Intelligent application of reserves to transactions

A method comprising:

• receiving, by one or more servers of a payment processing service and from a first computing device associated with a plurality of merchants, first transaction data associated with a plurality of transactions, wherein the first transaction data includes, for an individual transaction in the plurality of transactions, (1) an indication of whether the individual transaction was associated with a respective chargeback request, and (2) one or more characteristics of the individual transaction;
• training, by the one or more servers, a predictive model based at least in part on the first transaction data, wherein the predictive model is configured to output a value indicative of a level of risk that a given transaction will be associated with a chargeback request;
• receiving, by the one or more servers and from a second computing device associated with a merchant, second transaction data associated with a payment transaction between the merchant and a customer;
• determining, by the one or more servers, based at least in part on inputting the second transaction data to the predictive model, a particular value indicative of a particular level of risk associated with the payment transaction;
• determining, by the one or more servers, based at least in part on the particular value, to withhold an amount of funds from a settlement amount of the payment transaction;
• and
• causing, by the one or more servers, deposit of the amount of funds in a reserve account associated with the payment processing service.

One or more non-transitory computer-readable media storing instructions executable by one or more processors that, when executed by the one or more processors, cause the one or more processors to perform acts comprising:

• receiving, from a first computing device associated with a plurality of merchants, first transaction data associated with a plurality of transactions, wherein the first transaction data includes, for an individual transaction in the plurality of transactions, (1) an indication of whether the individual transaction was associated with a respective chargeback request, and (2) one or more characteristics of the individual transaction;
• training a predictive model based at least in part on the first transaction data, wherein the predictive model is configured to output a value indicative of a level of risk that a given transaction will be associated with a chargeback request;
• receiving, from a second computing device associated with a merchant, second transaction data associated with a payment transaction between the merchant and a customer;
• determining, based at least in part on inputting the second transaction data to the predictive model, a particular value indicative of a particular level of risk associated with the payment transaction;
• determining, based at least in part on the particular value, to withhold an amount of funds from a settlement amount of the payment transaction;
• and
• causing deposit of the amount of funds in a reserve account.