Identification of loss risk candidates for financial institutions

1. A method for determining loss risks, comprising:

employing a processor to execute executable instructions stored within a memory to perform the following:

receiving client data including client activity data associated with a set of client accounts;

filtering the client activity data into a subset of the client activity data associated with a subset of client accounts, wherein the filtering includes coarse filtering of client data and refined filtering of client activity data to generate the subset of client accounts;

calculating composite activity scores of the subset of client accounts according to the filtered subset of the client activity data;

generating a composite activity ranking from calculated composite activity scores of the subset of client accounts, wherein the composite activity ranking indicates a level of risk of loss;

identifying an activity rank target profile, wherein the activity rank target profile is based at least in part upon a composite of client accounts in the subset of client accounts that have similar composite activity rankings; and

identifying client accounts in the set of client accounts that are similar to the activity rank target profile.

2. The method of claim 1 , wherein the client activity data includes data fields that describe client account actions where at least one data field relates to a propensity for a loss risk.

3. The method of claim 1 , wherein the generating the composite activity rank comprises:

receiving weights for at least one data field; and

applying the weights to the subset of client activity data.

4. The method of claim 3 , wherein the composite activity rank is based at least in part on the weighted ranks for each data field for a client in relation to the plurality of client accounts.

5. The method of claim 1 , further comprising:

generating a heat map based at least in part on the client activity data; and

identifying client accounts that are similar to the activity rank target profile based at least in part on the composite activity rank and the generated heat map.

6. The method of claim 5 , wherein generating the heat map comprises:

receiving data points from the data fields;

calculating a data distribution of the data points, wherein the data distribution includes bins, wherein each bin represents an interval;

rendering a heat map based, at least in part, on the data distribution, wherein the heat map includes regions corresponding to the bins;

defining thresholds for different values of a data point density;

assigning each bin a threshold based, at least in part, on the data point density of the bin;

assigning a region a color based, at least in part, on the assigned threshold of the bin corresponding to the region; and

coloring the region with the assigned color.

7. The method of claim 5 , further comprises:

plotting a trend graph of the composite activity rank of the identified client accounts;

determining a client is a risk for loss; and

identifying the client as a target high risk client.

8. The method of claim 7 , determining the client is a risk for loss includes:

tracking the composite activity score of a client in relation to a predetermined threshold score including the time the score is below the threshold.

9. A system for determining loss risks, comprising:

a processor that executes computer executable components stored in a memory;

a client database comprising account information for a set of clients;

a client activity database comprising a set of client activity data for entries in the client database;

a filtering component that filters the client activity data into a subset of client activity data associated with a subset of client accounts, wherein the filtering includes coarse filtering of client data and refined filtering of client activity data to generate the subset of client accounts;

a ranking component that generates a composite activity ranking of the subset of client accounts according to the filtered subset of the client activity data wherein the composite activity ranking of the subset of client accounts identifies at least one loss risk;

a target component that identifies an activity rank target profile, the activity rank target profile is based on a composite of client accounts in the subset of client accounts that have similar composite activity rankings; and

a mapping component that identifies client accounts from the set of client accounts that are similar to the activity rank target profile.

10. The system of claim 9 , wherein the client activity database includes data fields describing client account actions where at least one data field relates to a propensity for a loss risk.

11. The system of claim 10 , comprising:

a weighting component that applies weights for at least one data field, wherein the filtering component uses the weights to filter the client activity data.

12. The system of claim 11 , wherein the composite activating rank is based at least in part on the weighted ranks for each data field for a client in relation to the plurality of client accounts.

13. The system of claim 9 , comprising:

a heat mapping component that generates a heat map based at least in part on the client activity data; and

the mapping component that identifies client accounts that are similar to the activity rank target profile based at least in part on the composite activity rank and the generated heat map.

14. The system of claim 13 , wherein the heat mapping component, comprises:

a data distribution logic configured to calculate a data distribution of the data points, wherein the data distribution has bins, wherein each bin corresponds to an interval; and

a render logic configured to render a heat map based, at least in part, on the data distribution.

15. The system of claim 13 , further comprising:

a trend graphing component that plots a trend graph of the composite activity rank of an identified client account to determine a client is a risk for loss.

16. The system of claim 15 , wherein the trend graphing component tracks the composite activity score of a client in relation to a predetermined threshold score including the time the score is above the threshold to make a loss risk determination.

17. A non-transitory computer readable medium having instructions to control one or more processors configured to:

receive client data including client activity data associated with a set of client accounts;

filter the client activity data into a subset of the client activity data associated with a subset of client accounts, wherein the filtering includes coarse filtering of client data and refined filtering of client activity data to generate the subset of client accounts;

receive weights for at least one data field; and

apply the weights to the subset of client activity data;

generate a composite activity ranking from calculated composite activity scores of clients according to the filtered subset of the client activity data, wherein the composite activity ranking of clients identifies at least one loss risk;

identify an activity rank target profile, the activity rank target profile is based on a composite of client accounts identified as a loss risk in the subset of client accounts that have similar composite activity rankings; and

identify client accounts from the set of client accounts that are similar to the activity rank target profile.