Contact our Trading Desk
(901) 261-5950

Responsive. Reliable. Experienced.

We help banks, credit unions, and asset managers energize their portfolios and reduce volatility.
  • Trade highly sought-after variable-rate government-guaranteed paper
  • Offer alternative strategies for clients to capitalize on the SBA/USDA market
  • Provide traditional fixed income financial service



Hanover Securities had a strong showing at

NAGGL 2019 Annual Conference



    From the desk of Joel Banes October 2019           

       download: Joel Banes' Jewels and Pools

In October 2017, the SBA began returning embedded principal to SBA pool owners. The structure of newly issued pools changed as well. The new pool structure prevents “excessive” embedded principal from building up again.



Now, after almost two years of history, what have we learned?  

We ran the following simulation using Bloomberg PSBA speeds for the first three years of October 2017 and September 2019. We then assumed a flat speed of 15%/20%/25%/30%/35% for the remainder of the life to create vectored prepayment speeds. We used the first three years, because there was minimal, if any, embedded principal.

First, let’s consider the difference between October 2017 and September 2019.

There is less than a 1% impact in the vector speeds in every circumstance. The only difference in the calculations are the first three years of speeds. As such, it demonstrates the stability of prepayment speeds in the critical first three years of SBA pools. Secondly, when looking at PSBA (September 2019) for YTD (21yr+ WAM) speeds on pools originating from 2009 to 2016, the speeds range from 19.70% to 26%. It is important to note that these pool speeds are inflated due to the return of excessive embedded principal. In projecting speeds after three years, would one expect pools with normal embedded principal* to pay as fast as the average 23% CPR pools with excessive embedded principal?

PSBA YTD speeds for 8-11yr+ WAM paper range from 20.7% to 35.1% (excluding the 2009 issue which is an outlier and so small it is insignificant). Again, would one use the average speed of pools, approximately 28%, with excessive embedded principal? Simply put, all other variables equal, pools that are returning normal embedded principal to investors will exhibit a slower prepayment speed than pools with excessive embedded principal. (As of September 2019, there was $42.25mm in embedded principal contained in $14.30bln of pools, or 0.30%.)

However, all other variables were not equal during the past few years. The most obvious variable is that Prime rate increased 2.25% (3.25% to 5.50%) from December 2015 – December 2018. The yield curve inverted as longer-term fixed rates did not follow the rise in the prime rate. Opportunities to refinance increased substantially.

Prime rate has subsequently decreased 50 basis points over the last 90 days with more rate cuts anticipated. These rate cuts lower the disparity between fixed and variable rate loans substantially reducing a borrower’s motivation to refinance. In making prepayment assumptions, please keep the following in mind:

            1) Vector calculations here demonstrate the immense stability of the critical first three years prepayment speeds of SBA pools

            2) Newly originated pools produce normal embedded principal, and

            3) The 2.25% rise in prime has been followed by a 50-basis point reduction with more rate cuts anticipated.

Virtually every disclaimer includes the verbiage “past performance is not indicative of future results.” As one looks to predict future prepayment rates of SBA pools, one must take into consideration many variables, of which we believe the two most critical variables are the; 

            1) excessive embedded principal versus normal embedded principal and 

            2) the disparity between fixed and variable rates.

We encourage all investors to use vectors, keep them updated, and not to judge new originated pools with normal embedded principal based on the prepayment speed of pools with excessive embedded principal.

The purchase of newly originated SBA pools on a regular basis remains the most critical part of building a solid portfolio and stabilizing the impact of prepayment speeds.

We are focused on the impact of SBA prepayment speeds, and we understand that our clients are as well. We believe you should know that the lifetime CPR (September 2019) of all SBA pools originated by Hanover Securities / Shay Financial Services since 2007 is 10.43%. That is the lowest CPR of any SBA approved pool assembler forming SBA pools in the same period. Please contact your Hanover representative if we can help in any way.

*Normal embedded principal: Pools issued 10-1-2017 or later have reduced embedded principal.


JOEL BANES (901) 261-5950


Although this information was derived from sources which we believe reliable, we do not guarantee its accuracy, and it may be incomplete or condensed. This is for informational purposes only and is not intended as an offer or solicitation with respect to the purchase or sale of any security. Securities which may be shown are subject to prior sale and change in price. Past performance is not indicative of future results. Changes in assumptions may have a material effect on projected results. Hanover Securities is a member of FINRA and SIPC.