Accessible Version
Mortgage Servicing Right Valuations Under Stress, Accessible Data
Figure 1. Bank Share of Agency Servicing, January 2014- May 2026
This line graph displays the bank market share of servicing for GNMA, the GSEs, and the combined agencies over a 12-year period from 2014 to 2026. The y-axis represents "share" ranging from 0.00 to 1.00 (0% to 100%), while the x-axis shows years from 2014 to 2026.
All three series begin in 2014 with share values between 0.60 and 0.65. The chart shows a general declining trend for all three categories over the 12-year period, though the rate and extent of decline varies by category.
All Agency (solid purple line): Begins at approximately 0.67 (67%) in 2014, declines steadily through 2017 to about 0.47 (47%), remains relatively stable around 0.47-0.48 through 2020, then continues a gradual decline to approximately 0.29 (29%) by 2026.
GNMA (dashed teal/blue line): Starts at approximately 0.67 (67%) in 2014, shows the steepest decline among all three lines, dropping continuously throughout the entire period to reach approximately 0.12 (12%) by 2026.
GSE (dashed orange line): Begins at approximately 0.64 (64%) in 2014, declines to about 0.50 (50%) by 2017, then rises slightly to peak around 0.55 (55%) during 2019-2020, followed by a gradual decline to approximately 0.37 (37%) by 2026.
Source: ICE, eMBS. eMBS Data.
Figure 2. Cumulative Mortgage Default Rates by Type of MBS Pool and Scenario
This line graph displays the cumulative probability of default over 36 projection quarters (9 years) for two types of securitized mortgages under baseline and severe stress scenarios. The y-axis represents "Cumulative P(Default)" ranging from 0.00 to above 0.20 (0% to over 20%), while the x-axis shows "Projection Quarter" from 0 to 36.
Four lines are plotted:
GNMA - baseline (solid purple line): Starts near 0.01 (1%) at quarter 0, increases steadily but gradually throughout the projection period, reaching approximately 0.13 (13%) by quarter 36.
GSE - baseline (solid gold line): Begins near 0.01 (1%) at quarter 0, shows the most gradual increase of all lines, reaching approximately 0.075 (7.5%) by quarter 36.
GNMA - severe (dashed purple line): Starts near 0.02 (2%) at quarter 0, rises steeply through the first 10 quarters, reaching approximately 0.17 (17%) by quarter 10, then continues to increase more gradually, ultimately reaching approximately 0.22 (22%) by quarter 36.
GSE - severe (dashed gold line): Begins near 0.01 (1%) at quarter 0, increases moderately in the early quarters, reaching about 0.10 (10%) by quarter 10, then continues rising steadily to approximately 0.12 (12%) by quarter 36.
Source: Authors’ calculations using data from Board of Governors of the Federal Reserve System, Capital Assessments and Stress Testing (FR Y-14).
Figure 3. Cumulative Mortgage Prepayment Rates by Type of MBS Pool and Scenario
This line graph displays the cumulative probability of prepayment over 36 projection quarters (9 years) for two types of securitized mortgaged under baseline and severe stress scenarios. The y-axis represents "Cumulative P(Prepayment)" ranging from 0.0 to 0.6 (0% to 60%), while the x-axis shows "Projection Quarter" from 0 to 36.
Four lines are plotted:
GNMA - baseline (solid purple line): Starts near 0.01 (1%) at quarter 0, increases steadily and nearly linearly throughout the projection period, reaching approximately 0.30 (30%) by quarter 36.
GSE - baseline (solid orange line): Begins near 0.01 (1%) at quarter 0, follows a very similar trajectory to GNMA baseline, nearly overlapping throughout and reaching approximately 0.29 (29%) by quarter 36.
GNMA - severe (dashed purple line): Starts near 0.01 (1%) at quarter 0, rises more steeply than in the baseline scenario, reaching approximately 0.20 (20%) by quarter 12, and continues to increase to approximately 0.50 (50%) by quarter 36.
GSE - severe (dashed orange line): Begins near 0.01 (1%) at quarter 0, shows the steepest increase among all lines, reaching approximately 0.28 (28%) by quarter 12, and continues rising to approximately 0.60 (60%) by quarter 36.
Source: Authors’ calculations using data from Board of Governors of the Federal Reserve System, Capital Assessments and Stress Testing (FR Y-14).
Figure 4. Modelled decline in MSR value from a 100bps CDR shock
The graph shows four time series with, annual values from 2013 to 2019, representing the average estimated percent change across banks in the projected value of MSR portfolios for a 100 bp CDR shock. The average incorporates a regression adjustment to account for the unbalanced panel of reporting banks. The four time series correspond to GSE versus GNMA portfolios, and alternative using simple average or averages weighted by the UPB of each bank's servicing portfolio. The y-axis represents percentage value declines ranging from -25% to 0%, while the x-axis shows years from 2013 to 2019.
Four lines are plotted:
GNMA - UPB Weighted (solid purple line): Starts at approximately -22% in 2013, improves (moves upward) steadily through 2016 to reach approximately -16%, remains relatively stable around -16% to -17% through 2018, then improves slightly to approximately -13% by 2019.
GSE - UPB Weighted (solid orange line): Begins at approximately -6.5% in 2013, remains relatively stable with minor fluctuations throughout the entire period, showing a slight dip to about -7% around 2017-2018, and ending at approximately -6.5% in 2019.
GNMA - Equal Weighted (dashed purple line): Starts at approximately -12% in 2013, remains relatively flat with minimal variation throughout the period, hovering between -12% and -13%, with a slight improvement toward -12% in 2016, and ending near -13% in 2019.
GSE - Equal Weighted (dashed orange line): Begins at approximately -5% in 2013, shows the most stable pattern of all lines, remaining nearly flat between -4.5% and -5% throughout the entire 7-year period, ending at approximately -5% in 2019.
Note: The time series shows the average estimated percent change across banks in the projected value of MSR portfolios for a 100 bp CDR shock. The average incorporates a regression adjustment to account for the unbalanced panel of reporting banks.
Source: Authors’ calculations using data from Board of Governors of the Federal Reserve System, Capital Assessments and Stress Testing (FR Y-14).