dimensions, the technical solution of this invention enables investors to construct a more balanced investment portfolio, ensuring that not only are returns optimal in uncertain markets, but risks are also controlled within a reasonable range.

[0068] This strategy addresses the shortcomings of traditional models that focus on a single objective, enhancing the comprehensiveness and robustness of asset allocation and making the portfolio perform more resilient when facing uncertain markets.

[0069] In some embodiments of the present invention, the iterative update of candidate asset weights includes: determining fitness evaluation results according to the multi-objective optimization criterion, and performing selection, crossover, and mutation operations based on the fitness evaluation results, wherein the crossover probability and / or mutation probability are adaptively adjusted according to the convergence state of the iterative update process.

[0070] For example, when fitness improvement slows down, the mutation intensity is increased to enhance search diversity; when fitness continues to improve, the selection pressure is increased to accelerate convergence, thereby achieving an adaptive balance between exploration and exploitation in the weight update process.

[0071] In this embodiment, by optimizing asset allocation weights through an adaptive genetic algorithm, dynamic optimization of the asset portfolio can be achieved under different market conditions, thereby achieving the optimal risk-reward ratio.

[0072] Genetic algorithms can continuously adjust and optimize asset allocation by simulating the process of evolution. When market conditions change drastically, adaptive algorithms can adjust the weight of the portfolio in a timely manner to prevent excessive risk exposure caused by market uncertainty.

[0073] The technical solution of this invention greatly improves the flexibility and dynamic adjustment capability of asset allocation, overcomes the decision lag problem caused by the inability of traditional models to optimize in real time, and thus effectively improves the stability of asset management returns and risk control capabilities.

[0074] Compared with the prior art, the beneficial effects of the present invention are as follows: The technical solution of this invention constructs a multivariate stochastic process model and parametrically characterizes the uncertainties of multi-asset returns, achieving joint modeling of the correlation and dynamic changes among assets. This results in a return profile that more closely resembles the true distribution under different market fluctuations and related structural changes, significantly improving the comprehensiveness and objectivity of asset return assessment. By introducing the Kelly formula into the quantitative assessment of risk and return, and combining Monte Carlo simulation to generate and statistically evaluate future returns multiple times under different scenarios, a stable quantitative output of the portfolio return-risk relationship under uncertain conditions is achieved, further obtaining more reliable risk measurement results and significantly improving the robustness and repeatability of risk-return assessment. By employing an adaptive optimization algorithm to iteratively update and dynamically adjust asset weights, the portfolio's adaptive allocation update is achieved under different market conditions, further optimizing the portfolio's risk-return ratio and significantly improving the consistency between the weight update process and the risk assessment results. By constructing a multi-objective optimization criterion based on real-time market state estimation and introducing multi-dimensional risk measurements such as return volatility, liquidity risk, and tail risk, dynamic optimization and comprehensive trade-offs of the asset portfolio are achieved, significantly improving the overall return performance and risk constraint adaptability of asset allocation in complex and uncertain markets. The technical solution of this invention enables risk measurement under uncertain environments, and improves the reliability of risk-return assessment and the adaptability of asset weight updates.

[0075] Figure 2 This is a flowchart illustrating a risk measurement system for an asset allocation model provided in an embodiment of the present invention.

[0076] Example 2, as Figure 2 As shown, the present invention also provides a risk measurement system for an asset allocation model, comprising: a data processing module S11, a return uncertainty modeling module S12, a market state modeling module S13, a simulation evaluation module S14, and a constraint optimization module S15.

[0077] The data processing module is used to acquire historical data of the assets to be configured and generate multi-asset return data based on the historical data of the assets. The return uncertainty modeling module is used to construct a multivariate stochastic process model to characterize return uncertainty based on the multi-asset return data, and output a set of return uncertainty parameters. The market state modeling module is used to determine the set of market states based on the historical asset data, construct a hidden Markov model to describe the evolution of market states, and generate a stochastic transition process of market states. The simulation evaluation module is used to perform Monte Carlo simulation based on the set of return uncertainty parameters and the random transition process of market state to generate a multi-scenario multi-asset return sample sequence; and to perform weighted summation of the multi-scenario multi-asset return sample sequence based on the candidate asset weights to generate a multi-scenario portfolio return sequence; and to perform a risk-return quantitative evaluation of the multi-scenario portfolio return sequence based on the Kelly formula to obtain the portfolio risk-return quantitative evaluation result. The constraint optimization module is used to obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, and use an adaptive optimization algorithm to iteratively update the weights of the candidate assets under the constraints of the multi-objective optimization criterion to generate updated asset weights; and output the corresponding risk measurement results based on the updated asset weights.

[0078] Example 3: The present invention also provides a risk measurement device for an asset allocation model. The device includes a computer device, which includes a processor and a memory. The processor stores computer instructions. When the computer instructions are executed, the device implements the risk measurement method for the asset allocation model.

[0079] Example 4, as Figure 3 As shown, the present invention also provides an electronic device 100 for implementing a risk measurement method for an asset allocation model.

[0080] The electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on at least one processor 102, and at least one communication bus 104.

[0081] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the risk measurement method of the asset allocation model described in the first aspect of the present invention by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0082] The memory 101 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created based on the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0083] At least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 102 may be a microprocessor or any conventional processor. Processor 102 is the control center of electronic device 100, connecting various parts of electronic device 100 via various interfaces and lines.

[0084] The memory 101 in the electronic device 100 stores multiple instructions to implement a risk measurement method for an asset allocation model, and the processor 102 can execute multiple instructions to achieve the following: Obtain historical data of the asset to be configured, and generate multi-asset return data based on the historical data of the asset. Based on the multi-asset return data, a multivariate stochastic process model for characterizing return uncertainty is constructed, and a set of return uncertainty parameters is obtained. Based on the historical asset data, a set of market states is determined, a hidden Markov model is constructed to describe the evolution of market states, and a random transition process of market states is generated. Based on the set of return uncertainty parameters output by the multivariate stochastic process model and the stochastic transition process of the market state, a multi-scenario multi-asset return sample sequence is generated by combining Monte Carlo simulation; the multi-scenario multi-asset return sample sequence is weighted and summarized based on the candidate asset weights to generate a multi-scenario portfolio return sequence; the risk and return of the multi-scenario portfolio return sequence is quantitatively evaluated based on the Kelly formula to obtain the portfolio risk and return quantitative evaluation result. Obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, use an adaptive optimization algorithm to iteratively update the weights of candidate assets under the constraints of the multi-objective optimization criterion, generate updated asset weights, and generate risk measurement results corresponding to the updated asset weights based on the updated asset weights.

[0085] Example 5: If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, and read-only memory (ROM).

[0086] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0088] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0089] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0090] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A risk measurement method for an asset allocation model, characterized in that, Includes the following steps: Obtain historical data of the asset to be configured, and generate multi-asset return data based on the historical data of the asset. Based on the multi-asset return data, a multivariate stochastic process model for characterizing return uncertainty is constructed, and a set of return uncertainty parameters is obtained. Based on the historical asset data, a set of market states is determined, a hidden Markov model is constructed to describe the evolution of market states, and a random transition process of market states is generated. Based on the set of return uncertainty parameters output by the multivariate stochastic process model and the stochastic transition process of the market state, a multi-scenario multi-asset return sample sequence is generated by combining Monte Carlo simulation; the multi-scenario multi-asset return sample sequence is weighted and summarized based on the candidate asset weights to generate a multi-scenario combined return sequence. Based on the Kelly formula, the risk-return quantitative assessment of the multi-scenario portfolio return sequence is performed to obtain the portfolio risk-return quantitative assessment result. Obtain risk preference parameters, construct a multi-objective optimization criterion including return and risk objectives based on the risk preference parameters and the combined risk-return quantitative assessment results, use an adaptive optimization algorithm to iteratively update the weights of candidate assets under the constraints of the multi-objective optimization criterion, generate updated asset weights, and generate risk measurement results corresponding to the updated asset weights based on the updated asset weights.

2. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The generation of multi-asset return data based on the historical asset data includes: The historical asset data is serialized to obtain multi-asset return data, wherein the serialization process includes time series analysis.

3. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The construction of the multivariate stochastic process model for characterizing the uncertainty of returns includes: Based on the multi-asset return data, the parameters of the multivariate stochastic process model are estimated to obtain the return uncertainty parameter set, wherein the return uncertainty parameter set includes at least the expected return parameter and the volatility parameter.

4. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The process of determining a set of market states based on the historical asset data, constructing a hidden Markov model to describe the evolution of market states, and generating a stochastic transition process for market states includes: The market state set includes at least two market states; the parameters of the hidden Markov model are estimated based on the historical asset data, and the random transition process of the market state is generated based on the parameters of the hidden Markov model.

5. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The aforementioned risk-return quantitative assessment includes: Based on the multi-scenario portfolio return sequence, the Kelly formula is used to calculate the Kelly evaluation value corresponding to the candidate asset weights, and the Kelly evaluation value is used as part of the portfolio risk-return quantitative assessment result.

6. The risk measurement method for the asset allocation model according to claim 5, characterized in that, The aforementioned risk-return quantitative assessment also includes: Based on the multi-scenario portfolio return series, calculate one or more of the maximum drawdown, Sharpe ratio, and Kalmar ratio, and combine the calculation results with the Kelly criterion to form the portfolio risk-return quantitative assessment result.

7. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The generation of the multi-scenario combined benefit sequence includes: Multi-scenario combined return sequences are repeatedly generated under different random states, and parameter sensitivity analysis is performed based on the repeatedly generated multi-scenario combined return sequences.

8. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The multi-objective optimization criteria include at least a return objective, a risk objective, a liquidity risk objective, and / or a tail risk objective.

9. The risk measurement method for the asset allocation model according to claim 1, characterized in that, The iterative update of candidate asset weights includes: The fitness evaluation result is determined according to the multi-objective optimization criterion, and selection, crossover and mutation operations are performed based on the fitness evaluation result, wherein the crossover probability and / or mutation probability are adaptively adjusted according to the convergence state of the iterative update process.

10. A risk measurement device for an asset allocation model, characterized in that, The apparatus includes a computer device, which includes a processor and a memory. The processor stores computer instructions, and when the computer instructions are executed, the apparatus implements the risk measurement method of the asset allocation model as described in any one of claims 1 to 9.

Images