The NAIC’s Capital Markets Bureau monitors developments in the capital markets globally and analyzes their potential impact on the investment portfolios of US insurance companies. A list of archived Capital Markets Bureau Special Reports is available via the index
Modeling of US Insurance Industry’s Holdings in Non-Agency Mortgage-Backed Securities
With year-end 2012, the NAIC successfully completed the modeling of non-agency residential mortgage-backed securities (RMBS) and commercial mortgage-backed securities (CMBS) held by the U.S. insurance industry. This was the fourth year for modeling RMBS and the third year for modeling CMBS. As in prior years, to the extent data was available, each individual holding was modeled for expected losses using different economic scenarios. The weighted average of those expected losses — assuming the bonds were held to maturity — were then translated into an expected recovery value for each individual bond. In lieu of using ratings by the nationally recognized statistical rating organizations (NRSROs), U.S. insurance companies are required to compare the carrying values of RMBS and CMBS holdings as of each year-end to the expected recovery values, to determine NAIC designations, which are then mapped to risk-based capital (RBC) factors. A small percentage of non-agency RMBS and CMBS were not modeled for different reasons. These non-modeled securities include interest-only strips, foreign transactions and some highly complex re-securitizations. For these bonds, U.S. insurers continued to rely on NRSRO ratings, but factored in carrying values in comparison to a fixed matrix of values.
As the Capital Markets Bureau has done in the past, the purpose of this special report is to consider the RMBS and CMBS modeling results and their impact on the U.S. insurance industry’s RBC, in comparison with the prior methodology that relied on NRSRO ratings, but that also assumed holdings were at par. Perhaps what is most noteworthy about the 2012 analysis is that there is nothing remarkable in comparison with the analysis from prior years.
Residential Mortgage-Backed Securities
As in prior years, at the end of 2012, after a public exposure and comment period, the NAIC’s Valuation of Securities (E) Task Force formally adopted assumptions to be used for modeling the U.S. insurance industry’s non-agency RMBS. One difference with respect to the assumptions is that, for year-end 2012, the Task Force decided to use only four scenarios for RMBS, as opposed to five scenarios that were used in prior years. CMBS modeling has always only used four scenarios. The RMBS assumptions were as follows:
Table 1: Assumptions for Year-End 2012 RMBS Modeling
For year-end 2012, the results for 18,901 unique CUSIPs were sent to U.S. insurers. The expected recovery values for these securities were used to determine NAIC designations for a total exposure of $110.5 billion in book/adjusted carrying value (BACV) across the entire industry. This BACV amount continued a trend of declining exposure from $123.2 billion as of year-end 2011, $127.7 billion as of year-end 2010 and $150.5 billion as of year-end 2009. While there was some acquisition activity during 2012 that totaled approximately $19.0 billion, the U.S. insurance industry’s exposure to non-agency RMBS continued to decline for three primary reasons: (1) even though there has been some recovery in new issuance in the marketplace, it continues to be far below historical levels; (2) there were amortizations on existing holdings; and (3) albeit smaller, some additional impairments were taken during the year. Other-than-temporary-impairments (OTTI) and fair value revaluations taken during 2012 totaled $693.6 million. This is compared with $2.9 billion in 2011, $2.8 billion in 2010, and $15.0 billion in 2009. OTTI taken in 2012 were $1.7 billion, partially offset by fair value revaluations, which were in aggregate positive, totaling $1.1 billion.
With respect to the 18,091 CUSIPs reported for year-end 2012, insurance companies held 16,793 of this total as of year-end 2011, which meant that they were modeled during that year-end process. The average expected recovery value of the common CUSIPs increased modestly from 81.91% of par to 82.06% of par from 2011 to 2012. Also, there were 14,876 CUSIPs that have been modeled in each of the four years ranging from 2009 to 2012. Within this population, 50.5% showed little or no change in expected recovery value from 2011 to 2012; 20.6% showed a moderate to significant increase. In addition, the expected recovery value for 15.9% of the CUSIPs experienced a moderate decrease. Of potential concern is 13.0% of the CUSIPs which reported a substantial decrease in expected recovery value from 2011 to 2012. During this time period, the average expected recovery value declined from 86.61% of par in 2009 to 83.96% of par in 2010, and to 81.25% of par in 2011; and it increased to 81.48% of par in 2012.
Based on the results in expected recovery values and the industry’s year-end 2012 BACV prices, the breakdown by NAIC designations were as follows:
Table 2: NAIC Designations Based on Expected Recovery Values and BACV Prices for RMBS
As was the case in the prior three years, factoring in the insurer’s BACV price into the determination made a fairly substantial difference in the profile of NAIC designations. Ignoring the different carrying values and relying solely on NRSRO ratings would have yielded the following breakdown:
Table 3: NAIC Designations Relying Solely on NRSRO Ratings for RMBS
As has been noted before, a key component of the process for assigning NAIC designations is the comparison of expected recovery values that result from the modeling with individual carrying values. Impairments taken by U.S. insurers, as well as purchases at what were depressed market prices following the market meltdown in 2008 and 2009, have resulted in average carrying values that are substantially discounted from par.
Table 4: Differentials in Carrying vs. Expected Recovery Values in RMBS (2010–2012)
Over the past three years, the carrying value of the industry’s RMBS exposure has gradually declined overall, as well as for those holdings with an assigned NAIC 1 designation. Over the same period, the expected recovery value has improved modestly for the overall RMBS exposure, while declining modestly for those holdings assigned an NAIC 1 designation. The net result is that the differential — which can be characterized as an overall cushion — for the aggregate RMBS exposure has improved by 1.68 percentage points for the overall exposure and by 2.47 percentage points for those with NAIC 1 designations.
RBC Impact for RMBS
The purpose of assigning NAIC designations, whether relying solely on NRSRO ratings or through the more finely tuned modeling approach, is to map each insurer’s holding to a RBC factor. Under the current NAIC process for assigning designations for non-agency RMBS, total RBC is lower than it would have been if the NAIC had continued to rely solely on NRSRO ratings. As shown in Table 4, in aggregate, the BACV price for the industry’s RMBS exposure was 79.49%, which compares favorably to the average expected recovery value of 86.12%. Without accounting for the impact of the covariance component of the RBC formula, a comparison using a breakdown of the industry’s RMBS exposure based on the modeling approach is detailed in the table below:
Table 5: Differential in RBC for RMBS
The $79.0 billion with an NAIC 1 designation under the modeling approach has an average BACV price of 77.02%, which was below the overall average price of 79.49%. Given the substantial discount to par and the favorable comparison to the modeled expected recovery value of 86.95%, there is a substantially lower RBC requirement for these bonds. In aggregate, actual realized recoveries on these NAIC 1 holdings (which represented 71.5% of the total exposure) could decrease more than 11% and the insurance company holders would still fully recover their current carrying value. The reverse is true at the lower end of Table 5, where the modeled results drive a higher RBC requirement relative to the results that would have been the case relying solely on NRSRO ratings. This relationship is reversed if the table is based on NRSRO rating equivalents.
Table 6: Differential in RBC for RMBS
Since 2009, the differential in RBC has continued to widen between the two approaches as an increasing percentage of the industry’s non-agency RMBS holdings have been downgraded by the NRSROs, while at the same time the industry’s margin between carrying value and expected recovery valuations has improved. In comparison, for year-end 2011, the exposures and, therefore, RBC differentials also reflected a modest shift from NAIC 4 and NAIC 5 designations to NAIC 2 and NAIC 3.
Table 7: RBC Before Covariance (for RMBS)
Fair Value Estimates
The general turmoil in the RMBS market since 2008 has also been reflected in the volatility of market values. Immediately upon the onset of the financial crisis, there was virtually no RMBS market at all. There has been some improvement since then. Generally market wide, this is reflected in the ABX, which is a credit default swap index comprised of subprime RMBS. As of the end of 2009, this index was trading at a price of 35% of notional, which then recovered to 50% of notional before dropping back down to 35% at the end of 2011. Since then, the ABX has been fairly consistently improving to 50% at the end of 2012, and more recently topping out at 58% of notional in May 2013.
The industry’s holdings have seen a similar progression. Based on reported fair values, the aggregate fair value as of year-end 2010 was 75.51% of par, and then dipped slightly to 75.19% of par at the end of 2011. It was reported at 82.01% of par at the end of 2012. The aggregate price of 82.01% is above the average carrying value of 79.49% of par, whereas carrying value exceeded reported fair value at the end of 2010 (80.94% versus 75.51%) and 2011 (80.07% versus 75.19%).
Markit ABX 07-1, AAA Tranche
Commercial Mortgage-Backed Securities
At the same time that the Valuation of Securities (E) Task Force adopted assumptions for year-end 2012 RMBS modeling, it also adopted property value assumptions to be used in CMBS modeling.
Table 8: Assumptions for Year-End 2012 Modeling of CMBS
For year-end 2012, the results for 4,294 unique CUSIPs were sent to U.S. insurers. The expected recovery values for these securities were used to determine the NAIC designation for a total exposure of $151.4 billion in BACV across the entire industry. This total BACV exposure compares with $161.9 billion at year-end of 2011, $171.6 billion at year-end of 2010 and $184.4 billion at year-end of 2009. CMBS acquisitions across the entire industry totaled approximately $30.5 billion in 2012, equal to roughly 3.5% of all bond acquisitions. Although there was some acquisition activity during 2012, the U.S. insurance industry’s exposure to non-agency CMBS continued to decline, albeit modestly, for several reasons: (1) despite a more significant recovery in new issue volume than in the RMBS new issuance market, the number and size of new CMBS transactions has remained modest compared to historical peaks; (2) there were amortizations on existing holdings; and (3) additional impairments were taken during the year. OTTI and fair value revaluations taken during 2012 totaled $784.6 million. This compares with $740 million in 2011, $4.0 billion in 2010 and $2.2 billion in 2009. OTTI taken in 2012 were $903.3 million, partially offset by fair value revaluations, which were in aggregate positive, and totaled $118.7 million.
With respect to the 4,294 CUSIPs reported for year-end 2012, insurance companies held 3,850 of these same CUSIPs as of year-end 2011, which meant that they were modeled during that year-end process. The average expected recovery value of the common CUSIPs decreased from 86.28% of par to 84.31% of par from 2011 to 2012. Also, there were 3,684 CUSIPs that had been modeled in each of the three years from 2010 to 2012. During this time period, the average expected recovery value declined from 86.27% of par in 2010, to 85.73% of par in 2011 and finally to 83.70% of par in 2012.
Based on the results in expected recovery values and the industry’s year-end 2012 BACV prices, the breakdown by NAIC designation was as follows:
Table 9: NAIC Designations Based on Expected Recovery Values and BACV Prices of CMBS
As was the case in prior years, factoring in the insurer’s BACV price into the determination resulted in some differences in the profile of NAIC designations. Ignoring the different carrying values and relying solely on NRSRO ratings would have yielded the following breakdown:
Table 10: NAIC Designations Relying Solely on NRSRO Ratings of CMBS
As has been the case every year, the shift in designations for CMBS was not as significant as it was for RMBS. Total exposure to bonds with an NAIC 1 designation increased 8.5 percentage points (using the current modeling approach) from the prior methodology that relied solely on NRSRO ratings. For RMBS, there was a comparable increase from 15.0% to 71.5%. Also notable, and in line with prior years, is that 89.4% of the industry’s CMBS holdings reflected no change in the NAIC designation between the two approaches. Where there was a difference, 9.5% reflected an improvement and 1.1% reflected a lower designation.
RBC Impact for CMBS
The purpose of assigning an NAIC designation, whether relying solely on NRSRO ratings, or through a more finely tuned modeling approach, is to map each insurer holding to a RBC factor. Under the current NAIC process for assigning designations for non-agency CMBS, total RBC is lower than it would have been if the NAIC had continued to rely solely on NRSRO ratings. In aggregate, the BACV price for the industry’s CMBS exposure was 96.02% of par, which compares favorably to the average expected recovery value of 96.41% of par. Without accounting for impact of the covariance component of the RBC formula, a comparison using a breakdown of the industry’s CMBS exposure based on the modeling approach is detailed in the table below:
Table 11: Differential in RBC for CMBS
The $146.6 billion with an NAIC 1 designation under the modeling approach has an average BACV price of 96.72% of par, which was slightly above the overall average of 96.02%. However, the positive differential to the modeled expected recovery value is wider, at 0.88 percentage points for the NAIC 1 category, versus 0.39 percentage points overall. Given the discount to par and the favorable comparison to the modeled expected recovery value, there is a lower RBC requirement. This positive differential, however, is smaller than it was relative to year-end 2011 modeling, by comparison. The differentials for 2011 were 1.75 percentage points for bonds with an NAIC 1 designation and 1.01 percentage points overall.
Table 12: Differential in RBC for CMBS
Table 13 shows an estimate of the impact on RBC over the past three years between the current modeling approach for expected recovery values and comparing them to carrying values, versus the prior approach of solely relying on NRSRO ratings. The differentials do not take into account the impact of covariance component of the RBC formula; therefore, this is for illustrative purposes only. The covariance component is important because it serves to smooth significant changes in RBC charges, especially when those changes are reflected in smaller portions of an insurer’s portfolio and are less correlated with other factors in the calculation.
Table 13: RBC Before Covariance (for CMBS)
In 2010 — the first year that the new modeling approach was applied to CMBS — the RBC requirement actually increased. The impact of modeling on RBC for CMBS, as opposed to the prior method (i.e., relying solely on NRSRO ratings), can be attributed to several factors, including a skewing of insurance industry holdings to more senior and super senior tranches in the capital structure and later impairments and revaluations. This latter factor is not surprising, given that the commercial real estate market tends to lag economic cycles.
As has been noted in previous Capital Markets Bureau special reports, the decision to change the process for assigning NAIC designations for non-agency RMBS resulted in a number of benefits. Most significant of these is a calibration of NAIC designations and the RBC factors that the holdings are mapped to, along with a greater level of sophistication that goes beyond simple credit risk. Continuing to rely solely on NRSRO ratings would have been particularly problematic for bonds that are carried at a substantial discount to par, as well as in situations often cited where the potential loss of principal was small in comparison with the overall size of the holding. In addition, there have been improvements in transparency and regulatory oversight of the process, as well as more accurate valuations by insurers. Assuming the NAIC continues to employ this approach, enhancements to the process may be considered.
The Investment Risk-Based Capital (E) Working Group (formerly the C1 Factor Review (E) Subgroup) is expecting to take under consideration the current framework as part of its charges. In its preliminary discussions on the topic, differences in volatility between individual RMBS and CMBS are one consideration. This was discussed in depth in two Capital Markets Bureau special reports published in 2012: “Potential for Volatility in U.S. Insurer Holdings of Residential Mortgage-Backed Securities,” published April 11, 2012; and “Potential for Volatility in U.S. Insurer Holdings of Commercial Mortgage-Backed Securities,” published July 12, 2012. In these two reports, we noted that some bonds displayed relatively stable profiles across the different economic scenarios, while others showed dramatically different results in the more conservative scenarios. The latter group was referred to as “cuspy” bonds. Another important factor will be how the current modeling process fits within the overall RBC framework, specifically as it relates to calibration and confidence levels.
Questions and comments are always welcome. Please contact the Capital Markets Bureau at CapitalMarkets@naic.org.
The views expressed in this publication do not necessarily represent the views of NAIC, its officers or members. NO WARRANTY IS MADE, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY OPINION OR INFORMATION GIVEN OR MADE IN THIS PUBLICATION.
© 1990 – 2018 National Association of Insurance Commissioners. All rights reserved.