Method and related device for predicting and optimizing growth of financial institution asset management size

Abstract

The application discloses a financial institution asset management scale growth prediction and optimization method and related equipment, first, customer behavior and market data are acquired; then, based on the market data, a time series analysis model is used to predict future asset return rate, and asset allocation strategies of different risk preference customer groups are optimized accordingly; then, expected maximum drawdown of each group is evaluated according to the optimized strategies, and customer churn rate is predicted in combination with the mapping relationship between the expected maximum drawdown and the customer churn rate; meanwhile, based on the predicted asset return rate, customer behavior data and risk attitude, a behavior decision model based on prospect theory is used to predict customer conversion rate; finally, the overall asset management scale growth prediction value of the financial institution is calculated by comprehensively integrating the above results, and an optimization suggestion report containing risk grouping, customer communication and product recommendation strategies is generated. The purpose of the application is to organically integrate market yield prediction, asset allocation optimization, churn rate prediction based on portfolio risk and conversion rate prediction based on customer behavior and preference, and realize result linkage and comprehensive optimization.

Description

Technical Field

[0001] This invention belongs to the field of asset management technology, specifically relating to a method and related equipment for predicting and optimizing the growth of asset management scale in financial institutions. Background Technology

[0002] As financial markets become increasingly complex and customer needs diversify, the assets under management (AUM) of financial institutions continue to grow. Traditional asset management models rely heavily on the personal experience and manual analysis of investment advisors or fund managers, making investment decisions and asset allocations based on subjective judgments of historical data and market trends. While this approach has developed a certain system over long-term practice, its evaluation process is difficult to systematize and quantify, resulting in imprecise measurements of returns and risks, limited optimization of asset allocation strategies, and difficulty in achieving the optimal balance between risk and return in dynamic markets.

[0003] To improve the scientific nature of decision-making, financial institutions have gradually introduced various financial models. For example, they use Markowitz mean-variance models and Black-Litterman models for asset allocation optimization, or utilize models such as Value at Risk (VaR) and Conditional Value at Risk (CVaR) for risk assessment. However, these models typically rely on historical data to estimate parameters, which inherently involves uncertainty. Moreover, most are built on static or deterministic assumptions, making it difficult to flexibly and promptly respond to rapid market changes, sentiment fluctuations, and macroeconomic policy adjustments. This results in insufficient stability of asset allocation strategies in actual operation, and the expected returns and risk control objectives often fail to be achieved.

[0004] Meanwhile, to reduce customer churn and improve customer conversion, some institutions have attempted to build customer churn prediction models or deploy investment decision support systems (DSS). While these technologies can analyze customer behavior characteristics or provide decision assistance to some extent, their predictive capabilities are generally limited, and they often operate independently, failing to be deeply integrated with core investment management processes such as asset allocation and return forecasting. They also lack effective mechanisms for continuous model optimization. Therefore, it is difficult to fundamentally and dynamically correlate the intrinsic causal relationship between market performance, portfolio risk (such as maximum drawdown), and customer behavior (churn, conversion), resulting in limited effectiveness in reducing churn and improving conversion rates.

[0005] In summary, existing technological solutions often focus on a single aspect of the asset management process, such as asset allocation, risk assessment, or customer behavior analysis. These aspects are relatively isolated, lacking a systematic approach that organically integrates market return forecasting, asset allocation optimization, churn rate forecasting based on portfolio risk, and conversion rate forecasting based on customer behavior and preferences, achieving synergistic and comprehensive optimization of results. This makes it difficult for financial institutions to accurately quantify and assess the comprehensive impact of different strategies on overall asset management scale, thus failing to provide decision support for the robust growth of AUM. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a method and related equipment for predicting and optimizing the growth of assets under management (AUM) for financial institutions. The purpose is to organically integrate market return prediction, asset allocation optimization, churn rate prediction based on portfolio risk, and conversion rate prediction based on customer behavior and preferences to achieve result linkage and comprehensive optimization. This enables financial institutions to accurately quantify and assess the comprehensive impact of different strategies on the overall AUM, providing decision support for the steady growth of AUM.

[0007] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: According to a first aspect of the present invention, a method for forecasting and optimizing the growth of asset management scale of financial institutions is provided, comprising: Obtain customer behavior data and market data from financial institutions; Based on the market data, a time series analysis model is used to predict future asset returns, and based on the predicted asset returns, asset allocation strategies are optimized for customer groups with different risk appetites. Based on the optimized asset allocation strategy, the expected maximum drawdown for each risk appetite customer group is assessed, and based on the mapping relationship between the expected maximum drawdown and the customer churn rate, the customer churn rate for each risk appetite customer group is predicted. Based on the predicted asset return rate, customer behavior data, and risk attitudes of various risk-preference customer groups, a prospect theory-based behavioral decision-making model is used to predict the customer conversion rate of each risk-preference customer group. Based on the optimized asset allocation strategy, the predicted customer churn rate, and the predicted customer conversion rate, the overall asset management scale growth forecast of the financial institution is calculated, and an optimization suggestion report including risk grouping strategy, customer communication strategy, and product recommendation strategy is generated.

[0008] In one possible implementation of the first aspect, the use of a time series analysis model to predict future asset returns specifically includes: The ARIMA and GARCH models were used for prediction, and the final asset return prediction was obtained by weighted summation. The weighted summation calculation is achieved through the following formula:

[0009] in, For time The projected final return on assets; and These are the predicted values ​​from the ARIMA and GARCH models, respectively. and The weighting coefficients and + =1; This is the error adjustment term.

[0010] In one possible implementation of the first aspect, the optimization of asset allocation strategies for customer groups with different risk appetites specifically includes: The customer base is divided into multiple risk levels based on a preset risk aversion coefficient; A multi-factor model combined with machine learning algorithms is used to dynamically determine the optimal asset allocation ratio for each risk level. Among them, the Expected return for each risk level Calculated using the following formula:

[0011] in, For the first The factor in the th Each risk level in time Factor values; For the first The factor in the th The weights corresponding to each risk level.

[0012] In one possible implementation of the first aspect, the mapping relationship between the expected maximum drawdown and the customer churn rate is modeled by a logical growth function; For the A risk-seeking customer group, its customer churn rate Calculated using the following formula:

[0013] in, This represents the expected maximum drawdown for the risk-averse client group. The preset maximum churn rate; For the first Natural churn rate of a risk-averse customer group; For the first The churn rate of a risk-averse customer group.

[0014] In one possible implementation of the first aspect, the use of a prospect theory-based behavioral decision-making model to predict the customer conversion rate for each risk-preference customer group specifically includes: For the A risk-averse customer group, when the expected rate of return At that time, its customer conversion rate Calculated using the following formula:

[0015] in, The scaling factor is a constant. For the first The risk aversion coefficient corresponding to each risk-seeking customer group.

[0016] In one possible implementation of the first aspect, calculating the projected growth value of the overall asset management scale of the financial institution specifically includes: Calculate the changes in assets under management brought about by the growth in existing assets, the reduction in churn, and the improvement in conversion rates, and then sum them up. The increase in assets under management brought about by the growth of existing assets Calculated using the following formula:

[0017] in, This represents the current total assets under management. For the first The asset allocation of each risk-averse customer group; For the first The projected increase in asset returns for each risk-averse customer group; The number of customers with a high risk appetite; The reduction in the amount of assets under management maintained due to the decrease in losses Calculated using the following formula: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. The expected maximum drawdown for a risk-averse client group; The increase in assets under management resulting from the transformation and improvement Calculated using the following formula: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. Expected rate of return for a risk-averse customer group; For the first A mapping function between the rate of return and conversion rate for a customer group with varying risk appetites; The projected growth value of the overall asset management scale of the financial institution is A comprehensive overview.

[0018] In one possible implementation of the first aspect, generating an optimization suggestion report that includes risk grouping strategies, customer communication strategies, and product recommendation strategies includes: Based on the predicted customer churn rate, generate differentiated customer communication strategies for customer groups with different risk preferences; Based on predicted customer conversion rates and customer behavior data, a personalized product recommendation list is generated using a recommendation algorithm. Based on the optimized asset allocation strategy, asset rebalancing recommendations are generated for client groups at different risk levels.

[0019] According to a second aspect of the present invention, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned method for predicting and optimizing the growth of asset management scale of a financial institution.

[0020] According to a third aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned method for predicting and optimizing the growth of asset management scale of a financial institution.

[0021] According to a fourth aspect of the present invention, a computer program product is provided, which, when executed by a processor, implements the aforementioned method for predicting and optimizing the growth of asset management scale in financial institutions.

[0022] Compared with the prior art, the present invention has at least the following beneficial effects: This invention presents a method for forecasting and optimizing the growth of asset management scale for financial institutions. By integrating key elements such as market return forecasting, asset allocation optimization based on forecast results, customer churn rate forecasting based on portfolio risk, and conversion rate forecasting based on customer behavior and preferences, it overcomes the limitations of existing technologies where various models operate in isolation and yield fragmented conclusions. This allows for a direct and quantitative assessment of the potential impact of asset allocation adjustments on customer churn and conversion, forming a complete analytical chain from front-end market judgment to back-end customer behavior response. Due to the tight coupling of each element, the forecast of asset management scale growth no longer relies on a single revenue estimate or simple addition or subtraction of customer numbers, but is based on the causal logic of changes in returns, changes in risk, and customer behavior responses. The growth forecast output by this solution integrates multiple contributions, including improved returns on existing customers, reduced churn, and new converted customers, making the forecast results more reflective of the actual effectiveness of market strategies. Simultaneously, it generates optimization suggestion reports including specific risk grouping configurations, differentiated customer communication, and personalized product recommendations. Financial institutions can directly adjust asset allocation strategies and launch targeted customer maintenance measures or marketing campaigns based on the report's content. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the specific embodiments of the present invention, the drawings used in the description of the specific embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a method for predicting and optimizing the growth of asset management scale in financial institutions according to the present invention; Figure 2 This is a schematic diagram of the effective frontier of the model; Figure 3 A trend chart of profitability versus customer conversion rate; Figure 4 Model optimization - AUM impact decomposition diagram; Figure 5 Maximum drawdown - churn rate trend chart; Figure 6 A diagram illustrating the holding period of wealth management products and the elements of the funds to be monitored; Figure 7 When a fund experiences losses or gains, does the user change the investment plan statistics chart? Figure 8 This is a graph of the value function. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] like Figure 1 As shown, this invention provides a method for predicting and optimizing the growth of asset management scale in financial institutions, specifically including the following steps: S1. Obtain customer behavior data and market data from financial institutions.

[0027] Specifically, customer behavior data includes customer transaction frequency, product holding duration, and query history. Market data includes historical prices, yield series, and macroeconomic indicators for various assets.

[0028] S2. Based on the market data, use a time series analysis model to predict future asset returns, and based on the predicted asset returns, optimize asset allocation strategies for customer groups with different risk preferences.

[0029] In other words, based on acquired market data, a time series analysis model is used to predict the trend of asset returns over a future period, such as the next quarter. Then, based on this predicted asset return and considering clients' risk attitudes, asset allocation strategies are optimized for client groups with different risk appetites. The optimization aims to improve expected returns while controlling risk, thereby improving the model's efficient frontier. For example,... Figure 2 As shown.

[0030] It should be noted that customers' risk attitudes are determined through risk assessment or behavioral clustering.

[0031] S3. Based on the optimized asset allocation strategy, assess the expected maximum drawdown for each risk appetite customer group, and predict the customer churn rate for each risk appetite customer group based on the mapping relationship between the expected maximum drawdown and the customer churn rate.

[0032] In other words, based on the optimized asset allocation strategy in step S2, the maximum expected drawdown of each asset portfolio under this market scenario is assessed. Then, based on the established maximum expected drawdown-customer churn rate mapping model, the customer churn rate of each risk appetite customer group is predicted at this maximum expected drawdown level.

[0033] S4. Based on the predicted asset return rate, customer behavior data, and risk attitudes of various risk-preference customer groups, a prospect theory-based behavioral decision-making model is used to predict the customer conversion rate of each risk-preference customer group.

[0034] In other words, based on the asset return rate predicted in step S2 as a reference for customer expected returns, and combined with customer behavior data obtained in step S1 and the known risk attitudes of each customer group, a prospect theory-based behavioral decision-making model is used to predict the customer conversion rate for each risk-preference customer group, i.e., the probability of additional investment or purchase of new products. The prospect theory-based behavioral decision-making model quantifies the impact of changes in return rate on customer decision-making psychology; its relationship curve is shown below. Figure 3 As shown.

[0035] S5. Based on the optimized asset allocation strategy, the predicted customer churn rate, and the predicted customer conversion rate, calculate the overall asset management scale growth forecast of the financial institution, and generate an optimization suggestion report including risk grouping strategy, customer communication strategy, and product recommendation strategy.

[0036] In other words, the asset allocation strategy optimized in step S2 is used to assess the returns on existing assets, the customer churn rate predicted in step S3, and the customer conversion rate predicted in step S4. These are then input into a comprehensive calculation model to calculate the financial institution's overall AUM growth forecast for the next stage. This calculation comprehensively considers three pathways: existing asset growth, churn reduction, and conversion improvement. Figure 4 As shown. Simultaneously, the system automatically generates an optimization suggestion report, which includes: specific asset allocation adjustment suggestions for different risk groups, i.e., risk grouping strategy optimization; differentiated customer maintenance and communication plans based on predicted churn rates, i.e., customer communication strategy generation and optimization; and a personalized product recommendation list based on predicted conversion rates and customer behavior profiles, i.e., optimized personalized recommendations.

[0037] This implementation method organically connects and optimizes the originally isolated revenue forecasting, asset allocation, churn forecasting, and conversion forecasting processes through the above-described process, forming a closed-loop AUM forecasting and optimization system that enables financial institutions to quantitatively assess and proactively manage AUM growth.

[0038] In one possible implementation, the use of a time series analysis model to predict future asset returns specifically includes: The ARIMA and GARCH models were used for prediction, and the final asset return prediction was obtained by weighted summation. The weighted summation calculation is achieved through the following formula:

[0039] in, For time The projected final return on assets; and The ARIMA model and the GARCH model, respectively, in terms of time... The predicted value; and These are the weight coefficients for the ARIMA and GARCH models, respectively. These weights are determined based on historical performance or optimization algorithms. For example, they can be dynamically adjusted based on the prediction errors of the two models over a past rolling window period, and must satisfy... + =1; This is an error adjustment term used to correct model system biases and optimize prediction results.

[0040] Specifically, to improve the robustness and accuracy of predictions, the ARIMA and GARCH models are applied in combination. The specific steps are as follows: First, the historical return series of the target asset are modeled using both the ARIMA and GARCH models, and then each model is used to predict the next time point. The rate of return, denoted as and Then, the final predicted value is obtained through weighted summation calculation.

[0041] This formula demonstrates how to improve prediction accuracy by combining the prediction results of different time series models and using a weighted average method for comprehensive analysis. This is achieved by adjusting the weights. , and the introduction of an error adjustment term This method can better capture market trends and predict asset returns, helping financial institutions optimize their asset allocation strategies, thereby maximizing returns and increasing overall asset management scale.

[0042] In one possible implementation, the optimization of asset allocation strategies for customer groups with different risk appetites specifically includes: dividing the customer group into multiple risk levels according to a preset risk aversion coefficient; dynamically determining the optimal asset allocation ratio corresponding to each risk level using a multi-factor model combined with a machine learning algorithm; wherein, the first... Each risk level in time Expected rate of return Calculated using the following formula:

[0043] in, For the first The factor in the th Each risk level in time Factor values; For the first The factor in the th The weights corresponding to each risk level. For example, factors include market risk, value factor, size factor, etc.

[0044] Specifically, the customer group is first clustered according to risk tolerance and attitude, and divided into multiple risk levels, for example, five levels from R1 (risk aversion) to R5 (risk preference). For example, the asset ratio of customers in each level is 1:2:4:2:1.

[0045] This invention optimizes asset allocation for each risk level. It combines multi-factor models with machine learning algorithms for optimization. For the [missing information - likely a specific risk level]... Each risk level, in time Expected rate of return The calculations are performed using the optimized multi-factor model described above. For example, the machine learning model employs the random forest algorithm.

[0046] Employing machine learning algorithms offers the following advantages: 1) It automatically selects the most important factors for predicting returns under the current market environment from the candidate factor pool and dynamically adjusts their weights to adapt to the specific needs of different risk groups; 2) It can capture the complex nonlinear relationship between factors and returns, making the model more accurate. Finally, through global optimization techniques such as genetic algorithms, it finds the optimal asset allocation ratio for each risk group under given constraints, achieving the best balance between risk and return.

[0047] Through the above improvements, this invention can effectively and accurately capture the asset allocation needs of different risk groups. By combining a multi-factor model with machine learning algorithms, factor weights are dynamically optimized, improving the model's adaptability. Global optimization techniques are used to find the optimal asset allocation strategy for each risk group, further enhancing overall returns. This improvement enables financial institutions to achieve significant growth in assets under management through sophisticated and intelligent asset allocation strategies in complex and volatile market environments.

[0048] In one possible implementation, the mapping between the expected maximum drawdown and the customer churn rate is modeled using a logistic growth function; for the ... A risk-seeking customer group, its customer churn rate Calculated using the following formula:

[0049] in, This represents the maximum expected drawdown for a risk-averse client group, i.e., the maximum expected drawdown of the portfolio held by that client group. The preset maximum churn rate; For the first The natural churn rate of a risk-averse customer group, i.e., the normal churn level without drawdown; For the first The churn rate of a risk-averse customer group reflects the sensitivity of the churn rate to the increase in drawdown.

[0050] Specifically, by analyzing historical data, a logistic growth function is used to model the non-linear relationship between customer churn rate and maximum portfolio drawdown. For the... A risk-seeking customer group, its customer churn rate The relationship between the maximum drawdown and the formula is as described above.

[0051] like Figure 5 As shown, the curves from R1 to R5 differ for different risk appetite groups. Customers with low risk appetite, such as R1, are more likely to be churned when faced with the same drawdown. The value is larger. During implementation, historical data needs to be used to determine the natural churn rate and churn rate of each group. The expected maximum drawdown for each group, as assessed in step S2, is then substituted into its corresponding function. This allows for the prediction of churn rate.

[0052] In one possible implementation, the prediction of customer conversion rates for each risk-preference customer group using a prospect theory-based behavioral decision-making model specifically involves: for the first... A risk-averse customer group, when the expected rate of return At that time, its customer conversion rate Calculated using the following formula:

[0053] in, The scaling factor is a constant. For the first Risk aversion coefficient corresponding to each risk-preference customer group; This refers to the projected or displayed rate of return on assets to clients. The risk aversion coefficient is a parameter reflecting an individual's or investor's degree of risk aversion. In economics and finance, the risk aversion coefficient is used to quantify an investor's risk preference when facing potential losses. A higher risk aversion coefficient indicates a greater degree of risk aversion and a stronger tendency to avoid risk. In prospect theory, the risk aversion coefficient is an important parameter describing an individual's behavioral characteristics when facing gains and losses.

[0054] Combination Figure 6 , Figure 7 and Figure 8As shown, based on prospect theory, investors are risk-averse when faced with returns. This theory is applied to conversion rate prediction. The formula describes the law that increased returns lead to increased conversion rates, but with diminishing marginal utility; the slope of the curve differs for different risk groups. Investors are all loss-averse. As returns increase, investors' marginal utility decreases, and the growth rate of customer conversion rates gradually slows down and converges. The more risk-averse a customer is, the slower the marginal utility diminishes, and the greater the marginal utility increase brought about by increased returns; that is, the conversion rate is more significantly affected by increased returns. In practice, the risk aversion coefficient can be obtained by fitting customer risk assessments or historical behavioral data. Substituting the asset return rate predicted in step S2, applicable to this group, into the formula yields the predicted conversion rate.

[0055] In one possible implementation, the calculation of the overall asset management scale growth forecast of the financial institution specifically includes: calculating the changes in asset management scale brought about by the three dimensions of stock growth, loss reduction and conversion improvement, and summing them up.

[0056] The increase in assets under management brought about by the growth of existing assets Calculated using the following formula:

[0057] in, This represents the current total assets under management. For the first The asset allocation of each risk-averse customer group; For the first The projected increase in asset returns for each risk-averse customer group; This refers to the number of customers with a high risk appetite.

[0058] Assume the company's total AUM is 10 billion; divide the customer group into 5 groups according to risk attitude, with their asset ratios being 1:2:4:2:1, as shown in Table 1 below (unit: 100 million).

[0059] Table 1

[0060] Assuming that the incremental returns for all risk customer groups are consistent after using the new model (Model B) (e.g.) Figure 2 The effective frontier of the model shifts upwards. Assuming the incremental return is 5%, that is:

[0061] We can obtain:

[0062] The reduction in the amount of assets under management maintained due to the decrease in losses The change in the model's risk coefficient leading to a change in the churn rate, and consequently the change in the total AUM, is calculated using the following formula: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. The expected maximum drawdown for each risk-averse client group, representing the values ​​under Model A and Model B respectively. The maximum expected drawdown for each risk group.

[0063] Assuming the asset allocation for clients with different risk appetites is 1:2:4:2:1, and the total AUM is 10 billion, the estimated average maximum drawdown for Model A and Model B is shown in Table 2 below. Table 2

[0064] The corresponding churn rate was calculated, as shown in Table 3 below: Table 3

[0065] The final estimated total decrease in AUM is 499 million.

[0066] Through the aforementioned model optimization strategy, this invention can significantly reduce customer churn rate. The improved risk management and churn rate prediction model can effectively reduce customer churn caused by market fluctuations and risk exposure. Furthermore, this invention can significantly improve customer satisfaction by enhancing the customer's investment experience and trust through personalized risk management and customer communication strategies, further reducing the likelihood of churn. Moreover, this invention can significantly stabilize and increase AUM; the reduction in customer churn directly leads to the maintenance and increase of AUM, while also attracting more new customers, thereby further increasing the scale of management.

[0067] The increase in assets under management resulting from the transformation and improvement The following formula calculates how changes in the model's profitability lead to changes in customer conversion rates, which in turn result in changes in AUM: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. The expected rate of return for each risk-averse customer group, representing the first and second returns under models A and B respectively. Expected rate of return for each risk group; For the first A mapping function between the rate of return and conversion rate for a risk-averse customer group.

[0068] Assuming the asset allocation for clients with different risk appetites is 1:2:4:2:1, and the total AUM is 10 billion, experiments show that increased returns lead to higher conversion rates, which in turn increases AUM. Assuming that the new model (Model B) can increase the rate of return by 5% compared to the original model (Model A), the conversion rates of customers grouped by risk level under the influence of the rate of return are shown in Table 4 below.

[0069] Table 4

[0070] Based on the conversion rate of each group of customers with different risk appetites, if AUM is 10 billion, substituting it into the formula, the AUM increment can be calculated to be 1.206 billion.

[0071] This invention improves customer conversion rates and further drives AUM growth by introducing model optimization, thereby bringing greater economic benefits to financial institutions. Experimental verification shows that applying the optimized model can significantly increase AUM. For example, assuming that AUM can increase by 1.206 billion after using the new model, it demonstrates that model optimization has a significant promoting effect on the growth of assets under management.

[0072] In one possible implementation, generating an optimized recommendation report including risk grouping strategies, customer communication strategies, and product recommendation strategies includes: generating differentiated customer communication strategies for customer groups with different risk preferences based on predicted customer churn rates; generating personalized product recommendation lists using recommendation algorithms based on predicted customer conversion rates and customer behavior data; and generating asset rebalancing recommendations for customer groups at each risk level based on optimized asset allocation strategies.

[0073] Profitability directly impacts customer conversion rates. This invention improves profitability through model optimization techniques, enabling financial institutions to more effectively identify and incentivize customers, thereby increasing customer conversion rates and achieving AUM growth. Specifically, the following are algorithmic improvement strategies for enhancing customer conversion rates: Leveraging machine learning and data analytics, we delve into customer behavioral data to identify key factors influencing conversion rates. We use clustering algorithms to segment customers based on their investment preferences, risk attitudes, and historical behavior. The characteristics of each segment help develop more precise marketing strategies. We comprehensively utilize behavioral prediction models to build predictive models to forecast customer conversion probabilities. These models can be trained based on customer behavioral data (such as transaction frequency, product holding time, etc.) to identify potential converting customers.

[0074] By providing personalized investment advice, we aim to improve customer conversion rates. At the recommendation algorithm level, we employ collaborative filtering, content recommendation, or hybrid recommendation algorithms to optimize recommendation combinations. Based on customers' historical investment behavior and preferences, we recommend suitable products or services, enhancing customer engagement and satisfaction, thereby increasing conversion rates. We also dynamically adjust recommendation strategies in real time based on customer feedback and market changes to ensure that the investment advice provided consistently aligns with customer needs and market trends.

[0075] By designing appropriate incentive mechanisms to reward customer behavior, automated marketing can be achieved. For example, new customers can be given rebates or discounts on their first investment, and products can be recommended through automated marketing methods. The effectiveness of different incentive programs can be evaluated, and the best strategy can be selected to improve customer conversion rates.

[0076] set up: Indicates customer conversion rate; This represents a vector of customer behavior characteristics (such as transaction frequency, holding time, etc.). Indicates the effectiveness of the recommendation strategy (based on incentive feedback, automated marketing, etc.); This indicates the effectiveness of communication incentive strategies (based on automated marketing and incentive mechanisms).

[0077] We obtain an optimized conversion rate formula, where, It is a comprehensive function that represents the combined impact of customer behavior characteristics, recommendation strategies, and communication incentive strategies on conversion rates:

[0078] Statistics show that yield is a significant factor influencing customer conversion rates, greatly impacting whether customers continue to hold the product. When a product incurs losses and yields reach a certain threshold, most investors will choose to adjust their investment plans to some extent.

[0079] This invention effectively enhances financial institutions' ability to assess and drive asset management scale growth. Specifically, the method first improves the accuracy of return forecasting by comprehensively applying multiple time series analysis algorithms such as ARIMA and GARCH, enabling financial institutions to more accurately grasp market trends and optimize asset allocation, thereby directly improving the asset returns of existing customers. By segmenting customers according to risk tolerance and introducing global optimization techniques such as multi-factor models optimized by machine learning and genetic algorithms, fine-tuning of asset allocation for different risk groups is achieved, effectively balancing risk and return. Addressing customer churn, this invention establishes a mapping model between maximum drawdown and churn rate, enabling institutions to predict and proactively manage churn risk caused by market fluctuations, improving customer satisfaction and reducing churn rate through personalized risk management and communication strategies. Simultaneously, based on improved returns and customer behavior analysis, combined with optimized clustering segmentation, personalized recommendation algorithms, and incentive mechanisms, this method effectively improves customer conversion rates. In summary, these improvements are not isolated but rather organically integrated by precise return forecasting, dynamic asset allocation, risk exposure-based attrition control, and behavioral analysis-based conversion enhancement. Together, they form a systematic AUM growth driver that ultimately helps financial institutions achieve robust and sustainable growth in assets under management in complex market environments.

[0080] The following is an explanation of some related terms: Assets under management (AUM): refers to the total value of client assets managed by a financial institution, including investment portfolios, funds, etc.

[0081] Asset allocation models are financial models used to determine how to allocate an investment portfolio across different asset classes, with the aim of optimizing the balance between risk and return to meet specific investment objectives.

[0082] Risk attitude refers to the degree of preference or aversion to risk when faced with uncertainty. It is usually divided into three types: risk aversion, risk neutrality, and risk preference.

[0083] Incremental returns: This refers to the increase in investment returns compared to a previous period or benchmark. In financial models, it is typically used to measure the improvement in the performance of an investment strategy or product.

[0084] Maximum drawdown: Drawdown refers to the process by which the market value of an investment portfolio falls from its peak to its trough, and is usually used to measure investment risk. Maximum drawdown is the largest decline within a given period.

[0085] Maximum churn rate: This typically refers to the highest percentage of customers who might churn under specific circumstances. In financial models, it describes the proportion of customers who might leave a financial institution in the worst-case scenario.

[0086] Natural churn rate: This refers to the rate of customer churn that occurs naturally without external intervention or market anomalies. It is typically the normal churn level caused by factors such as customer personal choices, changes in needs, or decreased satisfaction.

[0087] Churn rate of return (CRR) refers to the rate at which the number or percentage of customers churns within a given period. It is typically used to measure an accelerated rate of customer churn due to specific factors such as declining service quality, products no longer meeting needs, or increased competition. In the financial sector, CRR can be used to predict and manage customer churn risk, enabling appropriate measures to reduce customer loss.

[0088] Prospect Theory, also known as outlook theory, is a decision-making theory proposed by psychologists Daniel Kahneman and Amos Tversky. This theory describes people's decision-making behavior when faced with gains and losses, especially under risky conditions.

[0089] Risk aversion coefficient: This parameter reflects an individual's or investor's degree of risk aversion. In economics and finance, the risk aversion coefficient is used to quantify an investor's risk preference when facing potential losses. A higher risk aversion coefficient indicates a greater degree of risk aversion and a stronger tendency to avoid risk, even if it means forgoing some potential gains. In prospect theory, the risk aversion coefficient is an important parameter describing an individual's behavioral characteristics when facing gains and losses.

[0090] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor 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. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions from the computer storage medium to achieve a corresponding method flow or function. The processor described in this embodiment of the present invention can be used in the operation of a method for predicting and optimizing the growth of asset management scale in financial institutions.

[0091] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be Random Access Memory (RAM) or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the method for predicting and optimizing the growth of asset management scale of a financial institution in the above embodiments.

[0092] 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, optical storage, etc.) containing computer-usable program code.

[0093] 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.

[0094] 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.

[0095] 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.

[0096] This invention also provides a computer program product, which is used to execute any of the above-described methods for predicting and optimizing the growth of asset management scale in financial institutions. Since the computer program product provided by this invention and the above-described method for predicting and optimizing the growth of asset management scale in financial institutions belong to the same inventive concept, the computer program product provided by this invention has all the advantages of the above-described method for predicting and optimizing the growth of asset management scale in financial institutions. Therefore, the beneficial effects of the computer program product provided by this invention will not be elaborated upon here.

[0097] In this invention, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In this specification, the 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. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0098] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention.

Claims

1. A method for predicting and optimizing the growth of asset management scale of financial institutions, characterized in that, include: Obtain customer behavior data and market data from financial institutions; Based on the market data, a time series analysis model is used to predict future asset returns, and based on the predicted asset returns, asset allocation strategies are optimized for customer groups with different risk appetites. Based on the optimized asset allocation strategy, the expected maximum drawdown for each risk appetite customer group is assessed, and based on the mapping relationship between the expected maximum drawdown and the customer churn rate, the customer churn rate for each risk appetite customer group is predicted. Based on the predicted asset return rate, customer behavior data, and risk attitudes of various risk-preference customer groups, a prospect theory-based behavioral decision-making model is used to predict the customer conversion rate of each risk-preference customer group. Based on the optimized asset allocation strategy, the predicted customer churn rate, and the predicted customer conversion rate, the overall asset management scale growth forecast of the financial institution is calculated, and an optimization suggestion report including risk grouping strategy, customer communication strategy, and product recommendation strategy is generated.

2. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The method of using time series analysis models to predict future asset returns specifically includes: The ARIMA and GARCH models were used for prediction, and the final asset return prediction was obtained by weighted summation. The weighted summation calculation is achieved through the following formula: in, For time The projected final return on assets; and These are the predicted values ​​from the ARIMA and GARCH models, respectively. and The weighting coefficients and + =1; This is the error adjustment term.

3. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The optimized asset allocation strategies for client groups with different risk appetites specifically include: The customer base is divided into multiple risk levels based on a preset risk aversion coefficient; A multi-factor model combined with machine learning algorithms is used to dynamically determine the optimal asset allocation ratio for each risk level. Among them, the Expected return for each risk level Calculated using the following formula: in, For the first The factor in the th Each risk level in time Factor values; For the first The factor in the th The weights corresponding to each risk level.

4. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The mapping relationship between the expected maximum drawdown and the customer churn rate is modeled using a logical growth function; For the A risk-seeking customer group, its customer churn rate Calculated using the following formula: in, This represents the expected maximum drawdown for the risk-averse client group. The preset maximum churn rate; For the first Natural churn rate of a risk-averse customer group; For the first The churn rate of a risk-averse customer group.

5. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The method of using a prospect theory-based behavioral decision-making model to predict customer conversion rates for different risk-preference customer groups is as follows: For the A risk-averse customer group, when the expected rate of return At that time, its customer conversion rate Calculated using the following formula: in, The scaling factor is a constant. For the first The risk aversion coefficient corresponding to each risk-seeking customer group.

6. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The calculation of the overall asset management scale growth forecast of the financial institution specifically includes: Calculate the changes in assets under management brought about by the growth in existing assets, the reduction in churn, and the improvement in conversion rates, and then sum them up. The increase in assets under management brought about by the growth of existing assets Calculated using the following formula: in, This represents the current total assets under management. For the first The asset allocation of each risk-averse customer group; For the first The projected increase in asset returns for each risk-averse customer group; The number of customers with a high risk appetite; The reduction in the amount of assets under management maintained due to the decrease in losses Calculated using the following formula: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. The expected maximum drawdown for a risk-averse client group; The increase in assets under management resulting from the transformation and improvement Calculated using the following formula: [ ] in, and The first and second parts are respectively the first and second parts before and after optimization. Expected rate of return for a risk-averse customer group; For the first A mapping function between the rate of return and conversion rate for a risk-averse customer group; The projected growth value of the overall asset management scale of the financial institution is A comprehensive summary.

7. The method for predicting and optimizing the growth of asset management scale of financial institutions according to claim 1, characterized in that, The generated optimization suggestion report, which includes risk grouping strategies, customer communication strategies, and product recommendation strategies, includes: Based on the predicted customer churn rate, generate differentiated customer communication strategies for customer groups with different risk preferences; Based on predicted customer conversion rates and customer behavior data, a personalized product recommendation list is generated using a recommendation algorithm. Based on the optimized asset allocation strategy, asset rebalancing recommendations are generated for client groups at different risk levels.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a method for predicting and optimizing the growth of asset management scale of a financial institution as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a method for predicting and optimizing the growth of asset management scale of financial institutions as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, When the computer program product is executed by the processor, it implements a method for predicting and optimizing the growth of the asset management scale of a financial institution as described in any one of claims 1 to 7.

Images