SBA 7(a) prepayment vectors

SBA 7(a) loans have a tendency for very slow prepayment speeds during the initial years, followed by a spike before settling back down for the remaining term. Viewed on a graph, the tendency forms a bell-shaped curve. By placing a pool at the appropriate point on the curve, we can predict future returns.


Vector modeling

The Bloomberg Professional system’s PSBA report details the actual prepayment experience of all SBA 7(a) loans that have been pooled in the past 10 years. The report covers more than $40 billion worth of loans, and is updated monthly when SBA Pool factors are issued.

We use this data to calculate the weighted average of loans prepaid during each respective year of issuance and build vectors to reflect the historical performance. To account for seasoning, the vectors are advanced to match the seasoning of the subject SBA Pool.

We calculate three vector paths for each of the five original WAM categories: 

Actual Performance (CPR) calculates the actual weighted average prepayment speed of the loans during each respective year of existence.

Worst Performance (CPR) identifies the absolute worst performance posted during each respective year.

Best Performance (CPR) identifies the best performance posted for each respective year.

Prepayment Vectors for >21 Year No Cap

Factor date: June 2020

Prepayment Speed by Year CPR Equivalent
Year 1 2 3 4 5 6 7 8 9 10 11
Actual 2.63 5.74 12.71 19.82 20.30 19.59 20.09 22.11 19.91 17.58 17.10 13.10
Worst 4.8 7.0 16.1 27.1 26.5 27.2 25.0 26.6 24.1 24.2 17.1 16.25
Best 0.4 2.4 5.7 11.1 10.7 8.4 7.0 9.9 14.3 13.3 17.1 7.77

Past performance does not always indicate future results. Prevailing economic conditions and expectations for the broader economy and interest rates also factor into prepayment performance forecasts. While we can’t precisely predict the future, we can help investors make informed decisions.