A transaction strategy simulation evaluation system based on scenario generation

Abstract

The application discloses a transaction strategy simulation evaluation system based on scenario generation, and belongs to the technical field of financial transactions. The system comprises a parameter setting module, a data acquisition module, an economic scenario generation module and a strategy evaluation module. The economic scenario generation module constructs macro, meso and micro models through hierarchical modeling, couples each model by using a joint correlation matrix, and generates multiple future market scenario paths based on Monte Carlo simulation. The strategy evaluation module executes the strategy on each path and calculates the evaluation index. The system realizes joint dynamic simulation of multi-dimensional economic variables and asset yield, solves the technical problem that traditional models are difficult to depict cross-level statistical correlation, and improves the consistency and reliability of the simulation path. Through end-to-end automatic integration, manual intervention and system interaction overhead are significantly reduced, efficient strategy simulation testing of large scale and multiple scenarios is supported, and system computing throughput and testing efficiency are improved.

Description

Technical Field

[0001] This invention belongs to the field of financial transaction technology, and specifically relates to a scenario-based transaction strategy simulation and evaluation system. Background Technology

[0002] A trading strategy is a set of rules that guides investors' buying and selling decisions in the financial market. Based on fundamental analysis, technical analysis, or quantitative modeling, it covers scenarios from short-term trading to long-term investment. Its core is to reduce the impact of subjective decision-making through logical rules and improve the consistency and efficiency of trading.

[0003] Trading strategy evaluation is a crucial step in investment. It involves analyzing a strategy's performance under historical data and hypothetical scenarios to backtest historical performance, analyze risk-adjusted returns, verify stability and robustness, optimize parameters, and predict future performance. This requires combining metrics such as net return, maximum drawdown, win rate, and Sharpe ratio, as well as scenario testing methods like stress testing and Monte Carlo simulations. Its importance lies in its role as both the core of strategy development and improvement, and a key aspect of risk management. It identifies potential problems, quantifies the risk-return balance, and enhances the long-term adaptability of the strategy, making it indispensable for optimizing investment decisions and achieving sustainable returns.

[0004] Existing evaluation methods mainly include backtesting, stress testing, and simulated trading: backtesting uses historical data to simulate the past performance of a strategy and evaluate its effectiveness and robustness; stress testing focuses on extreme market conditions and tests the strategy's ability to withstand risks; however, existing methods have obvious drawbacks—over-reliance on historical data, insufficient coverage of extreme scenarios that have not occurred before, high testing costs and long cycles, and significant impact on evaluation accuracy when data is missing. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a scenario-based trading strategy simulation and evaluation system to solve the defects of the prior art, such as reliance on historical data, incomplete coverage of extreme scenarios, high testing costs, and low efficiency.

[0006] To achieve the above objectives, the present invention employs the following technical solution: A scenario-based trading strategy simulation and evaluation system includes: The parameter setting module is used to set economic scenario parameters and trading strategy parameters; The data acquisition module is used to acquire corresponding historical market data based on the asset types involved in the trading strategy parameters; The economic scenario generation module is used to generate multiple future market scenario simulation paths based on the economic scenario parameters and the historical market data through hierarchical modeling and joint simulation; wherein each path contains a complete evolution sequence of multiple target variables within a preset time period; The hierarchical modeling includes: constructing a macroeconomic model to simulate macroeconomic variables, constructing a mesoeconomic model to simulate asset characteristics, and constructing a microeconomic model to simulate the rate of return of specific assets. The joint simulation includes: coupling the macro-model, meso-model, and micro-model through a joint correlation matrix to form a joint model, and performing path simulation based on the joint model; the joint correlation matrix is ​​used to characterize the statistical association between the variables involved in the macro-model, meso-model, and micro-model; The strategy evaluation module is used to perform simulated trading calculations based on the asset return rate sequence in each of the future market scenario simulation paths, according to the trading strategy parameters, and generate evaluation data.

[0007] A further improvement of the present invention is that: Preferably, the parameter setting module is also used to set the number of paths, time periods, and time frequency for generating the future market scenario simulation path.

[0008] Preferably, the historical market data acquired by the data acquisition module includes market data, asset characteristic data, and historical asset return rate data.

[0009] Preferably, the economic scenario generation module includes a data processing unit for preprocessing and transforming the historical market data. The preprocessing includes data integration, data alignment, filling missing values, and noise removal. The data transformation is used to extract corresponding data from the data source according to the asset type and perform data format conversion.

[0010] Preferably, the macroeconomic model is used to simulate macroeconomic variables, including interest rates, unemployment rates, and inflation rates.

[0011] The meso-level model employs a linear regression factorization method to simulate the factor characteristics of assets; The micro-model employs a factor-based time series regression method to decompose the historical asset return data into factors and predict asset returns by combining the changes in factor returns.

[0012] Preferably, the stochastic process models adapted to the micro-model, meso-model, and macro-model include at least one of the following: Wiener process model, Vasicek model, Black-Karasinski model, CIR model, and Hull-White model.

[0013] Preferably, the joint correlation matrix is ​​estimated using the Copula method or a convolution method.

[0014] Preferably, the economic scenario generation module generates the multiple future market scenario simulation paths by performing multiple Monte Carlo simulations.

[0015] Preferably, the strategy evaluation module works as follows: on each of the future market scenario simulation paths, simulated trading is performed based on the trading strategy parameters; the evaluation data generated by the strategy evaluation module includes multiple preset quantitative indicators.

[0016] Preferably, the plurality of preset quantitative indicators include at least one of net return rate, Sharpe ratio, maximum drawdown, win rate, and profit / loss ratio.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a scenario-based trading strategy simulation and evaluation system, comprising: a parameter setting module, a data acquisition module, an economic scenario generation module, and a strategy evaluation module. The economic scenario generation module constructs macro, meso, and micro models using hierarchical modeling technology, and couples these models using a joint simulation technique via a joint correlation matrix, thereby generating multiple future market scenario simulation paths based on Monte Carlo simulation. The strategy evaluation module executes the strategy and calculates evaluation indicators on each path. This method, by constructing macro, meso, and micro hierarchical models and coupling them using a joint correlation matrix, is the first to technically achieve joint simulation of different types and dimensions of macroeconomic variables, asset characteristic factors, and specific asset returns within a unified dynamic framework. This architecture overcomes the technical shortcomings of existing single models or simple combination models that cannot accurately depict the complex statistical relationships between multi-level variables, thus ensuring higher economic logic consistency and statistical reliability within the generated large number of simulation paths. The system of this invention modularly integrates parameter setting, data acquisition, hierarchical modeling, joint simulation, path generation, and strategy execution calculation, forming an end-to-end automated calculation process. This design significantly reduces the overhead of manual intervention and cross-system data exchange, making it possible to conduct large-scale multi-scenario simulation tests on complex strategies, and effectively improving the overall system's computational throughput and testing efficiency.

[0018] Furthermore, this invention can generate diverse market scenarios, including extreme scenarios never seen before, without relying on complete historical data, thus technically expanding the boundaries of strategy stress testing. Even if some historical data is missing, the system can still generate meaningful simulated paths based on the model, significantly enhancing the comprehensiveness of strategy evaluation and its resilience to anomalies.

[0019] Furthermore, the series of indicators output by the system, such as the Sharpe ratio and the statistical distribution of the maximum drawdown, are not only used for investment decisions. Their technical role lies in objectively measuring the statistical characteristics of the generated simulated path and the stability of the strategy execution engine under different simulation environments, thereby providing data basis for verifying the reliability and consistency of the system's technical solution. Attached Figure Description

[0020] Figure 1 This is a block diagram of a scenario-based trading strategy simulation and evaluation system according to the present invention; Figure 2 This is a flowchart of the system of the present invention; Figure 3 This describes the relationships between the various models in hierarchical modeling. Detailed Implementation

[0021] Hereinafter, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," "third," or "fourth" may explicitly or implicitly include one or more of that feature.

[0022] The method provided in this application can be applied to mobile phones, tablets, wearable devices, in-vehicle devices, augmented reality (AR) / virtual reality (VR) devices, laptops, and ultra-mobile personal computers. In this application, the specific type of terminal device is not limited to terminal devices such as mobile personal computers (UMPCs), netbooks, and personal digital assistants (PDAs).

[0023] It should be noted that the terms "first," "second," etc., used in the specification and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] First, the basic terms involved in this invention will be explained.

[0025] Assets refer to assets that have economic value and can be traded, including stocks, bonds, foreign exchange, and derivatives, which are usually used by investors to obtain returns or diversify risks.

[0026] Factors, which are the driving forces that affect asset prices or returns, including macroeconomic variables or asset characteristics, are often used in quantitative investment and risk models; for example, macroeconomic variables include GDP growth and interest rates, and asset characteristics include market capitalization and momentum.

[0027] Net return rate refers to the total return on investment after deducting fees, taxes, or other costs, and usually represents the actual return obtained by the investor.

[0028] Maximum drawdown is the maximum percentage loss of an investment portfolio or asset from its highest to its lowest point over a period of time, reflecting the potential downside risk faced by the investment. For example, if a fund's net asset value drops from 100 yuan to 80 yuan and then rebounds to 90 yuan, the maximum drawdown is 20%.

[0029] Win rate, the percentage of profitable trades out of total trades, is used to measure the frequency of success of a trading strategy.

[0030] Sharpe ratio: A metric that measures excess return per unit of risk, calculated using the following formula:

[0031] Used to evaluate risk-adjusted returns.

[0032] Backtesting: An evaluation method that uses historical data to simulate the past performance of a trading strategy in order to assess the effectiveness and potential risks of the strategy.

[0033] Stress testing: A risk assessment method that simulates strategy performance under extreme market conditions to evaluate risk and robustness in high-stress environments.

[0034] See Figure 1 and Figure 2 This invention discloses a scenario-based trading strategy simulation and evaluation system, comprising: The parameter setting module is used for economic scenario parameters and trading strategy parameters to be evaluated; The data acquisition module is used to acquire the corresponding historical market data for the asset types involved in the trading strategy parameters; The economic scenario generation module is used to generate multiple future market scenario simulation paths based on the economic scenario parameters and the historical market data through hierarchical modeling and joint simulation; wherein each path contains a complete evolution sequence of multiple target variables within a preset time period; The hierarchical modeling includes: constructing a macroeconomic model to simulate macroeconomic variables, constructing a mesoeconomic model to simulate asset characteristics, and constructing a microeconomic model to simulate the rate of return of specific assets. The joint simulation includes: coupling the macro-model, meso-model, and micro-model through a joint correlation matrix to form a joint model, and performing path simulation based on the joint model; the joint correlation matrix is ​​used to characterize the statistical association between the variables involved in the macro-model, meso-model, and micro-model; The strategy evaluation module is used to execute the trading strategy parameters and conduct simulated trading on each of the aforementioned future market scenario simulation paths.

[0035] In some embodiments of the present invention, the parameter setting module is used to set parameters of the economic scenario environment, the simulation path of the future market scenario, and the trading strategy.

[0036] Specifically, economic scenarios are set according to needs, including but not limited to expected macroeconomic data such as interest rates, unemployment rates, and inflation rates, as well as the time period, time frequency, and number of future paths generated for each economic scenario. The time frequency can be monthly, quarterly, semi-annually, or annually.

[0037] The trading strategy includes asset type, trading rules, trading time, and trading standards. Asset type setting refers to the type of asset selected for trading, such as stocks, funds, futures, or a combination of assets. Trading rules refer to the buying and selling rules, including how to buy and sell, stop-loss settings, etc. Trading time refers to the specific time point or time condition specified in the trading strategy parameters for conducting simulated buying and selling operations, such as performing a specific trading action at a specific moment. Trading standards refer to the criteria used to select trading targets, such as which stocks to invest in.

[0038] In some embodiments of the present invention, the historical market data acquired by the data acquisition module includes macroeconomic data, asset characteristic data, and historical asset return data. The market data includes, but is not limited to, macroeconomic data, financial product data, financial market interest rate curves, and historical data of financial market pricing factors. Asset characteristics are factors that influence quantitative investment and model risk, serving as intermediate-dimensional indicators connecting macroeconomic variables and specific asset returns. Common types include, but are not limited to: market capitalization reflecting asset size, momentum reflecting the persistence of price trends, valuation levels including price-to-earnings ratio and price-to-book ratio, profitability including ROE and net profit growth, volatility reflecting price fluctuations, and liquidity reflecting trading activity. The historical asset return data refers to records of returns related to a certain type of tradable asset over a specific period of time, generated by price changes, dividends, interest distributions, etc., primarily reflecting the actual past returns of the asset. Examples of tradable assets include stocks, bonds, foreign exchange, and derivatives.

[0039] In some embodiments of the present invention, the economic scenario generation module includes a data processing unit, a macroeconomic model unit, a mesoeconomic model unit, a microeconomic model unit, a joint simulation unit, and a path generation unit. This module first processes the data. The economic scenario generation module adopts a waterfall-style modeling approach, using a layer-by-layer modeling technique to achieve the joint modeling of multiple asset markets. Specifically, the process involves first decomposing the assets, then gradually modeling them, starting with the basic assets and then progressively modeling more factors, ultimately providing a complete asset pricing framework.

[0040] The data processing unit is used to preprocess the historical market data. This preprocessing includes data integration, data alignment, missing value imputation, and noise removal. Specifically, the data processing unit performs preprocessing and variable transformation. First, it performs preprocessing, including raw data integration, data alignment, and missing value imputation. Variable transformation is then performed based on the trading asset class. The unit automatically extracts the required data from the data source and performs data format conversion as needed. Then, it removes noise through basic filtering and uses PCA (Principal Component Analysis) to extract mutually perpendicular signal sources and signal quantities. For example, if the asset type is stocks, data is automatically retrieved from stock exchange databases and macroeconomic statistics platforms based on the stock type. This includes historical stock data, macroeconomic data, and asset pricing factor data. Historical stock data includes daily closing prices, trading volumes, price-to-earnings ratios (PE), price-to-book ratios (PB), and returns on equity (ROE) for the past 10 years. Macroeconomic data includes monthly GDP growth rates, interest rates, and inflation rates for the past 10 years. Asset pricing factor data includes historical data on market momentum factors and size factors. Format conversion: The date format is unified to "YYYY-MM-DD", and numerical values ​​are retained to two decimal places. The scattered Excel and CSV format data are converted into a matrix structure to adapt to subsequent modeling. Preprocessing: Stock data and macroeconomic data are aligned and integrated by date, missing trading volume for a particular trading day is filled in, specifically by replacing it with the average of the past 5 days, and obvious outliers are removed, such as non-adjusted data with daily stock price fluctuations exceeding 20%. Variable transformation is performed, converting the daily closing price of the stock into the daily return. Filtering and denoising: Moving average filtering is used to process the return data and remove short-term market noise. PCA signal extraction: PCA analysis is performed on macroeconomic variables and stock characteristic data, such as PE, PB, and ROE, to extract three mutually perpendicular core signal sources, specifically including macroeconomic driving signals, valuation signals, and profit signals.

[0041] Furthermore, it also includes a data storage unit to store the formatted data in the appropriate location, improving the subsequent availability of the data.

[0042] See Figure 3 The macro, meso, and micro models described above can be modeled using different process models. During the modeling process, a corresponding stochastic process model is established for each set of asset or factor data. Generally, different types of assets or factors are suited to different stochastic process models. Currently, directly supported stochastic process models include the Wiener Process, Vasicek model, Black-Karasinski model, CIR model, Hull-White model, and vectorized high-dimensional versions of these models.

[0043] Macroeconomic Model Unit: Constructing various macroeconomic models. This includes establishing a monetary interest rate model, and building upon it to create other related macroeconomic models, such as economic growth models and inflation models. The monetary interest rate model can be modeled using the Vasicek or Hull-White model; the economic growth model can be modeled using the Cobb-Douglas production function; and the inflation model can be modeled using the Phillips curve. Various macroeconomic factors in the macroeconomic models include interest rate factors, GDP growth factors, inflation factors, unemployment rate factors, and exchange rate factors.

[0044] The meso-level model unit constructs various meso-level models. These models are used to model the characteristic factors of various assets, providing input for the return prediction of the micro-level model by quantifying the dynamic evolution of asset characteristics. They adapt to the differentiated characteristics of multiple asset types, including scenario simulations supporting mixed asset trading strategies such as stocks, bonds, foreign exchange, and derivatives. The modeling methods of this meso-level model are diverse, specifically employing a factorization method based on linear regression to model these tradable assets. The program module also reserves interfaces for using other prediction methods, introducing core macroeconomic factors such as interest rates, GDP growth, inflation, unemployment rate, and exchange rates output from the macro-level model unit as inputs to the asset characteristic factors. The meso-level factors in the meso-level model are indicators that directly quantify asset characteristics, such as the price-to-earnings ratio factor, return on equity factor, momentum factor, and volatility factor for stocks; the duration factor and credit spread factor for bonds; and the interest rate spread sensitivity factor for foreign exchange.

[0045] Micro-model Unit: Establishing return models for various assets. This part mainly adopts an asset return prediction method based on asset characteristics. Time series regression is performed on the historical returns of the modeled assets to obtain factor decomposition. Then, combined with the return change process of asset characteristics, a prediction is made for the asset itself. Specifically, firstly, the standardized asset characteristic factor sequences output by the meso-level model, such as PE factor, ROE factor, and momentum factor for stocks; duration factor and credit spread factor for bonds, are organized into time series data according to the time frequency generated by the scenario. Then, historical asset return data is obtained, and regression analysis is used to quantify the contribution of asset characteristic factors to "historical returns," decomposing the core driving asset characteristic factors of returns.

[0046] Micro-level return factors in micro-models are used to decompose quantitative indicators of asset returns. For example, by performing time series regression on historical stock returns using asset characteristic factors, valuation contribution factors, momentum contribution factors, and profit contribution factors are obtained. These factors are extensions of the return dimension and are used to predict the future returns of specific assets.

[0047] In some embodiments of this invention, the joint simulation unit is used to construct a joint model: based on processed even data, it reasonably estimates the joint correlation matrix of all factors, thereby achieving a complete economic environment simulation. The estimation of the joint correlation matrix can generally be accomplished using methods such as Copula or convolution. The specific process is as follows: S301 integrates the core asset characteristic factors and factor dimensions of various micro-model units, unifies all factor data according to the time dimension, and performs standardization processing. S302 uses Pearson linear correlation coefficient and Spearman rank correlation coefficient to determine the factor linkage relationship. If the factor has a linear relationship, the covariance matrix and Gaussian copula are used to obtain the correlation relationship. If the factor has a nonlinear relationship, the t-copula or Clayton copula are used to obtain the factor linkage characteristics between N-dimensional factors.

[0048] S303 outputs an N×N dimensional factor joint correlation matrix using either the Copula method or a convolution method. The Copula method is suitable for continuous factors, while the convolution method is suitable for discrete factors.

[0049] Using the above method, a joint model including macroscopic, mesoscopic, and microscopic models is obtained, and these models are coupled and coordinated through a jointness matrix.

[0050] The path generation unit generates multiple simulated paths for future market scenarios by performing multiple Monte Carlo simulations. For example, inputting economic scenario parameters generates multiple simulated paths, each representing a set of possible future market economic environments. For instance, if the expected interest rate decreases by 5% over three years, the time frequency is monthly, the generation period is 50 periods, and 10,000 paths are generated. The final result is a data matrix of m assets * number of generation periods * number of paths, representing dynamic random paths for asset situations.

[0051] The economic scenario generation module of this invention, based on the influence relationships between models, combines the aforementioned models to reasonably estimate the joint correlation matrix of all risk factors, thereby achieving a complete economic environment simulation. It implements multiple data processing steps according to a workflow and ultimately generates a comprehensive simulation result. The factors are used to reconstruct the situation of each asset, returning a complete simulated market environment.

[0052] In some embodiments of the present invention, the specific steps of the strategy evaluation module are as follows: S401 Strategy Simulation: Based on the input trading strategy and the generated asset information, execute the trading strategy and record the trading situation for each period. S402 Indicator Calculation: This function statistically calculates relevant evaluation indicators for trading strategies under various paths, including net return rate, Sharpe ratio, maximum drawdown, win rate, and profit / loss ratio.

[0053] S403 Output: Outputs the evaluation results of the trading strategy, including the average, maximum and minimum values ​​of each indicator, and the distribution chart of each indicator; the average value of each indicator is the average of the results of multiple paths.

[0054] This invention uses an economic scenario generator to generate multiple possible future paths, simulating market economic conditions to help analyze and evaluate the performance of trading strategies under different market scenarios. The advantage of using the economic scenario generation module to evaluate trading strategies lies in its comprehensiveness and flexibility. It can generate multiple future paths by simulating the joint dynamic changes of macroeconomic and microeconomic variables, such as GDP growth, inflation rate, and interest rates, and microeconomic variables such as stock returns and credit spreads, covering different economic cycles and market scenarios. This allows for the quantification of the performance of trading strategies under diverse economic conditions, reduces testing costs, and enhances the robustness and adaptability of strategies in complex economic environments.

[0055] The following is a further explanation using a specific embodiment.

[0056] In this embodiment, the user configures economic scenario parameters and trading strategy parameters through the parameter setting module.

[0057] Economic scenario parameters: Macroeconomic variables: Expected interest rate to decline from the current 2.5% to 2.0% over the next three years; inflation rate to remain within the 2.5%-3.5% range; unemployment rate to rise from 5.0% to 6.0%. Time period: 36 months. Number of simulation paths: 10,000.

[0058] Trading Strategy Parameters: Asset Type: Stocks included in the CSI 300 Index. Trading Rules: On the first trading day of each month, buy if the stock's price-to-earnings ratio (P / E ratio) is lower than 90% of its average over the past 12 months; sell if the P / E ratio of the held stocks is higher than 110% of its average over the past 12 months. Stop-Loss Condition: Forced liquidation when the loss on a single stock exceeds 15%.

[0059] The data acquisition module automatically retrieves historical market data from the following data sources, covering the period from January 2013 to December 2022. This includes macroeconomic data: monthly GDP growth rate, interest rates, and CPI data from the National Bureau of Statistics; asset characteristic data: price-to-earnings ratio, price-to-book ratio, return on equity, and momentum factor of the CSI 300 constituent stocks; and asset return data: daily closing price data of the CSI 300 constituent stocks, used to calculate daily returns.

[0060] The data processing and modeling unit in the economic scenario generation module first performs the following preprocessing on the acquired historical data: Data integration and alignment: Macroeconomic and stock data are aligned by month using the moving average method, and missing values ​​are filled in; Variable transformation: Stock closing prices are converted into monthly returns; Standardization is performed on macroeconomic variables and asset characteristic data; Noise removal: Low-pass filtering is used to smooth the return series; PCA signal extraction: Principal component analysis is performed on the standardized macroeconomic variables and asset characteristics to extract three orthogonal signals, representing macroeconomic drivers, valuation factors, and momentum factors, respectively.

[0061] Subsequently, a hierarchical model was constructed: Macroeconomic Model: The Hull-White model is used to model interest rates, and the Cobb-Douglas production function is used to simulate GDP growth. The Phillips curve is used to simulate the inflation rate. Mesoeconomic Model: Linear regression is used to model the price-to-earnings ratio (PE), return on equity (ROE), and momentum factor, generating a standardized factor sequence. Microeconomic Model: Based on the mesoeconomic factor sequence and historical return data, time-series regression is used to decompose three microeconomic return factors: valuation contribution, earnings contribution, and momentum contribution.

[0062] The joint simulation unit constructs a joint model through the following steps: integrating time-series data of macro, meso, and micro factors and standardizing them; estimating the joint correlation matrix between factors using the Gaussian Copula method, with a dimension of 9×9; and, based on the joint model, executing 10,000 Monte Carlo simulations to generate multi-factor joint evolution paths over 36 months. Each path includes: a series of macroeconomic variables, such as interest rates, inflation rates, and unemployment rates; a series of mesoeconomic factors, such as PE, ROE, and momentum; and a series of microeconomic returns, such as valuation contributions, earnings contributions, and momentum contributions. Finally, a simulated monthly return series for each stock is synthesized.

[0063] The strategy simulation and evaluation module executes a preset trading strategy on each simulated path. Specifically, on the first trading day of each month, it performs buy and sell operations according to the PE rule, recording the profit, position changes, and whether stop-loss orders are triggered for each trade. It calculates evaluation metrics for each path, including net return rate (average annualized return of 12.5%), Sharpe ratio (average of 1.2), maximum drawdown (average of 18.3%), win rate (average of 54.7%), and profit / loss ratio (average of 1.8). The strategy evaluation module summarizes the evaluation results of 10,000 paths and outputs the following: statistical values ​​of each indicator, including average, maximum, minimum and standard deviation; distribution histogram of net return and maximum drawdown; box plot of Sharpe ratio and win rate; and economic scenario analysis report of the best and worst performing paths in the simulation.

[0064] This embodiment generates 10,000 future market paths covering diverse economic scenarios through hierarchical modeling and joint simulation, and comprehensively evaluates a stock trading strategy based on the PE factor on each path. Results show that the strategy is robust in most scenarios, but experiences significant drawdowns in extreme scenarios involving high inflation and rapidly rising interest rates. The system provides quantitative assessments and visualizations, offering a reliable basis for strategy optimization and risk management.

[0065] 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 scenario-based trading strategy simulation and evaluation system, characterized in that, include: The parameter setting module is used to set economic scenario parameters and trading strategy parameters; The data acquisition module is used to acquire corresponding historical market data based on the asset types involved in the trading strategy parameters; The economic scenario generation module is used to generate multiple future market scenario simulation paths based on the economic scenario parameters and the historical market data through hierarchical modeling and joint simulation; wherein each path contains a complete evolution sequence of multiple target variables within a preset time period; The hierarchical modeling includes: constructing a macroeconomic model to simulate macroeconomic variables, constructing a mesoeconomic model to simulate asset characteristics, and constructing a microeconomic model to simulate the rate of return of specific assets. The joint simulation includes: coupling the macro-model, meso-model, and micro-model through a joint correlation matrix to form a joint model, and performing path simulation based on the joint model; the joint correlation matrix is ​​used to characterize the statistical association between the variables involved in the macro-model, meso-model, and micro-model; The strategy evaluation module is used to perform simulated trading calculations based on the asset return rate sequence in each of the future market scenario simulation paths, according to the trading strategy parameters, and generate evaluation data.

2. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The parameter setting module is also used to set the number of paths, time periods, and time frequency for generating the future market scenario simulation path.

3. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The historical market data acquired by the data acquisition module includes market data, asset characteristic data, and historical asset return data.

4. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The economic scenario generation module includes a data processing unit for preprocessing and transforming the historical market data. The preprocessing includes data integration, data alignment, filling in missing values, and noise removal. The data transformation is used to extract corresponding data from the data source according to the asset type and perform data format conversion.

5. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The macroeconomic model is used to simulate macroeconomic variables, including interest rates, unemployment rates, and inflation rates. The meso-level model employs a linear regression factorization method to simulate the factor characteristics of assets; The micro-model employs a factor-based time series regression method to decompose the historical asset return data into factors and predict asset returns by combining the changes in factor returns.

6. The scenario-based trading strategy simulation and evaluation system according to claim 5, characterized in that, The stochastic process models adapted to the micro, meso, and macro models include at least one of the following: Wiener process model, Vasicek model, Black-Karasinski model, CIR model, and Hull-White model.

7. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The joint correlation matrix is ​​estimated using the Copula method or a convolution method.

8. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The economic scenario generation module generates multiple future market scenario simulation paths by performing multiple Monte Carlo simulations.

9. The scenario-based trading strategy simulation and evaluation system according to claim 1, characterized in that, The strategy evaluation module works as follows: on each of the future market scenario simulation paths, simulated trading is executed based on the trading strategy parameters; the evaluation data generated by the strategy evaluation module includes multiple preset quantitative indicators.

10. The scenario-based trading strategy simulation and evaluation system according to claim 9, characterized in that, The preset quantitative indicators include at least one of net return rate, Sharpe ratio, maximum drawdown, win rate, and profit / loss ratio.

Images