Common use of Random Sampling Clause in Contracts

Random Sampling. A PFI may utilize a statistical sampling method to determine its random sample size. This method must be designed to achieve a ninety-five percent (95%) level of confidence. If the PFI uses a statistical sampling method for its selection process in lieu of the standard 10% random selection process, it must document the methodology and provide to its MPF Bank upon request a detailed written justification of the methodology which must including the following: • Method for making a statistical selection; • Variables used in the selection model and how they are defined; • The results of periodic evaluations of the process and variables; and • Establishment of time periods for the evaluations. The MPF Bank or the MPF Provider may require adjustments to the statistical sampling methodology based upon a review of the statistical sampling methodology. A PFI that does not use a statistical sampling method must use the following post-Closing random sample selection procedures: • Select for review at least ten percent (10%) of one of the following origination populations: o Total annual home mortgage production; o Total annual secondary market home mortgage production; or o Total annual MPF Program originations. • The selection of the post-Closing QC Sample from within the chosen population must be made throughout the year based on the following schedule: Production Sample Selection Timing 10 or more Mortgage Loans closed in a month The selection of at least 10% of the mortgages closed within that month must be made no later than the last day of the following month. The PFI must notify the MPF Bank if their sampling cycle is in, or will be in, significant arrears. Example: In December, the PFI closes 12 loans; the sample must be selected by January 31. Less than 10 Mortgage Loans closed in a month The PFI may defer the sample selection until the month after which a combined 10 or more mortgages have closed since the last sample was selected. However, the PFI must select a sample no less than every 90 days. If after three consecutive months the combined production is <10 mortgages, the PFI must select one Mortgage for QC review.

Random Sampling. A PFI may utilize a statistical sampling method to determine its random sample size. This method must be designed to achieve a ninety-five percent (95%) level of confidence. If the PFI uses a statistical sampling method for its selection process in lieu of the standard 10% random selection process, it must document the methodology and provide to its MPF Bank upon request a detailed written justification of the methodology which must including the following: • Method for making a statistical selection; • Variables used in the selection model and how they are defined; • The results of periodic evaluations of the process and variables; and • Establishment of time periods for the evaluations. The MPF Bank or the MPF Provider may require adjustments to the statistical sampling methodology based upon a review of the statistical sampling methodology. A PFI that does not use a statistical sampling method must use the following post-Closing random sample selection procedures: • Select for review at least ten percent (10%) of one of the following origination populations: o Total annual home mortgage production; o Total annual secondary market home mortgage production; or o Total annual MPF Program originations. • The selection of the post-Closing QC Sample from within the chosen population must be made throughout the year based on the following schedule: Production Sample Selection Timing 10 or more Mortgage Loans closed in a month The selection of at least 10% of the mortgages closed within that month must be made no later than the last day of the following month. The PFI must notify the MPF Bank if their sampling cycle is in, or will be in, significant arrears. 70 MPF Announcement 2017-13 (3/28/17) Production Sample Selection Timing Example: In December, the PFI closes 12 loans; the sample must be selected by January 31. Less than 10 Mortgage Loans closed in a month The PFI may defer the sample selection until the month after which a combined 10 or more mortgages have closed since the last sample was selected. However, the PFI must select a sample no less than every 90 days. If after three consecutive months the combined production is <10 mortgages, the PFI must select one Mortgage for QC review. The selection must then be made no later than the end of the month following the accumulation of 10 or more closed mortgage or three consecutive months.

Random Sampling. A PFI may utilize a statistical sampling method to determine its random sample size. This method must be designed to achieve a ninety-five percent (95%) level of confidence. If the PFI uses a statistical sampling method for its selection process in lieu of the standard 10% random selection process, it must document the methodology and provide to its MPF Bank upon request a detailed written justification of the methodology which must including the following: • Method for making a statistical selection; • Variables used in the selection model and how they are defined; • The results of periodic evaluations of the process and variables; and • Establishment of time periods for the evaluations. The MPF Bank or the MPF Provider may require adjustments to the statistical sampling methodology based upon a review of the statistical sampling methodology. A PFI that does not use a statistical sampling method must use the following post-Closing random sample selection procedures: • Select for review at least ten percent (10%) of one of the following origination populations: ; o Total annual home mortgage production; o Total annual secondary market home mortgage production; or o Total annual MPF Program originations. • The selection of the post-Closing QC Sample from within the chosen population must be made throughout the year based on the following schedule: Production Sample Selection Timing 10 or more Mortgage Loans closed in a month The selection of at least 10% of the mortgages closed within that month must be made no later than the last day of the following month. The PFI must notify the MPF Bank if their sampling cycle is in, or will be in, significant arrears. Example: In December, the PFI closes 12 loans; the sample must be selected by January 31. Less than 10 Mortgage Loans closed in a month The PFI may defer the sample selection until the month after which a combined 10 or more mortgages have closed since the last sample was selected. However, the PFI must select a sample no less than every 90 days. If after three consecutive months the combined production is <10 mortgages, the PFI must select one Mortgage for QC review. The selection must then be made no later than the end of the month following the accumulation of 10 or more closed mortgage or three consecutive months.