NMI HOLDINGS, INC. The following analysis should be read in conjunction with our unaudited condensed consolidated financial statements and the notes thereto included in this report and our audited financial statements, notes thereto and "Management's Discussion and Analysis of Financial Condition and Results of Operations" included in our 2019 10-K, for a more complete understanding of our financial position and results of operations. In addition, investors should review the "Cautionary Note Regarding Forward-Looking Statements" above and the "Risk Factors" detailed in Part II, Item 1A of this report and in Part I, Item 1A of our 2019 10-K, as subsequently updated in other reports we file with theSEC , for a discussion of those risks and uncertainties that have the potential to affect our business, financial condition, results of operations, cash flows or prospects in a material and adverse manner. Our results of operations for interim periods are not necessarily indicative of results to be expected for a full fiscal year or for any other period. Overview We provide private MI through our wholly-owned insurance subsidiaries NMIC and Re One. NMIC and Re One are domiciled inWisconsin and principally regulated by the Wisconsin OCI. NMIC is our primary insurance subsidiary and is approved as an MI provider by the GSEs and is licensed to write coverage in all 50 states and D.C. Re One provides reinsurance to NMIC on insured loans after giving effect to third-party reinsurance. Our subsidiary, NMIS, provides outsourced loan review services to mortgage loan originators. MI protects lenders and investors from default-related losses on a portion of the unpaid principal balance of a covered mortgage. MI plays a critical role in theU.S. housing market by mitigating mortgage credit risk and facilitating the secondary market sale of high-LTV (i.e., above 80%) residential loans to the GSEs, who are otherwise restricted by their charters from purchasing or guaranteeing high-LTV mortgages that are not covered by certain credit protections. Such credit protection and secondary market sales allow lenders to increase their capacity for mortgage commitments and expand financing access to existing and prospective homeowners. NMIH, aDelaware corporation, was incorporated inMay 2011 , and we began start-up operations in 2012 and wrote our first MI policy in 2013. Since formation, we have sought to establish customer relationships with a broad group of mortgage lenders and build a diversified, high-quality insured portfolio. As ofMarch 31, 2020 , we had master policies with 1,501 customers, including national and regional mortgage banks, money center banks, credit unions, community banks, builder-owned mortgage lenders, internet-sourced lenders and other non-bank lenders. As ofMarch 31, 2020 , we had$101.0 billion of total insurance-in-force (IIF), including primary IIF of$98.5 billion , and gross RIF of$25.3 billion , including primary RIF of$25.2 billion . We believe that our success in acquiring a large and diverse group of lender customers and growing a portfolio of high-quality IIF traces to our founding principles, whereby we aim to help qualified individuals achieve their homeownership goals, ensure that we remain a strong and credible counter-party, deliver a unique customer service experience, establish a differentiated risk management approach that emphasizes the individual underwriting review or validation of the vast majority of the loans we insure, utilizing our proprietary Rate GPSSM pricing platform to dynamically evaluate risk and price our policies, and foster a culture of collaboration and excellence that helps us attract and retain experienced industry leaders. Our strategy is to continue to build on our position in the private MI market, expand our customer base and grow our insured portfolio of high-quality residential loans by focusing on long-term customer relationships, disciplined and proactive risk selection and pricing, fair and transparent claim payment practices, responsive customer service, financial strength and profitability. Our common stock trades on the NASDAQ under the symbol "NMIH." Our headquarters is located inEmeryville, California . As ofMarch 31, 2020 , we had 327 full- and part-time employees. Our corporate website is located at www.nationalmi.com. Our website and the information contained on or accessible through our website are not incorporated by reference into this report. We discuss below our results of operations for the periods presented, as well as the conditions and trends that have impacted or are expected to impact our business, including new insurance writings, the composition of our insurance portfolio and other factors that we expect to impact our results. COVID-19 Developments OnJanuary 30, 2020 , the WHO declared the outbreak of COVID-19 a global health emergency and characterized the outbreak as a global pandemic onMarch 11, 2020 . In an effort to stem contagion and control the COVID-19 pandemic, the 26 -------------------------------------------------------------------------------- population at large has severely curtailed day-to-day activity and local, state and federal regulators have imposed a broad set of restrictions on personal and business conduct nationwide. The COVID-19 pandemic, along with the widespread public and regulatory response, has caused a dramatic slowdown inU.S. and global economic activity. In the weeks following the outbreak, non-essential businesses across theU.S. have been shuttered and capital markets have experienced a significant spike in volatility and sell-off in valuations. A record number of Americans have been furloughed or laid-off, and unemployment claims have increased dramatically. The global dislocation caused by COVID-19 is unprecedented and, while there is broad hope for a medical advance that relieves the crisis and provides for a quick return to normalized activity, it is not known how long the dislocation will persist. In response to the COVID-19 outbreak and continuing uncertainties, we activated our business continuity program to ensure our employees were safe and able to continue serving our customers and their borrowers without interruption. We have also sought to broadly assess the potential impact the COVID-19 outbreak will have on theU.S. economy and housing market, and the implications for the mortgage insurance market, and our business performance and financial position. Potential Impact on theU.S. Housing Market and Mortgage Insurance Industry We expect the COVID-19 outbreak will have a direct effect on theU.S. housing market, with existing homeowners facing challenges related to COVID-19, and the volume and timing of future housing transactions negatively impacted as potential sellers re-evaluate or postpone planned sales (housing supply) and potential buyers reassess their ability and willingness to purchase homes (demand). The market may be further constrained as transactions between committed buyers and sellers encounter operational challenges, such as delayed title searches due to the closure of local government offices. As a market (like all others) where valuation is driven by supply/demand dynamics, this dislocation has the potential to impact housing prices. While it is too early for us to estimate the magnitude of any potential housing price decline (either nationally or in local markets), we have looked to the 2008 Financial Crisis as a point of comparison. In the wake of the 2008 Financial Crisis, national home price indices declined by approximately 30% from their pre-crisis peaks and remained severely dislocated for several years, causing significant stress for homeowners, lenders, mortgage insurers and the broader economy. We observe several differences in the current environment as compared to that in the period leading up to and through the 2008 Financial Crisis that we believe may lessen the relative housing price dislocation experienced in the aftermath of the COVID-19 outbreak, including: (i) the generally higher quality borrower base (as measured by weighted average FICO scores and LTV ratios) and tighter underwriting standards (with, among other items, full-documentation required to verify borrower income and asset positions) that prevail in the current market; (ii) the lower concentration of higher risk loan structures, such as negative amortizing, interest-only or short-termed option adjusted-rate mortgages being originated and outstanding in the current market; (iii) the meaningfully higher proportion of loans used for lower risk purposes, such as the purchase of a primary residence or rate-term refinancing in the current market, as opposed to cash-out refinancings, investment properties or second home purchases, which prevailed to a far greater degree in the lead up to the 2008 Financial Crisis; (iv) the availability and immediate application by politicians, regulators, lenders, loan servicers and others of a broad toolkit of resources designed to aid distressed borrowers, including forbearance, foreclosure moratoriums and other assistance programs codified under the CARES Act enacted onMarch 27, 2020 ; and (v) the broader and equally immediate application of a massive amount of fiscal and monetary stimulus by the federal government under the CARES Act and across a range of other programs designed to assist unemployed individuals and distressed businesses, as well as support the smooth functioning of various capital and risk markets. We also perceive the house price environment in the period leading up to the current COVID-crisis to be anchored by more balanced market fundamentals than that in the period leading up to the 2008 Financial Crisis. We believe the 2008 Financial Crisis was directly precipitated by irresponsible behavior in the housing market in which home prices were driven to unsustainable heights (a so called "bubble"). We see a causal link between the housing market and the 2008 Financial Crisis that we do not see in the COVID-19 outbreak, and we believe this will further lessen the relative housing price dislocation experienced going forward. 27 -------------------------------------------------------------------------------- Notwithstanding the relative differences we observe in the current environment as compared to the 2008 Financial Crisis, we do expect that housing prices will be impacted by the COVID-19 outbreak. We also expect the mix of purchase and refinancing mortgage originations will shift in the near-term due to the current pandemic. We anticipate that purchase mortgage origination volume will be lower in the aftermath of the COVID-19 outbreak than it otherwise would have been as buyers and sellers postpone or more broadly reassess current and future transactions, while we expect refinancing origination volume will increase significantly as declining mortgage rates create refinancing opportunities for existing borrowers. Historically,U.S. mortgage insurance industry penetration of the purchase origination market has been meaningfully higher than the refinancing origination market, as new home buyers have generally required a greater degree of down payment support, while refinancing borrowers have typically benefited from increasing equity in their existing homes. In recent periods, however, mortgage insurance industry penetration of the refinancing market has increased significantly and the composition of industry new insurance written (NIW) volume between purchase and refinancing loans has shifted. Refinancing originations accounted for approximately 30% of total mortgage insurance industry NIW volume for the quarter-endedDecember 31, 2019 , compared to approximately 6% for the quarter-endedDecember 31, 2018 . We expect this trend to continue in the near-term as purchase origination volume slows and refinancing origination volume increases, particularly as those borrowers who capitalize on the emerging refinancing opportunity may not have significant equity in their homes (either because they are refinancing so soon after closing on a new home purchase or because of a decline in the appraised value of their property) and may need to rely on mortgage insurance support to complete their transactions. In this context, we expect to see a decline in totalU.S. mortgage insurance industry NIW volume in the near-term, with the impact of declining purchase origination volume partially offset by a dramatic increase in, and increasing mortgage insurance penetration of, refinancing origination volume. Potential Impact on NMI's Business Performance and Financial Position Operations We had 327 employees atMarch 31, 2020 , including 173 who typically work at our corporate headquarters inEmeryville, CA and 154 who typically work from home in locations across the country. In response to the COVID-19 outbreak, we activated our business continuity program and instituted additional work-from-home practices for our 173 Emeryville-based staff. We have transitioned our operations seamlessly and continue to positively engage with customers on a remote basis. Our IT environment, underwriting capabilities, policy servicing platform and risk architecture have continued without interruption, and our internal control environment and internal controls over financial reporting are unchanged. We have achieved this transition without incurring additional capital expenditures or operating expenses and we believe our current operating platform can continue to support our newly distributed needs for an extended period without further investment beyond that planned in the ordinary course. New Business Production We expect that the volume, composition, credit profile and pricing of our new business production will change due to the COVID-19 crisis. Our NIW has expanded significantly in past periods, driven by growth in the overall mortgage insurance market and the success we have had further developing our customer franchise. During the three-year period betweenJanuary 1, 2017 andDecember 31, 2019 , we activated 340 new lenders, growing our franchise by over 45%, and successfully deepened our engagement with many existing customers. In the same period, totalU.S. mortgage insurance NIW volume increased by 19% annually - peaking at$384 billion for the year-endedDecember 31, 2019 . Notwithstanding our recent gains and the customer success we have achieved, we expect that our new business volume will decline in the near-term along with the broader decline we anticipate in totalU.S. mortgage insurance industry NIW volume. We have broadly defined underwriting standards and loan-level eligibility criteria that are designed to limit our exposure to higher risk loans, and have used Rate GPS to actively shape the mix of our new business production and insured portfolio by, among other risk factors, borrower FICO score, debt-to-income (DTI) ratio and LTV ratio. AtMarch 31, 2020 , the weighted average FICO score of our IIF was 751 and we had a 4% mix of below 680 FICO score risk. Similarly, atMarch 31, 2020 , the weighted average LTV ratio (at origination) of our insured portfolio was 91.9% and we had a 10% mix of 97% LTV risk. In the weeks since the outbreak of COVID-19, we have adopted changes to our underwriting guidelines, including changes to our loan documentation requirements, our asset reserve requirements, our employment verification process and our income continuance determinations, that we expect will further strengthen the credit risk profile of our new business production. 28 -------------------------------------------------------------------------------- We set our premium rates based on a broad range of individual and market variables, including property type, type of loan product, borrower credit characteristics, and lender profile. Given the significant economic dislocation caused thus far by the COVID-19 outbreak, and the uncertain duration and ultimate global impact of this crisis, we have taken action to increase the premium rates we charge on all new business production, in accordance with our filed rates and applicable rating rules. We expect the pricing changes we have instituted will increase the rate received on NIW volume in future periods and further enhance the credit mix of our new business production. Delinquency Trends and Claims Expense We had 1,449 NODs in our primary insured portfolio as ofMarch 31, 2020 , which represented a 0.38% delinquency rate against our 376,852 total policies in-force. Given the recency of the COVID-19 outbreak, most borrowers (even those immediately impacted by the crisis) have not defaulted on their mortgages (i.e., missed two or more payments as due). We expect that we will see a significant increase in our default population going forward as borrowers face challenges related to COVID-19 and access the forbearance program for federally backed loans codified under the CARES Act or other programs made available by private lenders. As ofApril 30, 2020 , our default population had increased to 1,610, which represented a 0.43% delinquency rate. We are not yet able to forecast the ultimate level of forbearance-driven delinquencies we will receive or the timing in which they will develop. While early, we have observed a correlation between the risk profile of the underlying borrowers and the incidence of forbearance in the data received throughApril 30, 2020 . Borrowers with weaker credit profiles appear to be accessing forbearance programs with notably higher frequency. This trend is further supported by the forbearance data being reported by Black Knight McDash, which indicates that a meaningfully higher concentration of mortgages insured by the FHA andVA are in forbearance status compared to the loans purchased by the GSEs. Mortgages insured by the FHA andVA are generally of meaningfully lower credit quality than those purchased by the GSEs. We are monitoring this trend with particular focus given the high-quality credit profile of our insured portfolio. We establish reserves for claims and allocated claim expenses when we are notified that a borrower is in default (i.e., has missed two or more mortgage payments). As our default population grows in future periods, we expect to establish increasing reserves and incur additional claims expense. The size of the reserve we establish for each defaulted loan (and by extension our aggregate reserve and claims expense) will reflect our best estimate of the future claim payment to be made under each individual policy. Our future claims exposure is a function of the number of delinquent loans that progress to claim payment (which we refer to as frequency) and the amount to be paid to settle such claims (which we refer to as severity). Our estimates of claims frequency and severity are not formulaic, rather they are broadly synthesized based on historical observed experience for similarly situated loans and assumptions about future macroeconomic factors. We generally observe that forbearance programs are an effective tool to bridge dislocated borrowers from a time of acute stress to a future date when they can resume timely payment of their mortgage obligations. Our reserve setting process in future periods will consider this dynamic along with our expectations for house prices (and current views of general resiliency), interest rates and unemployment rates. Regulatory Capital Position As an approved mortgage insurer andWisconsin -domiciled carrier, we are required to satisfy financial and/or capitalization requirements stipulated by each of the GSEs and the Wisconsin OCI. The financial requirements stipulated by the GSEs are outlined in the PMIERs. Under the PMIERs, we must maintain available assets that are equal to or exceed a minimum risk-based required asset amount, subject to a minimum floor of$400 million . AtMarch 31, 2020 , we reported$1,070 million available assets against$912 million risk-based required assets for a$157 million of "excess" funding position. The risk-based required asset amount under PMIERs is determined at an individual policy-level based on the risk characteristics of each insured loan. Loans with higher risk factors, such as higher LTVs or lower borrower FICO scores, are assessed a higher charge. Non-performing loans that have missed two or more payments are generally assessed a significantly higher charge than performing loans, regardless of the underlying borrower or loan risk profile; however, special consideration is given under PMIERs to loans that are delinquent on homes located in an area declared by theFederal Emergency Management Agency (FEMA) to be a Major Disaster zone. The PMIERs charge on non-performing loans that enter delinquent status up to 30 days prior and 90 days after aFEMA -declared Major Disaster is adjusted by a 30% multiplier (inversely, a 70% haircut).FEMA has made a Major Disaster Declaration in all 50 states in response to the COVID-19 pandemic. As such, the PMIERs risk-based required asset charge for all newly delinquent loans nationwide (including those that go delinquent under a federal or private forbearance program) will be reduced by 70%. Our PMIERs minimum risk-based required asset amount is also adjusted for our reinsurance transactions (as approved by the GSEs). Under our quota share reinsurance treaties, we receive credit for the PMIERs risk-based required asset amount on 29 -------------------------------------------------------------------------------- ceded RIF. As our gross PMIERs risk-based required asset amount on ceded RIF increases, our PMIERS credit for ceded RIF automatically increases as well (in an unlimited amount). Under our ILN transactions, we generally receive credit for the PMIERs risk-based required asset amount on ceded RIF to the extent such requirement is within the subordinated coverage (excess of loss detachment threshold) afforded by the transaction. We have structured our ILN transactions to be overcollateralized, such that there are more ILN notes outstanding and cash held in trust than we currently receive credit for under the PMIERs. To the extent our PMIERs risk-based required asset amount on RIF ceded under the ILN transactions grows, we receive increased PMIERs credit under the treaties. The increasing PMIERs credit we receive under the ILN treaties is further enhanced by their delinquency lockout triggers. In the event delinquencies exceed 4% of ceded RIF, the ILN notes stop amortizing and the cash held in trust is secured for our benefit. As the underlying RIF continues to run-off, this has the effect of increasing the overcollateralization within, and excess PMIERs capacity provided by, each ILN structure. The following table provides detail on the level of overcollateralization of each of our ILN structures atMarch 31, 2020 : ($ values in thousands) 2017 ILN
Transaction 2018 ILN Transaction 2019 ILN Transaction Ceded RIF
$ 3,087,267 $ 3,853,559 $ 5,044,194
Current First Layer Retained Loss 122,810 124,311 123,424 Current Reinsurance Coverage 46,990 174,340 259,047 Eligible Coverage $ 169,800 $ 298,651 $ 382,471 Subordinated Coverage 5.50% 7.75% 7.50% PMIERs Charge on Ceded RIF 5.20% 6.63% 6.89% Overcollateralization $ 19,522 $ 43,192 $ 35,138 Delinquency Trigger 4.0% 4.0% 4.0% AtMarch 31, 2020 , we had an aggregate$98 million of overcollateralization available across our three ILN transactions to absorb an increase in the PMIERs risk-based required asset amount on ceded RIF. Assuming the 4% "delinquency lockout trigger" is activated in each deal and our underlying RIF continues to run-off at the same rate as it did during the month endedMarch 31, 2020 , we estimate that our total overcollateralization would increase by up to$70 million per quarter. Our PMIERs funding requirement will go up in future periods based on the volume and risk profile of our new business production, and performance of our in-force insurance portfolio. We estimate, however, that we will remain in compliance with our PMIERs asset requirements even if forbearance-driven default rates on our in-force portfolio materially exceed the current forbearance rates noted by Black Knight McDash for each of the GSEs and FHA/VA , given our$157 million excess available asset position atMarch 31, 2020 , the nationwide applicability of the 70%FEMA -disaster haircut on newly delinquent policies, the increasing PMIERs relief automatically provided under each of our quota share and ILN treaties, and our expectation for declining new business funding requirements (with lower NIW volume coming through at incrementally higher credit quality). NMIC is also subject to state regulatory minimum capital requirements based on its RIF. Formulations of this minimum capital vary by state, however, the most common measure allows for a maximum ratio of RIF to statutory capital (commonly referred to as "risk-to-capital" or "RTC") of 25:1. The RTC calculation does not assess a different charge or impose a different threshold RTC limit based on the underlying risk characteristics of the insured portfolio. Non-performing loans are treated the same as performing loans under the RTC framework. As such, the PMIERs generally imposes a stricter financial requirement than the state RTC standard, and we expect this to remain the case in the aftermath of the COVID-19 outbreak. Liquidity We evaluate our liquidity position at both a holding company (NMIH) and primary operating subsidiary (NMIC) level. As ofMarch 31, 2020 , we had$1.2 billion of consolidated cash and investments, including$44 million of cash and investments at NMIH. NMIH's principal liquidity demands include funds for the payment of (i) certain corporate expenses, (ii) certain reimbursable expenses of our insurance subsidiaries, including NMIC, and (iii) principal and interest as due on our outstanding debt. NMIH generates cash interest income on its investment portfolio and benefits from tax, expense-sharing and debt service agreements with its subsidiaries. Such agreements have been approved by the Wisconsin OCI and provide for the reimbursement of substantially all of NMIH's annual cash expenditures. While such agreements are subject to revocation by theWisconsin OCI, we do not expect such action to be taken at this time. The Wisconsin OCI has refreshed its approval of the debt service agreement providing for the additional reimbursement by NMIC of interest expense due under our amended and upsized 2020 Revolving 30 -------------------------------------------------------------------------------- Credit Facility. The Wisconsin OCI has also provided for the allocation of incremental holding company interest expenses to NMIC should we choose to pursue additional debt financing opportunities. NMIH also has access to$100 million of undrawn revolving credit capacity under the 2020 Revolving Credit Facility and$16.1 million of aggregate ordinary course dividend capacity available from NMIC and Re One without the prior approval of the Wisconsin OCI. Amounts drawn under the 2020 Revolving Credit Facility are available as directed for NMIH needs or may be down-streamed to support the requirements of our operating subsidiaries if we so decide. See Item 1, "Financial Statements - Notes to Condensed Consolidated Financial Statements - Note 4, Debt." NMIC's principal sources of liquidity include (i) premium receipts on its insured portfolio and new business production, (ii) interest income on its investment portfolio and principal repayments on maturities therein, and (iii) existing cash and cash equivalent holdings. AtMarch 31, 2020 , NMIC had$1.1 billion of cash and investments, including$76 million of cash and equivalents. NMIC's principal liquidity demands include funds for the payment of (i) reimbursable holding company expenses, (ii) premiums ceded under our reinsurance transactions, and (iii) claims payments. NMIC's cash inflow is generally significantly in excess of its cash outflow in any given period. During the twelve-month period endedMarch 31, 2020 , NMIC generated$216 million of cash flow from operations and received an additional$252 million of cash flow on the maturity, sale and redemption of securities held in its investment portfolio. NMIC is not a party to any contracts (derivative or otherwise) that require it to post an increasing amount of collateral to any counterparty and NMIC's principal liquidity demands (other than claims payments) generally develop along a scheduled path (i.e., are of a contractually predetermined amount and due at a contractually predetermined date). NMIC's only use of cash that develops along an unscheduled path is claims payments. Given the breadth and duration of forbearance programs available to borrowers, separate foreclosure moratoriums that have been enacted at a local, state and federal level, and the general duration of the default to foreclosure to claim cycle, we do not expect NMIC to pay meaningful cash claims in the near-term. Premiums paid to NMIC on monthly policies are generally collected and remitted by loan servicers. We have noted the broad discussion about the liquidity challenges loan servicers may themselves face in the event of widespread borrower utilization of forbearance programs. We do not currently believe that loan servicer liquidity issues will have a material impact on NMIC's premium receipts or liquidity profile. Loan servicers are contractually obligated to advance mortgage insurance premiums in a timely manner, even if the underlying borrowers fail to remit their monthly mortgage payments. We believe the GSEs are considering options to continue remitting premiums to us (and other mortgage insurers) on delinquent loans in the event a loan servicer fails to fulfill its obligation, including directly advancing such mortgage insurance premiums and/or moving the servicing obligation to a more adequately funded loan servicer with the capacity to continue making premium payments. Investment portfolio AtMarch 31, 2020 , we had$1.1 billion of invested assets and$110 million of cash and equivalents. Our investment strategy equally prioritizes capital preservation alongside income generation, and we have a long-established investment policy that sets conservative limits for asset types, industry sectors, single issuers and instrument credit ratings. AtMarch 31, 2020 , our investment portfolio was comprised of 100% fixed income assets with 100% of our holdings rated investment grade and our portfolio having an average rating of "A+." AtMarch 31, 2020 , our portfolio was highly liquid and highly diversified with no Level 3 asset positions and no single issuer concentration greater than 1.5%. Despite the severe sell-off generally seen across the capital markets, our investment portfolio was in a$11 million aggregate unrealized gain position at quarter end. We may experience future fluctuation in the value of our investment holdings, but did not record any allowance for credit losses in the portfolio in the quarter-endedMarch 31, 2020 , as we expect to recover the amortized cost basis of all securities held. The pre-tax book yield on our investment portfolio was 3.0% for the three months endedMarch 31, 2020 . We calculate book yield as annualized net investment income divided by the average amortized cost of the investment portfolio. Given the current interest rate environment and our focus on high-grade assets, the yield we are capturing on new investments is below that of the total portfolio. We expect our total portfolio yield will migrate down over time if this dynamic holds. Taxes The CARES Act, among other things, includes provisions relating to refundable payroll tax credits, deferment of employer side social security payments, net operating loss carryback periods, alternative minimum tax credit refunds, modifications to the net interest deduction limitations, increased limitations on qualified charitable contributions, and technical corrections to tax depreciation methods for qualified improvement property. We continue to monitor the impact that the CARES Act may have on our business, financial condition and results of operations. 31 --------------------------------------------------------------------------------
New Insurance Written, Insurance-In-Force and Risk-In-Force
NIW is the aggregate unpaid principal balance of mortgages underpinning new
policies written during a given period. Our NIW is affected by the overall size
of the mortgage origination market and the volume of high-LTV mortgage
originations. Our NIW is also affected by the percentage of such high-LTV
originations covered by private versus government MI or other alternative credit
enhancement structures and our share of the private MI market. NIW, together
with persistency, drives our IIF. IIF is the aggregate unpaid principal balance
of the mortgages we insure, as reported to us by servicers at a given date, and
represents the sum total of NIW from all prior periods less principal payments
on insured mortgages and policy cancellations (including for prepayment,
nonpayment of premiums, coverage rescission and claim payments). RIF is related
to IIF and represents the aggregate amount of coverage we provide on all
outstanding policies at a given date. RIF is calculated as the sum total of the
coverage percentage of each individual policy in our portfolio applied to the
unpaid principal balance of such insured mortgage. RIF is affected by IIF and
the LTV profile of our insured mortgages, with lower LTV loans generally having
a lower coverage percentage and higher LTV loans having a higher coverage
percentage. Gross RIF represents RIF before consideration of reinsurance. Net
RIF is gross RIF net of ceded reinsurance.
Net Premiums Written and Net Premiums Earned
We set our premium rates on individual policies based on the risk
characteristics of the underlying mortgage loans and borrowers, and in
accordance with our filed rates and applicable rating rules. On June 4, 2018 , we
introduced a proprietary risk-based pricing platform, which we refer to as Rate
GPS. Rate GPS considers a broad range of individual variables, including
property type, type of loan product, borrower credit characteristics, and lender
and market factors, and provides us with the ability to set and charge premium
rates commensurate with the underlying risk of each loan that we insure. We
introduced Rate GPS in June 2018 to replace our previous rate card pricing
system. While most of our new business is priced through Rate GPS, we also
continue to offer a rate card pricing option to a limited number of lender
customers who require a rate card for operational reasons. We believe the
introduction and utilization of Rate GPS provides us with a more granular and
analytical approach to evaluating and pricing risk, and that this approach
enhances our ability to continue building a high-quality mortgage insurance
portfolio and delivering attractive risk-adjusted returns.
Premiums are generally fixed for the duration of our coverage of the underlying
loans. Net premiums written are equal to gross premiums written minus ceded
premiums written under our reinsurance arrangements, less premium refunds and
premium write-offs. As a result, net premiums written are generally influenced
by:
•NIW;
•premium rates and the mix of premium payment type, which are either single,
monthly or annual premiums, as described below;
•cancellation rates of our insurance policies, which are impacted by payments or
prepayments on mortgages, refinancings (which are affected by prevailing
mortgage interest rates as compared to interest rates on loans underpinning our
in force policies), levels of claim payments and home prices; and
•cession of premiums under third-party reinsurance arrangements.
Premiums are paid either by the borrower (BPMI) or the lender (LPMI) in a single
payment at origination (single premium), on a monthly installment basis (monthly
premium) or on an annual installment basis (annual premium). Our net premiums
written will differ from our net premiums earned due to policy payment type. For
single premiums, we receive a single premium payment at origination, which is
earned over the estimated life of the policy. Substantially all of our single
premium policies in force as of March 31, 2020 were non-refundable under most
cancellation scenarios. If non-refundable single premium policies are canceled,
we immediately recognize the remaining unearned premium balances as earned
premium revenue. Monthly premiums are recognized in the month billed and when
the coverage is effective. Annual premiums are earned on a straight-line basis
over the year of coverage. Substantially all of our policies provide for either
single or monthly premiums.
The percentage of IIF that remains on our books after any 12-month period is
defined as our persistency rate. Because our insurance premiums are earned over
the life of a policy, higher persistency rates can have a significant impact on
our net premiums earned and profitability. Generally, faster speeds of mortgage
prepayment lead to lower persistency. Prepayment speeds and the relative mix of
business between single and monthly premium policies also impact our
profitability. Our premium rates include certain assumptions regarding repayment
or prepayment speeds of the mortgages underlying our policies. Because premiums
are paid at origination on single premium policies and our single premium
policies are generally non-refundable on cancellation, assuming all other
factors remain constant, if single premium loans are repaid earlier than
expected, our profitability
32
--------------------------------------------------------------------------------
on these loans is likely to increase and, if loans are repaid slower than
expected, our profitability on these loans is likely to decrease. By contrast,
if monthly premium loans are repaid earlier than anticipated, we do not earn any
more premium with respect to those loans and, unless we replace the repaid
monthly premium loan with a new loan at the same premium rate or higher, our
revenue is likely to decline.
Effect of reinsurance on our results
We utilize third-party reinsurance to actively manage our risk, ensure
compliance with PMIERs, state regulatory and other applicable capital
requirements, and support the growth of our business. We currently have both
quota share and excess-of-loss reinsurance agreements in place, which impact our
results of operations and regulatory capital and PMIERs asset positions. Under a
quota share reinsurance agreement, the reinsurer receives a premium in exchange
for covering an agreed-upon portion of incurred losses. Such a quota share
arrangement reduces premiums written and earned and also reduces RIF, providing
capital relief us and reducing our incurred claims in accordance with the terms
of the reinsurance agreement. In addition, reinsurers typically pay ceding
commissions as part of quota share transactions, which offset our acquisition
and underwriting expenses. Certain quota share agreements include profit
commissions that are earned based on loss performance and serve to reduce ceded
premiums. Under an excess-of-loss agreement, typically we are responsible for
losses up to an agreed-upon threshold and the reinsurer then provides coverage
in excess of such threshold up to a maximum agreed-upon limit. We expect to
continue to evaluate reinsurance opportunities in the normal course of business.
Quota share reinsurance
Effective September 1, 2016 , NMIC entered into the 2016 QSR Transaction with a
syndicate of third-party reinsurers. Under the terms of the 2016 QSR
Transaction, NMIC cedes premiums written related to 25% of the risk on eligible
primary policies written for all periods through December 31, 2017 and 100% of
the risk under our pool agreement with Fannie Mae, in exchange for reimbursement
of ceded claims and claim expenses on covered policies, a 20% ceding commission,
and a profit commission of up to 60% that varies directly and inversely with
ceded claims.
Effective January 1, 2018 , NMIC entered into the 2018 QSR Transaction with a
syndicate of third-party reinsurers. Under the 2018 QSR Transaction, NMIC cedes
premiums earned related to 25% of the risk on eligible policies written in 2018
and 20% of the risk on eligible policies written in 2019, in exchange for
reimbursement of ceded claims and claim expenses on covered policies, a 20%
ceding commission, and a profit commission of up to 61% that varies directly and
inversely with ceded claims.
Under the terms of the QSR Transactions, NMIC may elect to selectively terminate
its engagement with individual reinsurers on a run-off basis (i.e., reinsurers
continue providing coverage on all risk ceded prior to the termination date,
with no new cessions going forward) or cut-off basis (i.e., the reinsurance
arrangement is completely terminated with NMIC recapturing all previously ceded
risk) under certain circumstances. Such selective termination rights arise
when, among other reasons, a reinsurer experiences a deterioration in its
capital position below a prescribed threshold and/or a reinsurer breaches (and
fails to cure) its collateral posting obligations under the relevant agreement.
Effective April 1, 2019 , NMIC elected to terminate its engagement with one
reinsurer under the 2016 QSR Transaction on a cut-off basis. In connection with
the termination, NMIC recaptured approximately $500 million of previously ceded
primary RIF and stopped ceding new premiums written with respect to the
recaptured risk. With this termination, ceded premiums written under the 2016
QSR Transaction decreased from 25% to 20.5% on eligible policies. The
termination had no effect on the cession of pool risk under the 2016 QSR
Transaction.
Excess-of-loss reinsurance
NMIC has secured aggregate excess-of-loss reinsurance coverage on defined
portfolios of mortgage insurance policies written during discrete periods
through a series of mortgage insurance-linked note offerings by the Oaktown Re
Vehicles. Under each agreement, NMIC retains a first layer of aggregate loss
exposure on covered policies and the respective Oaktown Re Vehicle then provides
second layer loss protection up to a defined reinsurance coverage amount. NMIC
then retains losses in excess of the respective reinsurance coverage amounts.
NMIC applies claims paid on covered policies against its first layer aggregate
retained loss exposure under each excess-of-loss agreement. The respective
reinsurance coverage amounts provided by the Oaktown Re Vehicles generally
decrease from the inception of each agreement over a 10-year period as the
underlying insured mortgages are amortized or repaid, and/or the mortgage
insurance coverage is canceled. The respective outstanding reinsurance coverage
amounts stop amortizing if certain credit enhancement or delinquency thresholds,
as defined in each agreement are triggered, referred to as a lock-out feature.
When
33
--------------------------------------------------------------------------------
a lock-out trigger is breached, the respective ILN notes stop amortizing and the
assets held in trust are frozen for our benefit while the underlying risk
continues to run-off.
The following table presents the inception date, covered production period,
initial and current reinsurance coverage amount, and initial and current first
layer retained aggregate loss under each of the ILN Transactions.
Initial Initial First Current First
Reinsurance Current Reinsurance Layer Retained Layer Retained
($ values in thousands) Inception Date Covered Production Coverage Coverage Loss Loss
2017 ILN Transaction May 2, 2017 1/1/2013 - 12/31/2016 $ 211,320 $ 46,990 $ 126,793 $ 122,810
2018 ILN Transaction July 25, 2018 1/1/2017 - 5/31/2018 264,545 174,340 125,312 124,311
2019 ILN Transaction July 30, 2019 6/1/2018 - 6/30/2019 326,905 259,047 123,424 123,424
See Item 1, "Financial Statements - Notes to Condensed Consolidated Financial
Statements - Note 5, Reinsurance" for further discussion of these third-party
reinsurance arrangements.
Portfolio Data
The following table presents primary and pool NIW and IIF as of the dates
and for the periods indicated. Unless otherwise noted, the tables below do not
include the effects of our third-party reinsurance arrangements described above.
Primary and pool IIF and NIW As of and for the three months ended
March 31, 2020 March 31, 2019
IIF NIW IIF NIW
(In Millions)
Monthly $ 81,347 $ 10,461 $ 55,995 $ 6,211
Single 17,147 836 17,239 702
Primary 98,494 11,297 73,234 6,913
Pool 2,487 - 2,838 -
Total $ 100,981 $ 11,297 $ 76,072 $ 6,913
For the three months ended March 31, 2020 , primary NIW increased 63%
compared to the three months ended March 31, 2019 , primarily due to growth in
our monthly policy production tied to increased penetration of existing customer
accounts, new customer account activations and a larger mortgage insurance
market. The increase also contributed to the growth in single premium policy
production. For the three months ended March 31, 2020 , monthly primary NIW
increased 68% compared to the three months ended March 31, 2019 .
For the three months ended March 31, 2020 , monthly premium polices accounted
for 93% of our NIW. As of March 31, 2020 , monthly premium policies accounted for
83% of our primary IIF, as compared to 76% at March 31, 2019 . We expect the
break-down of monthly premium policies and single premium policies (which we
refer to as "mix") in our primary IIF will continue to trend toward an increased
monthly mix over time given the composition of our NIW.
Total IIF increased 33% as of March 31, 2020 compared to March 31, 2019 ,
primarily due to the NIW generated between such measurement dates, partially
offset by the run-off of our in-force policies. Our persistency rate decreased
to 71.7% at March 31, 2020 from 87.2% at March 31, 2019 , reflecting the impact
of increased refinancing activity tied to record low interest rates.
34
--------------------------------------------------------------------------------
The following table presents net premiums written and earned for the periods
indicated.
Primary and pool premiums written and earned For the three months ended
March
31, 2020
(In Thousands)
Net premiums written $ 91,371 $ 71,923
Net premiums earned 98,717 73,868
For the three months ended March 31, 2020 , net premiums written and earned
increased 27% and 34%, respectively, compared to the three months ended
March 31, 2019 . The increases in net premiums written and earned are primarily
due to the growth of our IIF and increased monthly policy production, partially
offset by increased cessions under the QSR Transactions tied to the growth of
our direct premium volume and the inception of the 2019 ILN Transaction. Net
premiums earned grew at an accelerated pace compared to net premiums written due
to the amortization of unearned premiums and earnings on single premium policy
cancellations, partially offset by new single premium policy originations.
Pool premiums written and earned for the three months ended March 31, 2020
and 2019, were $0.7 million and $0.8 million before giving effect to the 2016
QSR Transaction, under which all of our written and earned pool premiums have
been ceded. A portion of our ceded pool premiums earned are recouped through
profit commission under the 2016 QSR Transaction.
Portfolio Statistics
Unless otherwise noted, the portfolio statistics tables presented below do not
include the effects of our third-party reinsurance arrangements described above.
The table below highlights trends in our primary portfolio as of the dates and
for the periods indicated.
Primary portfolio trends As of and for the three months ended
December 31, September 30,
March 31, 2020 2019 2019 June 30, 2019 March 31, 2019
($ Values In Millions)
New insurance written $ 11,297 $ 11,949 $ 14,100 $ 12,179 $ 6,913
Percentage of monthly premium 93 % 93 % 92 % 91 % 90 %
Percentage of single premium 7 % 7 % 8 % 9 % 10 %
New risk written $ 2,897 $ 3,082 $ 3,651 $ 3,183 $ 1,799
Insurance-in-force (1) 98,494 94,754 89,713 81,708 73,234
Percentage of monthly premium 83 % 81 % 80 % 78 % 76 %
Percentage of single premium 17 % 19 % 20 % 22 % 24 %
Risk-in-force (1) $ 25,192 $ 24,173 $ 22,810 $ 20,661 $ 18,373
Policies in force (count) (1) 376,852 366,039 350,395 324,876 297,232
Average loan size (1) $ 0.261 $ 0.259 $ 0.256 $ 0.252 $ 0.246
Coverage percentage (2) 25.6 % 25.5 % 25.4 % 25.3 % 25.1 %
Loans in default (count) (1) 1,449 1,448 1,230 1,028 940
Percentage of loans in default (1) 0.38 % 0.40 % 0.35 % 0.32 % 0.32 %
Risk-in-force on defaulted loans (1) $ 84 $ 84 $ 70 $ 58 $ 53
Average premium yield (3) 0.41 % 0.41 % 0.43 % 0.43 % 0.42 %
Earnings from cancellations $ 8.6 $ 8.0 $ 7.4 $ 4.5 $ 2.3
Annual persistency (4) 71.7 % 76.8 % 82.4 % 86.0 % 87.2 %
Quarterly run-off (5) 8.0 % 7.7 % 7.5 % 5.1 % 3.3 %
(1) Reported as of the end of the period.
(2) Calculated as end of period RIF divided by end of period IIF.
(3) Calculated as net premiums earned divided by average primary IIF for the
period, annualized.
(4) Defined as the percentage of IIF that remains on our books after a given
12-month period.
(5) Defined as the percentage of IIF that is no longer on our books after a
given three month period.
35
--------------------------------------------------------------------------------
The table below presents a summary of the change in total primary IIF for
the dates and periods indicated.
Primary IIF For the three months ended
(In Millions)
IIF, beginning of period $ 94,754 $ 68,551
NIW 11,297 6,913
Cancellations, principal repayments and other reductions (7,557) (2,230)
IIF, end of period $ 98,494 $ 73,234 We consider a "book" to be a collective pool of policies insured during a
particular period, normally a calendar year. In general, the majority of
underwriting profit, calculated as earned premium revenue minus claims and
underwriting and operating expenses, generated by a particular book year emerges
in the years immediately following origination. This pattern generally occurs
because relatively few of the claims that a book will ultimately experience
typically occur in the first few years following origination, when premium
revenue is highest, while subsequent years are affected by declining premium
revenues, as the number of insured loans decreases (primarily due to loan
prepayments), and by increasing losses.
The table below presents a summary of our primary IIF and RIF by book year
as of the dates indicated.
Primary IIF and RIF As of March 31, 2020 As of March 31, 2019
IIF RIF IIF RIF
(In Millions)
March 31, 2020 $ 11,236 $ 2,882 $ - $ -
2019 39,485 10,259 6,872 1,789
2018 17,545 4,464 25,609 6,492
2017 13,656 3,398 18,353 4,514
2016 10,962 2,763 14,750 3,652
2015 and before 5,610 1,426 7,650 1,926
Total $ 98,494 $ 25,192 $ 73,234 $ 18,373
We utilize certain risk principles that form the basis of how we underwrite and
originate NIW. We have established prudential underwriting standards and
loan-level eligibility matrices which prescribe the maximum LTV, minimum
borrower FICO score, maximum borrower debt-to-income (DTI) ratio, maximum loan
size, property type, loan type, loan term and occupancy status of loans that we
will insure and memorialized these standards and eligibility matrices in our
Underwriting Guideline Manual that is publicly available on our website. Our
underwriting standards and eligibility criteria are designed to limit the
layering of risk in a single insurance policy. "Layered risk" refers to the
accumulation of borrower, loan and property risk. For example, we have higher
credit score and lower maximum allowed LTV requirements for investor-owned
properties, compared to owner-occupied properties. We monitor the concentrations
of various risk attributes in our insurance portfolio, which may change over
time, in part, as a result of regional conditions or public policy shifts.
36
--------------------------------------------------------------------------------
The tables below present our primary NIW by FICO, LTV and purchase/refinance
mix for the periods indicated. We calculate the LTV of a loan as the percentage
of the original loan amount to the original purchase value of the property
securing the loan.
Primary NIW by FICO For the three months ended
March 31, 2020 March 31, 2019
(In Millions)
>= 760 $ 6,290 $ 3,057
740-759 1,615 1,224
720-739 1,579 1,044
700-719 1,038 792
680-699 565 553
<=679 210 243
Total $ 11,297 $ 6,913
Weighted average FICO 757 749
Primary NIW by LTV For the three months ended
March 31, 2020 March 31, 2019
(In Millions)
95.01% and above $ 721 $ 569
90.01% to 95.00% 5,009 3,424
85.01% to 90.00% 4,082 2,241
85.00% and below 1,485 679
Total $ 11,297 $ 6,913
Weighted average LTV 91.3 % 92.2 %
Primary NIW by purchase/refinance mix For the three months ended
March 31, 2020 March 31, 2019
(In Millions)
Purchase $ 7,991 $ 6,383
Refinance (1) 3,306 530
Total $ 11,297 $ 6,913 (1) The amount of cash-out refinance loans insured in our portfolio was de minimis for the periods presented
37
--------------------------------------------------------------------------------
The tables below present our total primary IIF and RIF by FICO and LTV and total
primary RIF by loan type as of the dates indicated.
Primary IIF by FICO As of
March 31, 2020 March 31, 2019
($ Values In Millions)
>= 760 $ 47,340 48 % $ 33,902 46 %
740-759 16,060 16 12,160 17
720-739 14,002 14 10,096 14
700-719 10,518 11 8,122 11
680-699 6,879 7 5,435 7
<=679 3,695 4 3,519 5
Total $ 98,494 100 % $ 73,234 100 %
Weighted average FICO 751 749
Primary RIF by FICO As of
March 31, 2020 March 31, 2019
($ Values In Millions)
>= 760 $ 12,076 48 % $ 8,506 46 %
740-759 4,121 16 3,076 17
720-739 3,626 14 2,550 14
700-719 2,696 11 2,036 11
680-699 1,760 7 1,357 7
<=679 (1) 913 4 848 5
Total $ 25,192 100 % $ 18,373 100 %
Weighted average FICO 751 749
(1) There were no loans with a FICO <=620 insured in our portfolio for the
periods presented.
Primary IIF by LTV As of
March 31, 2020 March 31, 2019
($ Values In Millions)
95.01% and above $ 8,838 9 % $ 7,204 10 %
90.01% to 95.00% 46,318 47 34,024 46
85.01% to 90.00% 31,729 32 22,208 30
85.00% and below 11,609 12 9,798 14
Total $ 98,494 100 % $ 73,234 100 %
Weighted average LTV 91.9 % 91.9 %
38--------------------------------------------------------------------------------
Primary RIF by LTV As of
March 31, 2020 March 31, 2019
($ Values In Millions)
95.01% and above $ 2,478 10 % $ 1,928 10 %
90.01% to 95.00% 13,587 54 9,923 54
85.01% to 90.00% 7,767 31 5,384 30
85.00% and below 1,360 5 1,138 6
Total $ 25,192 100 % $ 18,373 100 %
Weighted average LTV 92.7 % 92.7 %
Primary RIF by Loan Type As of
March 31, 2020 March 31, 2019
Fixed 98 % 98 %
Adjustable rate mortgages
Less than five years - -
Five years and longer 2 2
Total (1) 100 % 100 %
(1) There were no interest-only mortgages insured in our portfolio for the
periods presented.
The table below presents selected primary portfolio statistics, by book year, as
of March 31, 2020 .
As of March 31, 2020
Original Remaining % Remaining of Number of Number of Incurred Loss Current
Insurance Insurance in Original Policies Ever in Policies in Loans in # of
Claims Ratio (Inception Cumulative Default Rate
Book year Written Force Insurance Force Force Default Paid to Date) (1) Default Rate (2) (3)
($ Values in Millions)
2013 $ 162 $ 20 12 % 655 115 - 1 0.2 % 0.2 % - %
2014 3,451 747 22 % 14,786 4,081 40 44 4.1 % 0.6 % 1.0 %
2015 12,422 4,843 39 % 52,548 23,277 158 97 3.0 % 0.5 % 0.7 %
2016 21,187 10,962 52 % 83,626 47,687 254 97 2.4 % 0.4 % 0.5 %
2017 21,582 13,656 63 % 85,897 59,356 441 53 3.4 % 0.6 % 0.7 %
2018 27,295 17,545 64 % 104,043 73,620 429 24 4.6 % 0.4 % 0.6 %
2019 45,141 39,485 87 % 148,423 133,291 127 - 2.3 % 0.1 % 0.1 %
2020 11,297 11,236 99 % 35,581 35,425 - - - % - % - %
Total $ 142,537 $ 98,494 525,559 376,852 1,449 316
(1) Calculated as total claims incurred (paid and reserved) divided by
cumulative premiums earned, net of reinsurance.
(2) Calculated as the sum of the number of claims paid ever to date and number
of loans in default divided by policies ever in force.
(3) Calculated as the number of loans in default divided by number of policies
in force.
39
--------------------------------------------------------------------------------Geographic Dispersion
The following table shows the distribution by state of our primary RIF as of
the periods indicated. As of March 31, 2020 , our RIF continues, although
declining, to be relatively more concentrated in California , primarily as a
result of the size of the California mortgage market relative to the rest of the
country and the location and timing of our acquisition of new customers. The
distribution of our primary RIF as of March 31, 2020 is not necessarily
representative of the geographic distribution we expect in the future.
Top 10 primary RIF by state As of
March 31, 2020 March 31, 2019
California 11.5 % 12.7 %
Texas 8.2 8.3
Florida 5.9 5.2
Virginia 5.3 5.0
Illinois 3.8 3.4
Arizona 3.7 4.8
Pennsylvania 3.7 3.6
Colorado 3.6 3.4
Michigan 3.4 3.6
Maryland 3.4 3.2
Total 52.5 % 53.2 %
Insurance Claims and Claim Expenses
Insurance claims and claim expenses represent estimated future payments on newly
defaulted insured loans and any change in our claim estimates for previously
existing defaults. Claims and claim expenses are generally affected by a variety
of factors, including the macroeconomic environment, national and regional
unemployment trends, changes in housing values, borrower risk characteristics,
LTV ratios and other loan level risk attributes, the size and type of loans
insured, and the percentage of coverage on insured loans.
Reserves for claims and allocated claim expenses are established for reported
mortgage loan defaults, which we refer to as case reserves, when we are notified
that a borrower has missed two or more mortgage payments (i.e., an NOD). We also
make estimates of IBNR defaults, which are defaults that have been incurred but
have not been reported by loan servicers, based on historical reporting trends,
and establish IBNR reserves for those defaults. We also establish reserves for
unallocated claim expenses not associated with a specific claim. Claim expenses
consist of the estimated cost of the claim administration process, including
legal and other fees as well as other general expenses of administering the
claim settlement process.
Reserves are established by estimating the number of loans in default that will
result in a claim payment, which is referred to as claim frequency, and the
amount of the claim payment expected to be paid on each such loan in default,
which is referred to as claim severity. Claim frequency and severity estimates
are established based on historical observed experience regarding certain loan
factors, such as age of the default, cure rates, size of the loan and estimated
change in property value. Reserves are released the month in which a loan in
default is brought current by the borrower, which is referred to as a cure.
Adjustments to reserve estimates are reflected in the period in which the
adjustment is made. Reserves are also ceded to reinsurers under the QSR
Transactions. We will not cede claims under the ILN Transactions unless losses
exceed the respective retained coverage layers. Reserves are not established for
future claims on insured loans which are not currently in default. Our pool
insurance agreement with Fannie Mae contains a claim deductible through which
Fannie Mae absorbs specified losses before we are obligated to pay any claims.
We have not established any claims or claim expense reserves for pool exposure
to date.
Based on our experience and industry data, we believe that claims incidence for
mortgage insurance is generally highest in the third through sixth years after
loan origination. As of March 31, 2020 , approximately 81% of our primary IIF
related to business written since March 31, 2017 . Although the claims experience
on our IIF to date has been modest, we expect incurred claims to increase as a
greater amount of our existing insured portfolio reaches its anticipated period
of highest claim frequency.
The actual claims we incur as our portfolio matures are difficult to predict and
depend on the specific characteristics of our current in-force book (including
the credit score and DTI of the borrower, the LTV ratio of the mortgage and
geographic
40
--------------------------------------------------------------------------------
concentrations, among others), as well as the risk profile of new business we
write in the future. In addition, claims experience will be affected by
macroeconomic factors such as housing prices, interest rates, unemployment rates
and other events, such as natural disasters.
The following table provides a reconciliation of the beginning and ending reserve balances for primary insurance claims and claim expenses.
For the three months ended
March
31, 2020
(In Thousands)
Beginning balance $ 23,752 $ 12,811
Less reinsurance recoverables (1) (4,939) (3,001)
Beginning balance, net of reinsurance recoverables 18,813 9,810
Add claims incurred:
Claims and claim expenses incurred:
Current year (2) 7,558 3,909
Prior years (3) (1,861) (1,166)
Total claims and claim expenses incurred 5,697 2,743
Less claims paid:
Claims and claim expenses paid:
Current year (2) - -
Prior years (3) 1,224 694
Total claims and claim expenses paid 1,224 694
Reserve at end of period, net of reinsurance recoverables 23,286 11,859
Add reinsurance recoverables (1) 6,193 3,678
Ending balance $ 29,479 $ 15,537
(1) Related to ceded losses recoverable under the QSR Transactions, included in
"Other assets" on the condensed consolidated balance sheets. See Item 1,
"Financial Statements - Notes to Condensed Consolidated Financial Statements -
Note 5, Reinsurance" for additional information.
(2) Related to insured loans with their most recent defaults occurring in the
current year. For example, if a loan had defaulted in a prior year and
subsequently cured and later re-defaulted in the current year, that default
would be included in the current year. Amounts are presented net of reinsurance.
(3) Related to insured loans with defaults occurring in prior years, which have
been continuously in default before the start of the current year. Amounts are
presented net of reinsurance.
The "claims incurred" section of the table above shows claims and claim
expenses incurred on NODs for current and prior years, including IBNR reserves
and is presented net of reinsurance. The amount of claims incurred for current
year NODs represents the estimated amount to be ultimately paid on such loans in
default. The decreases during the periods presented in reserves held for prior
year defaults represent favorable development and are generally the result of
NOD cures. We perform an ongoing analysis of recent loss development trends and
may increase or decrease our original estimates as we gather additional
information about individual defaults and claims and continue to observe and
analyze loss development trends in our portfolio, including factors affected by
the recent COVID-19 developments. Gross reserves of $19.9 million related to
prior year defaults remained as of March 31, 2020 .
41
--------------------------------------------------------------------------------
The following table provides a reconciliation of the beginning and ending
count of loans in default.
For the three months ended
March 31, 2020 March 31, 2019
Beginning default inventory 1,448 877
Plus: new defaults 512 574
Less: cures (475) (474)
Less: claims paid (34) (37)
Less: claims denied (2) -
Ending default inventory 1,449 940
The increase in the ending default inventory at March 31, 2020 compared to
March 31, 2019 , primarily relates to an increase in new defaults tied to the
growth in the number of policies in force and the aging of our earlier book
years. Our total policies in force were 376,852 and 297,232 for the three months
ended March 31, 2020 and March 31, 2019 , respectively. The increase in new
defaults was partially offset by cure activity on our beginning NOD population.
The following table provides details of our claims paid, before giving
effect to claims ceded under the QSR Transactions, for the periods indicated.
For the three months ended
March 31, 2020 March 31, 2019
($ In Thousands)
Number of claims paid (1) 34 37
Total amount paid for claims $ 1,503 $ 926
Average amount paid per claim $ 44 $ 25
Severity (2) 83 % 64 %
(1) Count includes one claim settled without payment for the three months ended
March 31, 2020 , and three claims settled without payment for the three months
ended March 31, 2019 .
(2) Severity represents the total amount of claims paid including claim expenses
divided by the related RIF on the loan at the time the claim is perfected, and
is calculated including claims settled without payment.
The number of claims paid for the three months ended March 31, 2020
decreased compared to the three months ended March 31, 2019 , primarily due to
the timing of when claims were received and paid in respect to respective period
end. Our claims severity for the three months ended March 31, 2020 was 83%,
compared to 64% for the three months ended March 31, 2019 . The increase in
claims severity for the three months ended March 31, 2020 can be attributed to
fewer claims settled with zero payment compared to the three months ended March
31, 2019 .
The following table provides detail on our average reserve per default,
before giving effect to reserves ceded under the QSR Transactions, as of the
dates indicated.
Average reserve per default: As of March 31, 2020 As of March 31, 2019
(In Thousands)
Case (1) $ 18 $ 15
IBNR(2) 2 2
Total $ 20 $ 17
(1) Defined as the gross reserve per insured loan in default. (2) Amount includes claims adjustment expenses.
The average reserve per default at March 31, 2020 increased from March 31,
2019 , primarily tied to an increase in claim rates resulting from updates to key
assumptions as a result of COVID-19 and the aging our NOD population.
42
--------------------------------------------------------------------------------
GSE Oversight
As an approved insurer, NMIC is subject to ongoing compliance with the PMIERs
established by each of the GSEs (italicized terms have the same meaning that
such terms have in the PMIERs, as described below). The PMIERs establish
operational, business, remedial and financial requirements applicable to
approved insurers. The PMIERs financial requirements prescribe a risk-based
methodology whereby the amount of assets required to be held against each
insured loan is determined based on certain loan-level risk characteristics,
such as FICO, vintage (year of origination), performing vs. non-performing
(i.e., current vs. delinquent), LTV ratio and other risk features. In general,
higher quality loans carry lower asset charges.
Under the PMIERs, approved insurers must maintain available assets that equal or
exceed minimum required assets, which is an amount equal to the greater of (i)
$400 million or (ii) a total risk-based required asset amount. The risk-based
required asset amount is a function of the risk profile of an approved insurer's
RIF, assessed on a loan-by-loan basis and considered against certain risk-based
factors derived from tables set out in the PMIERs, which is then adjusted on an
aggregate basis for reinsurance transactions approved by the GSEs, such as with
respect to our ILN Transactions and QSR Transactions. The aggregate gross
risk-based required asset amount for performing, primary insurance is subject to
a floor of 5.6% of performing primary adjusted RIF, and the risk-based required
asset amount for pool insurance considers both factors in the PMIERs tables and
the net remaining stop loss for each pool insurance policy.
By April 15th of each year, NMIC must certify it met all PMIERs requirements
as of December 31st of the prior year. We certified to the GSEs by April 15,
2020 that NMIC was in full compliance with the PMIERs as of December 31, 2019 .
NMIC also has an ongoing obligation to immediately notify the GSEs in writing
upon discovery of a failure to meet one or more of the PMIERs requirements. We
continuously monitor NMIC's compliance with the PMIERs.
The following table provides a comparison of the PMIERs available assets and
risk-based required asset amount as reported by NMIC as of the dates indicated.
As of
March 31, 2020 March 31, 2019
(In Thousands)
Available assets $ 1,069,695 $ 817,758
Risk-based required assets 912,321 607,325
Available assets were $1,070 million at March 31, 2020 , compared to $818 million
at March 31, 2019 . The increase in Available assets of $252 million was
primarily driven by our positive cash flow from operations in the intervening
periods.
The increase in the risk-based required asset amount was due to the growth of
our gross RIF, partially offset by the increased cession of risk under our
third-party reinsurance agreements. The risk-based required asset amount further
increased in 2020 as a result of the termination of our engagement with and the
recapture of previously ceded primary RIF from one reinsurer under the 2016 QSR
Transaction effective April 1, 2019 .
LIBOR Transition
In July 2017 , the U.K. Financial Conduct Authority announced that it intends to
stop persuading or compelling banks to submit LIBOR rates after 2021, which is
expected to render these widely used reference rates unavailable or unreliable.
We have exposure to LIBOR-based financial instruments, such as LIBOR-based
securities held in our investment portfolio. The 2018 Term Loan, 2020 Revolving
Credit Facility, and premiums paid on our ILN Transactions are also LIBOR-based.
We are in the process of reviewing our LIBOR-based contracts that extend beyond
2021 and transitioning to a set of alternative reference rates. We will continue
to monitor, assess and plan for the phase out of LIBOR; however, we currently do
not know the impact it will have to our operations or financial results.
Capital Position of Our Insurance Subsidiaries and Financial Strength Ratings
In addition to GSE-imposed asset requirements, NMIC is subject to state
regulatory minimum capital requirements based on its RIF. While formulations of
this minimum capital may vary by jurisdiction, the most common measure allows
for a
43
--------------------------------------------------------------------------------
maximum permitted RTC ratio of 25:1.
As of March 31, 2020 , NMIC's performing primary RIF, net of reinsurance, was
approximately $17.9 billion . NMIC ceded 100% of its pool RIF pursuant to the
2016 QSR Transaction. Based on NMIC's total statutory capital of $1.0 billion
(including contingency reserves) as of March 31, 2020 , NMIC's RTC ratio was
17.7:1. Re One had total statutory capital of $36 million as of March 31, 2020 ,
and a RTC ratio of 1.3:1. We continuously monitor our compliance with state
capital requirements.
In October 2019 , Moody's Investors Service (Moody's) upgraded its financial
strength rating of NMIC from "Baa3" to "Baa2" and upgraded its rating of NMIH's
$150 million senior secured 2018 Term Loan and 2018 Revolving Credit Facility
from "Ba3" to "Ba2". The outlook for Moody's ratings is stable. In June 2019 ,
S&P upgraded its financial strength and long-term counter-party credit ratings
of NMIC from "BBB-" to "BBB" and upgraded its long-term counter-party credit
rating of NMIH from "BB-" to "BB". In March 2020 , the S&P updated its outlook
from stable to negative for the mortgage insurance sector, including NMIH.
Information and Technology Support Function
Effective March 31, 2020 , we entered into a services agreement to outsource our
IT support and expertise functions to Tata Consultancy Services (TCS). The
agreement provides for TCS to provide critical IT services over a seven-year
period including such functions as application development and support,
infrastructure support, information security, transition and transformation. We
expect this engagement will allow us to continue our strategy to provide
innovative development and IT optimization, as the business and industry
continues to evolve, while realizing cost efficiencies by leveraging TCS's
resources and operating model. As a result of this agreement, the majority of
our IT employees will become employees of TCS during the second quarter of 2020,
with guaranteed continued employment for a 12-month period effective from the
date of transition. If we were to terminate the services agreement with TCS at
any point before that 12-month period has expired, we would be obligated to
rehire the affected employees for the remainder of their guaranteed employment
period. In addition, the agreement provides that we may terminate it at any time
with 120 days' notice, subject to our payment of a termination fee as specified
in the agreement.
Competition
The MI industry is highly competitive and currently consists of six private mortgage insurers, including NMIC, as well as government MIs such as the FHA,USDA orVA . Private MI companies compete based on service, customer relationships, underwriting and other factors, including price, credit risk tolerance and information technology capabilities. We expect the private MI market to remain competitive, with pressure for industry participants to maintain or grow their market share. The private MI industry overall competes more broadly with government MIs who significantly increased their share in the MI market following the 2008 financial crisis. Although there has been broad policy consensus toward the need for increasing private capital participation and decreasing government exposure to credit risk in theU.S. housing finance system, it remains difficult to predict whether the combined market share of government MIs will recede to pre-2008 levels. A range of factors influence a lender's and borrower's decision to choose private over government MI, including among others, GSE demand, premium rates and other charges, loan eligibility requirements, cancelability, loan size limits and the relative ease of use of private MI products compared to government MI alternatives. 44
--------------------------------------------------------------------------------
© Edgar Online, source
