Optimizing interest accrual between a user's financial accounts

Abstract

Techniques are disclosed utilizing cognitive computing to improve banking experiences. A user's financial account(s) may be monitored to predict when a surplus of funds is unnecessarily present and for how long this will likely be the case. Once this is determined, techniques include automatically drafting funds from the account to another account having a higher interest rate where the funds may accrue more interest. The techniques also include predicting when an overdraft may occur and taking appropriate action when such a prediction is made. Predictions may be based upon different weighted inputs used in accordance with a predictive modeling system, which may attempt to predict for a particular user, location, and retailer, whether the user will spend an anticipated amount in excess of the user's current balance. If so, passive (e.g., notifications) and active (e.g., transferring cover funds) actions may be performed.

1 . A computer-implemented method, comprising:
 training, by one or more processors, and based on historical financial data profiles associated with a set of users, a machine learning model to determine predictive characteristics that are predictive of user spending habits indicated by the historical financial data profiles;   tracking, by the one or more processors, financial account data associated with a user, the financial account data indicating financial transactions and account balances associated with a checking account and a savings account of the user;   determining, by the one or more processors, an average daily account balance associated with the checking account based upon the financial account data;   detecting, by the one or more processors at a first time, that the checking account has a surplus by determining that a checking account balance of the checking account exceeds, at the first time, the average daily account balance by at least a threshold buffer value;   identifying, by the one or more processors, and based on the financial account data, at least one predictive characteristic of the user that is included in the predictive characteristics that are predictive of the user spending habits;   predicting, by the one or more processors, and using the trained machine learning model, forecasted checking account balances of the checking account over a future time period occurring between the first time and a second time after the first time, the one or more processors predicting the forecasted checking account balances based on:
 the at least one predictive characteristic of the user, and 
 indications in the financial account data of at least one of:
 a spending history of the user, 
 the financial transactions associated with the checking account, 
 a frequency of recurring deposits to the checking account, or 
 recurring deposit amounts associated with the recurring deposits to the checking account; 
 
   determining, by the one or more processors, and based on the forecasted checking account balances, statistical probabilities of the checking account continuing to have the surplus through different numbers of days after the first time;   determining, by the one or more processors, a particular number of days, of the different numbers of days, that has a statistical probability, of the statistical probabilities, that exceeds a threshold likelihood; and   initiating, by the one or more processors, an automatic transfer of an amount of funds, corresponding to the threshold buffer value, from the checking account to the savings account, wherein at least a portion of the amount of funds is held in the savings account for up to the particular number of days.   
     
     
         2 . The computer-implemented method of  claim 1 , wherein the checking account accrues interest at a rate that is less than that of the savings account. 
     
     
         3 . The computer-implemented method of  claim 1 , further comprising initiating, by the one or more processors, a second automatic transfer of the at least the portion of the amount of funds from the savings account to the checking account after the particular number of days have passed following transfer of the amount of funds to the savings account. 
     
     
         4 . The computer-implemented method of  claim 3 , wherein the at least the portion of the amount of funds brings the checking account balance to the average daily account balance. 
     
     
         5 . The computer-implemented method of  claim 1 , further comprising reducing, by the one or more processors, the particular number of days based on a time for the at least the portion of the amount of funds to be transferred back to the checking account and be available in the checking account. 
     
     
         6 . The computer-implemented method of  claim 1 , further comprising:
 identifying, by the one or more processors at the first time, one or more pending transactions associated with the checking account; and   determining, by the one or more processors at the first time, that the checking account has the surplus at least in part by determining that the checking account balance will, upon the one or more pending transactions being cleared, exceed the average daily account balance by at least the threshold buffer value.   
     
     
         7 . The computer-implemented method of  claim 1 , further comprising transmitting, by the one or more processors, a notification to a client device associated with the user indicating that the amount of funds has been transferred from the checking account to the savings account. 
     
     
         8 . The computer-implemented method of  claim 1 , further comprising:
 identifying, by the one or more processors, one or more pending transactions associated with the checking account;   determining, by the one or more processors, that the one or more pending transactions, upon being cleared, would result in the checking account being overdrawn; and   initiating, by the one or more processors, and before the particular number of days has passed following the automatic transfer of the amount of funds to the savings account, a second automatic transfer of the at least the portion of the amount of funds from the savings account to the checking account to prevent the checking account becoming overdrawn.   
     
     
         9 . (canceled) 
     
     
         10 . (canceled) 
     
     
         11 . A computer system comprising:
 one or more processors; and   memory storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to:
 train, based on historical financial data profiles associated with a set of users, a machine learning model to determine predictive characteristics that are predictive of user spending habits indicated by the historical financial data profiles; 
 determine demographic data associated with a user; 
 track financial account data associated with the user, the financial account data indicating financial transactions and account balances associated with a checking account and a savings account of the user; 
 determine an average daily account balance associated with the checking account based upon the financial account data; 
 detect, at a first time, that the checking account has a surplus, by determining that a checking account balance of the checking account exceeds, at the first time, the average daily account balance by at least a threshold buffer value; 
 identify, based on the financial account data and demographic data, at least one predictive characteristic of the user, wherein the at least one predictive characteristic:
 is included in the predictive characteristics that are predictive of the user spending habits, and 
 corresponds to one or more of the historical financial data profiles; 
 
 predict, using the trained machine learning model, and based on the at least one predictive characteristic of the user, forecasted checking account balances of the checking account over a future time period occurring between the first time and a second time after the first time; 
 determine, based on the forecasted checking account balances, statistical probabilities of the checking account continuing to have the surplus through different numbers of days after the first time; 
 determine a particular number of days, of the different numbers of days, that has a statistical probability, of the statistical probabilities, that exceeds a threshold likelihood; and 
 automatically transfer an amount of funds, corresponding to the threshold buffer value, from the checking account to the savings account, wherein at least a portion of the amount of funds is held in the savings account for up to the particular number of days. 
   
     
     
         12 . The computer system of  claim 11 , wherein the checking account accrues interest at a rate that is less than that of the savings account. 
     
     
         13 . The computer system of  claim 11 , wherein the computer-executable instructions further cause the one or more processors to automatically transfer the at least the portion of the amount of funds from the savings account to the checking account after the particular number of days have passed following transfer of the amount of funds to the savings account. 
     
     
         14 . The computer system of  claim 13 , wherein the at least the portion of the amount of funds brings the checking account balance to the average daily account balance. 
     
     
         15 . The computer system of  claim 11 , wherein the computer-executable instructions further cause the one or more processors to reduce the particular number of days based on a time for the at least the portion of the amount of funds to be transferred back to the checking account and be available in the checking account. 
     
     
         16 . The computer system of  claim 11 , wherein the computer-executable instructions further cause the one or more processors to:
 identify, at the first time, one or more pending transactions associated with the checking account; and   determine, at the first time, that the checking account has the surplus at least in part by determining that the checking account balance will, upon the one or more pending transactions being cleared, exceed the average daily account balance by at least the threshold buffer value.   
     
     
         17 . The computer system of  claim 11 , wherein the computer-executable instructions further cause the one or more processors to transmit a notification to a client device associated with the user indicating that the amount of funds has been transferred from the checking account to the savings account. 
     
     
         18 . The computer system of  claim 11 , wherein the computer-executable instructions further cause the one or more processors to:
 identify one or more pending transactions associated with the checking account;   determine that the one or more pending transactions, upon being cleared, would result in the checking account being overdrawn; and   automatically transfer, before the particular number of days has passed following transfer of the amount of funds to the savings account, the at least the portion of the amount of funds from the savings account to the checking account to prevent the checking account becoming overdrawn.   
     
     
         19 . (canceled) 
     
     
         20 . The computer system of  claim 11 , wherein the machine learning model predicts the forecasted checking account balances based at least in part upon at least one of:
 a spending history of the user,   the financial transactions associated with the checking account,   a frequency of recurring deposits to the checking account, or   recurring deposit amounts associated with the recurring deposits to the checking account.   
     
     
         21 . The computer system of  claim 11 , wherein the particular number of days is determined by:
 identifying a highest number of consecutive days, among the different numbers of days, that have the statistical probabilities that exceed the threshold likelihood; and   selecting the highest number of consecutive days as the particular number of days.   
     
     
         22 . The computer-implemented method of  claim 1 , wherein the particular number of days is selected by:
 identifying, by the one or more processors, a highest number of consecutive days, among the different number of days, that have the statistical probabilities that exceed the threshold likelihood; and   selecting, by the one or more processors, the highest number of consecutive days as the particular number of days.   
     
     
         23 . The computer-implemented method of  claim 1 , wherein determining the average daily account balance comprises predicting, using the machine learning model, the average daily account balance based on at least one of:
 the spending history of the user,   the financial transactions associated with the checking account,   the frequency of the recurring deposits to the checking account, or   the recurring deposit amounts associated with the recurring deposits.