## _cr07163 ## Source details **Canonical URL:** [_cr07163](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07163.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07163.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07163.pdf.json) --- ### Introduction - Paper analyzes performance and development of the Mexican pension annuity market stemming from the 1997 pension reform (SAR-97). - Key observations: - Mexican experience offers lessons for decumulation design of second-pillar systems. - Identifies technical and policy concerns that could hamper future market development. - Methodology: - Interviews with CNSF, SHCP, CONSAR, IMSS and annuity company management. - Written material provided by CNSF. ### The 1992 and 1997 Pension Reforms; Contribution Rates - Reform milestones: - SAR-92 introduced mandatory defined contribution individual accounts; two accounts (retiro and vivienda). - SAR-97 (effective July 1997) covered private sector workers, abolished old-age DB for private sector under LSS-73, restructured IMSS benefits into RCV and IV lines. - Contribution rates (percentage of base salary / other specifics preserved): - Retiro: SAR-92 = 2.0 percent; SAR-97 = 2.0 percent (employer funds 100 percent). - RCV (Cesantía y Vejez): SAR-92 = 4.5 percent (employer funds 3.15 percent, employee funds 1.125 percent, Federal Government funds 0.225 percent). - Cuota Social: MXN 2.76 per working day (indexed to inflation; value indicative of end of 2005). - IV (Invalidez y Vida): SAR-92 = 2.5 percent; SAR-97 = 2.5 percent (employer 1.75 percent, employee 0.625 percent, Federal Government 0.125 percent). - RT (Riesgo de Trabajo): SAR-97 = Min. 0.25 percent, Max. 15 percent (variable per Article 72 of LSS-97). - Vivienda: SAR-92 = 5.0 percent; SAR-97 = 5.0 percent. ### Accumulated Savings in the AFORE System (1998–2005) - Aggregate accumulation (RCV, Voluntary, Housing accounts; values in MXN 000,000; percentage of GDP preserved where reported): - 1998: AFOREs = 13; RCV = 54,470.5; Voluntary = 46.7; Housing = 51,419.0; Total = 2.75 percent of GDP. - 1999: AFOREs = 14; RCV = 104,194.8; Voluntary = 331.3; Housing = 81,039.7; Total = 4.03 percent of GDP. - 2000: AFOREs = 13; RCV = 158,806.2; Voluntary = 846.0; Housing = 110,294.0; Total = 4.91 percent of GDP. - 2001: AFOREs = 13; RCV = 242,242.2; Voluntary = 1,513.8; Housing = 148,148.2; Total = 6.74 percent of GDP. - 2002: AFOREs = 11; RCV = 315,322.4; Voluntary = 1,900.5; Housing = 178,808.2; Total = 7.91 percent of GDP. - 2003: AFOREs = 12; RCV = 392,881.8; Voluntary = 2,221.3; Housing = 210,071.6; Total = 8.78 percent of GDP. - 2004: AFOREs = 13; RCV = 469,145.8; Voluntary = 2,031.0; Housing = 246,336.9; Total = 9.40 percent of GDP. - 2005: AFOREs = 18; RCV = 577,008.7; Voluntary = 1,981.4; Housing = 284,599.1. (Total/GDP percent not provided in extracted content.) - Observation: - To assess private annuity demand, add IMSS reserves for IV and RT to AFORE reserves. ### Decumulation Phase: Cohorts and Choices - Cohort treatment: - New cohort (entrants after July 1997): - If disabled or dies: must buy annuity from private specialized annuity provider. - If old-age retirement (65+) or severance at old age (60+): choice between programmed withdrawal (AFORE) or annuity (private). - If balance insufficient to obtain annuity ≥ PMG, worker required to select programmed withdrawal; when exhausted, Government pays PMG. - Transitional cohort (≥ one pre-reform contribution): - Can opt for old-age retirement benefits under LSS-73. - Originally required to buy annuity if disabled or dies (changed in 2002; see Section IV). - Distribution in 2004: - Transitional generation: 7,743,057 workers — 63 percent. - Actual (new) generation: 4,604,993 workers — 37 percent. - Total: 12,348,050 workers — 100 percent. - Implication: - Expected that all transitional cohort retirees will elect LSS-73 benefits because LSS-73 old-age benefits are on average more generous than LSS-97. - Short- to medium-term private annuity market growth depends heavily on accumulated IV and RT funds. ### Market Structure, Concentration and Size (1997–end-2005) - Providers and ownership (end-2005): - 1997: 13 insurance companies licensed. - End-2005: 11 specialized pension annuity companies active (7 subsidiaries of foreign companies; 4 part of financial groups). - 2005 market shares (Percentage of Market Premium; total = 100.0): - Pensiones BBVA Bancomer: 19.7 - Pensiones Inbursa: 17.0 - Profuturo GNP Pensiones: 15.6 - Pensiones Banamex: 10.5 - Pensiones Banorte Generali: 9.0 - HSBC Pensiones: 6.8 - Pensiones Comercial América: 6.6 - Metlife Pensiones México: 6.1 - Principal Pensiones: 4.4 - Allianz Rentas Vitalicias: 2.9 - Royal & SunAlliance Pensiones (México): 1.4 - Market concentration: - Foreign subsidiaries increased in importance (counts by year preserved). - CR5 (largest five providers' share, percent): 1997 = 85.3; 1998 = 77.0; 1999 = 73.7; 2000 = 71.4; 2001 = 72.2; 2002 = 72.0; 2003 = 72.5; 2004 = 72.9; 2005 = 71.8. - Finding: Concentration decreased from 85 percent to 72 percent between 1997 and 2001 and remained fairly stable thereafter. - Market size and penetration: - Insurance penetration (2005): GPI/GDP = 1.7 percent; GPI/Pop. = MXN 1,389 per capita. - Pension annuity share of gross premium income: 1997 = 3.2 percent → 2001 = 13.2 percent → 2005 = 3.3 percent. ### Pension Annuity Market Growth Metrics (1997–2005) - Eligible insured (IV and RT eligible): 1997 = 6,068; 1998 = 23,905; 1999 = 24,782; 2000 = 27,384; 2001 = 30,778; 2002 = 14,879; 2003 = 5,998; 2004 = 6,365; 2005 = 7,917. - Annuitants (bought private annuity): 1997 = 4,410; 1998 = 23,457; 1999 = 24,599; 2000 = 27,188; 2001 = 30,824; 2002 = 15,798; 2003 = 6,153; 2004 = 6,187; 2005 = 7,927. - GPI for pension annuities (MXN mln., Real 2005 pesos): 1997 = 2,522; 1998 = 12,913; 1999 = 13,389; 2000 = 14,529; 2001 = 16,641; 2002 = 10,261; 2003 = 3,497; 2004 = 5,225; 2005 = 4,504. - o/w IV (percent of pension annuity GPI): 1997 = 96; 1998 = 90; 1999 = 87; 2000 = 86; 2001 = 86; 2002 = 86; 2003 = 86; 2004 = 85; 2005 = 85. - GPI/GDP (percent) for pension annuities: 1997 = 0.08; 1998 = 0.33; 1999 = 0.29; 2000 = 0.26; 2001 = 0.28; 2002 = 0.16; 2003 = 0.05; 2004 = 0.07; 2005 = 0.05. - Finding: - Eligible annuitants rose to 30,778 in 2001 and fell to ~7,900 in 2005; private annuitants followed similar pattern. - 2004 premium spike due to congressional approval of increased IV and RT annuities requiring retroactive payments as of March 2004, transferring almost $1,700 million MXN single premium to annuity companies. ### Product Design, Pricing Parameters, and Regulation - Product features: - Immediate premium level annuities indexed to inflation. - Components: (a) annuity for retirement, unemployment at old age, disability, incapacity, worker’s compensation, orphans, widows/widowers; (b) survivor benefit. - Pricing and technical rules: - Separate premium calculation for annuity and survivor benefit; survivor uses death conditional probability of pensioner. - Mortality tables: - IV pensioners: EMSSIM-97 (male), EMSSIH-97 (female) by IMSS. - Non-disabled pensioners: EMSSAM-97 (male), EMSSAH-97 (female) by CNSF using CONAPO. - Gender-specific dynamic tables approved under Art. 81 of LSAR; methodology revised by CONAC. - Technical rate required: 3.5 percent real (unique). - Indexation: annuities indexed every February to prior fiscal year's inflation; adjustments account for timing and monthly crediting. - Two mark-ups on premiums: - 2 percent for contingency reserve (mortality deviations). - 1 percent for acquisition and administration costs. - Single pricing system (Sistema Único de Cotización (SUC)) mandatory — ensures uniform basic IV and RT annuity pricing across providers. - Rationale: eliminate unfair commercial practices and protect IMSS reserves. - Counterpoint: general life insurers could better hedge longevity risk by natural offset with mortality risk in pure life portfolios. ### Additional Benefits, Reserving and Technical Provisions - Companies may offer additional benefits but cannot alter basic pension price; costs absorbed via spread between technical rate and expected asset returns. - Additional benefits: - Additional economic benefits: reserves must use 1 percent real technical rate. - Life insurance products for pensioner/beneficiaries: purchased from life insurers as single premium; life insurers constitute corresponding reserves. - Additional benefits must be registered with CNSF and posted. - Technical provisions (1997–2005, MXN mln., real 2005 pesos): - NPR: 1997 = 2,332.4; 1998 = 14,774.8; 1999 = 27,417.3; 2000 = 40,728.7; 2001 = 55,791.0; 2002 = 63,875.1; 2003 = 65,429.0; 2004 = 68,592.6; 2005 = 70,624.7. - Contingency reserve: 1997 = 46.7; 1998 = 295.0; 1999 = 548.3; 2000 = 814.5; 2001 = 1,115.0; 2002 = 1,277.0; 2003 = 1,308.0; 2004 = 1,371.0; 2005 = 1,412.0. - Special mathematical reserve: 1997 = 0.3; 1998 = 9.6; 1999 = 32.7; 2000 = 82.1; 2001 = 139.5; 2002 = 226.2; 2003 = 315.5; 2004 = 445.0; 2005 = 574.4. - Investment fluctuation reserve: 1997 = 1.8; 1998 = 33.7; 1999 = 156.2; 2000 = 234.1; 2001 = 333.6; 2002 = 628.0; 2003 = 744.9; 2004 = 747.1; 2005 = 817.8. - Loss reserves: 1997 = 7.4; 1998 = 9.7; 1999 = 15.2; 2000 = 33.8; 2001 = 70.6; 2002 = 97.1; 2003 = 133.2; 2004 = 167.5; 2005 = 203.6. - NPR/Total (percent): 1997 = 96; 1998 = 97; 1999 = 97; 2000 = 97; 2001 = 97; 2002 = 97; 2003 = 96; 2004 = 96; 2005 = 96. - Aggregate provisions: - Total technical provisions increased from MXN 2,440 million in 1997 to MXN 57,493 million in 2001; total reached MXN 73,640.41 in December 2005 (as reported in figure caption). ### Investment Regulation, Asset Composition and ALM - Admitted assets and investment limits (percent limits preserved): - Papers issued/backed by Federal Government: 100 - Papers issued/backed by credit institutions or international financial bodies: 60 - Papers issued by SINCAS: 1 - Papers issued/backed by other institutions: 30 - Repurchasing: 30 - Foreign debt papers: 2 - Investment limits by issuer (percent): - Federal Government: 100.0 - Banks or international financial bodies: 18.0 - SINCAS: 0.5 - Institutions different from (1),(2),(3): 7.0 - Groups and institutions in same economic sector (shares): 10.0 - Financial institution (shares): 5.0 - Securities registered in SIC of Mexican Stock Exchange: 1.0 - Investment composition evolution (1997–2005; percent and MXN mln. preserved): - Government Sec. (percent): 1997 = 78.5; 1998 = 82.2; 1999 = 84.9; 2000 = 80.7; 2001 = 82.8; 2002 = 80.6; 2003 = 73.7; 2004 = 68.5; 2005 = 76.3. - Private Sec. (percent): 1997 = 20.0; 1998 = 16.3; 1999 = 10.4; 2000 = 14.4; 2001 = 10.4; 2002 = 11.6; 2003 = 15.7; 2004 = 16.2; 2005 = 11.9. - Total (MXN mln.): 1997 = 3,602; 1998 = 16,191; 1999 = 29,713; 2000 = 43,743; 2001 = 60,039; 2002 = 69,061; 2003 = 78,112; 2004 = 84,112; 2005 = 79,852. - Currency and inflation risk (percent): - Local real: 1997 = 60.0; 1998 = 58.0; 1999 = 84.4; 2000 = 93.1; 2001 = 92.2; 2002 = 92.6; 2003 = 97.4; 2004 = 94.2; 2005 = 91.3. - Local nominal and foreign nominal values preserved in Table 12. - Matching and maturity: - Average maturity of total bond portfolio: 1997 = 5.80; 1998 = 7.30; 1999 = 6.50; 2000 = 11.80; 2001 = 13.70; 2002 = 15.00; 2003 = 13.40; 2004 = 13.60; 2005 = 14.20. - Average maturity of government bond portfolio (selected): 1997 = 6.14; 1998 = 7.51; 1999 = 6.73; 2000 = 13.07; 2001 = 14.71; 2002 = 16.36; 2003 = 14.76; 2004 = 14.88; 2005 = 14.53. - Average maturity of liabilities (expected longevity) = 37 years — creating maturity mismatch with assets. - Investment behavior: - Most technical provisions invested in government bonds (UDIBONOS, PICs, CBICs). - Shift over time from nominal and foreign assets to real and domestic assets, reducing inflation and currency risk. ### Interest Rate Environment, Pricing Rigidity and Risks - Long-term real interest rate movements (selected): - 10 yr Udibonos: 7.14 percent in October 1999 → 4.30 percent in December 2005 → 5.09 percent in June 2006. - 20 and 30 yr CBICs: 6.16 and 6.21 percent in January 2003 → 4.59 and 4.68 percent in December 2005. - Regulatory technical rate: - Technical rate for pricing fixed at 3.5 percent real; has not been modified. - Risks highlighted: - Fixed regulatory technical parameters can create unbalanced basic annuity product; technical profit margins erode if technical rate not adjusted to market yields. - Mortality tables are experimental; need annuitant-specific tables as experience accumulates. - Fixed pricing may hide solvency differences — price does not reflect offering company solvency. - In low long-term real rate environment without timely regulatory adjustments, companies and annuitants may face undesirable risk exposure. ### Accounting Classification, Liquidity and Custody - Investment classification: - For sale: finance operations/short-term liabilities; marked to market. - To hold until maturity: back longer-term liabilities; valued at constant discount rate; valued at maturity. - On average 90 percent of pension annuity portfolios classified as held to maturity. - Liquidity rules: - Short-term technical reserves must be covered by equivalent assets; historical shares of reserves to be invested in liquid instruments varied (e.g., 100 percent for loss reserve; 6 percent for NPR, contingency, special mathematical reserves). - Starting 2006, liquidity requirements calculated from asset–matching analysis for pensions. - Custody: - Domestic securities administered by credit institutions and kept in custody by depository institutions. - For foreign-currency investments outside country, Mexican Financial Entities or subsidiaries act as intermediaries; foreign custodians authorized in applicable jurisdiction. ### Duration, Convexity, Key-Rate Duration and Supervisory Considerations - Duration used to measure sensitivity of option-free bond portfolios; liabilities are not interest-rate sensitive, so duration is meaningful. - Limitations: - Duration accurate only for small, parallel yield curve shifts. - Extended metrics recommended: - Convexity to capture second-order effects for large rate changes. - Key-rate/partial durations to address non-parallel yield curve movements. - Supervisory implications: - Require a diversity of risk metrics beyond duration to accommodate tactical business decisions while ensuring oversight. ### Underwriting, Investment Performance and Profitability - Underwriting trends (selected): - Average loss ratio (LR) in 1997 exceeded 140 percent. - Underwriting performance deteriorated until 2002 when combined ratio (CR) exceeded 100. - Deterioration driven by drastic fall in gross premium income since 2002 and delayed cost adjustments. - Evolution of average underwriting costs (GPI and components, Real 2005 pesos): - GPI: 2001 = 16,640.59; 2002 = 10,261.26; 2003 = 3,501.71; 2004 = 5,224.7; 2005 = 4,504.33. - Total Underwriting: 2001 = 1,152.48 (Percent 9.5); 2002 = 984.68 (Percent 9.6); 2003 = 506.18 (Percent 14.5); 2004 = 318.64 (Percent 6.1); 2005 = 380.93 (Percent 8.5). - Investment returns: - Record nominal investment return of 20 percent in 1998; average nominal rates over 2002–2005 decreased to 6.1 percent. - Link to solvency: - Combined ratios often > 100 percent; profitability depends on investment performance. - Deterioration in investment returns contributed to weakened company profitability. ### Solvency Framework, Minimum Guarantee Capital (CMG) and Weighting System - CMG purpose: cover claim deviations, reinsurer failures, adverse asset valuation and asset–liability mismatches. - CNSF weighting system (weights for year of projection k; March 2006): - k <= 5 : 1 - 5 < k <= 6 : 0.96 - 6 < k <= 7 : 0.94 - 7 < k <= 8 : 0.85 - 8 < k <= 9 : 0.78 - 9 < k <= 12 : 0.51 - 12 < k <= 14 : 0.46 - 14 < k <= 15 : 0.42 - 15 < k <= 17 : 0.37 - 17 < k <= 18 : 0.30 - 18 < k <= 19 : 0.25 - 19 < k <= 20 : 0.24 - 20 < k <= 21 : 0.23 - 21 < k <= 23 : 0.20 - 23 < k <= 24 : 0.19 - 24 < k <= 25 : 0.17 - 25 < k <= 27 : 0.13 - 27 < k <= 28 : 0.08 - 28 < k <= 29 : 0.02 - 29 < k <= 30 : 0.01 - k > 30 : 0.00 - Assessment: - Maturity-gap weighting approach is improvement over ad hoc capital rules but may underestimate long-term interest rate risk by assigning near-zero weights to long maturities. - Recommendation: complement with economic duration-based ALM and consider unit weights at all maturities for prudential scenarios. - Solvency and profitability indicators (1997–2005): - Solvency Margin (MXN million): 1997 = 667.2; 1998 = 292.5; 1999 = 811.3; 2000 = 862.0; 2001 = 570.8; 2002 = 829.1; 2003 = 2,176.8; 2004 = 1,761.2; 2005 = 2,927.5. - ROE (percent): 1997 = -8; 1998 = -29.2; 1999 = -20.2; 2000 = -19.8; 2001 = -6.4; 2002 = 15.6; 2003 = 22.5; 2004 = 26.1; 2005 = 24.3. - ROA (percent): 1997 = -2.9; 1998 = -3; 1999 = -1.2; 2000 = -0.9; 2001 = -0.3; 2002 = 0.7; 2003 = 2.9; 2004 = 4.0; 2005 = 1.9. - Puzzling divergence: - Since 2002: premiums and underwriting worsened, yields fell, yet regulatory solvency margins improved — possible explanations include growth of core portfolio, legacy high-yield investments, or regulatory weighting understating economic risk. - Policy implication: adopt economic-capital-based solvency metrics for prudential monitoring. ### Market Drivers, Legal Reforms and Behavioral Observations - Key legal changes affecting IV benefits and market growth: - 2001: indexation base of LSS-73 changed from minimum salary to INPC, increasing relative generosity of LSS-73. - 2002: reform allowed transitional retirees to withdraw 2 percent RCV subaccount balance as lump sum; IMSS extended LSS-73 option to IV benefits. - Market outcome: - IMSS policy changes and election options caused sudden drop in gross premium income to private annuity providers after 2002. - Behavioral puzzle: - Table comparisons suggest most individuals should prefer LSS-97 on average, yet many elected LSS-73; causes include informational gaps, administrative ambiguity in LSS-97, IMSS not proactively comparing benefits, tightened eligibility and increased temporary pensions. - Consequences: - (a) Fewer potential private annuitants. - (b) Deterioration of the annuitant risk pool (delayed entry, longer-lived annuitants). ### Investment Market Frictions and Less-Urgent Issues - Market frictions: - Bond stripping possible in principle but counterparties scarce. - Short-maturity inflation-indexed instruments (below one year) reported very illiquid. - Suggested government/authority actions: - Consider issuing "zero coupon CPI-indexed bonds". - Consider issuing "longevity bonds". - Consider allowing annuity companies, under regulatory framework, to buy interest rate swaps in OTC market to manage duration mismatches. - CNSF could assess short-term inflation risk and consider relaxing inflation matching for short-term liabilities due to illiquidity. ### Policy Recommendations and Priority Actions (preserved wording and implications) - Enhance transparency in choosing benefits under LSS-73 and LSS-97 via a comprehensive election system for disabled workers. - Clarify government liability so IMSS is indifferent to worker choice between old and new systems. - Require IMSS to compare and disclose disability benefits under the two systems unless the worker requests otherwise. - Consider deregulating annuity prices for private sector providers so competition focuses on basic benefits; pair deregulation with transparency and market conduct measures. - Improve ALM practices and link them to capital requirements and solvency via independent assessment and technical assistance; consider partial/key durations and convexities. - Explore instruments to diversify longevity risk: longevity-linked instruments, longevity participating annuities, and possible government or international issuance of longevity-linked securities. - Consider allowing general life insurers to operate in annuity market to enable natural hedging across life and annuity portfolios, while recognizing remaining net longevity exposures. *Italic: Source: _cr07163 - References (extracted content from the IMF chapter on the Mexican pension annuity market, 1997–2005).* ### References.............................................................................................................. ### _cr07163 - References ### Introduction - Paper analyzes the performance and development of the Mexican pension annuity market that stemmed from the 1997 pension reform (SAR-97). - Highlights: - Mexican experience offers lessons for countries designing the decumulation phase of second-pillar systems. - Identifies technical and policy concerns that could hamper future market development. - Methodology: - Benefited from interviews with CNSF, SHCP, CONSAR, IMSS and annuity company management. - Written material provided by CNSF listed in the reference section. ### The 1992 and 1997 Pension Reforms - SAR-92 (1992): - Introduced mandatory defined contribution individual accounts complementary to existing defined benefit schemes for private and public sector workers. - Established two individual accounts for old age (retiro) and housing (vivienda). - SAR-97 (effective July 1997): - Covered only private sector workers. - Abolished old age defined benefit scheme for private sector workers under LSS-73. - Reformed disability, old age, and life benefits managed by IMSS into two lines: RCV (retiro, cesantía en edad avanzada y vejez) and IV (invalidez y vida). - Contributions under SAR-92 and SAR-97 (percentage of base salary): - Retiro: SAR-92 = 2.0 percent; SAR-97 = 2.0 percent. - RCV (Cesantía y Vejez): SAR-92 = 4.5 percent. - Cuota Social: MXN 2.76 per working day (indexed to inflation and updated every quarter; value reported is indicative of end of 2005). - IV (Invalidez y Vida): SAR-92 = 2.5 percent; SAR-97 = 2.5 percent. Note on funding: Employer funds 1.75 percent, employee funds 0.625 percent, Federal Government funds 0.125 percent. - RT (Riesgo de Trabajo): SAR-97 = Min. 0.25 percent, Max. 15 percent (variable as defined in Article 72 of LSS-97). - Vivienda: SAR-92 = 5.0 percent; SAR-97 = 5.0 percent. - Notes on funding composition where specified: - Retiro employer funds 100 percent of this contribution. - RCV employer funds 3.15 percent, employee funds 1.125 percent, Federal Government funds 0.225 percent. - IV employer funds 1.75 percent, employee funds 0.625 percent, Federal Government funds 0.125 percent. ### Accumulated Savings in the AFORE System (1998–2005) - As of end-2005, around 10 percent of GDP is accumulated in the AFORE system (note: table reports up to 2004 explicitly and partial 2005 numbers). - Accumulated balances by year (RCV account, Voluntary account, Housing account, Total, Percentage of GDP): - 1998: Number of AFOREs = 13; RCV account = 54,470.5 (MXN 000,000); Voluntary account = 46.7 (MXN 000,000); Housing account = 51,419.0 (MXN 000,000); Total = 2.75 percent of GDP. - 1999: Number of AFOREs = 14; RCV account = 104,194.8; Voluntary account = 331.3; Housing account = 81,039.7; Total = 4.03 percent of GDP. - 2000: Number of AFOREs = 13; RCV account = 158,806.2; Voluntary account = 846.0; Housing account = 110,294.0; Total = 4.91 percent of GDP. - 2001: Number of AFOREs = 13; RCV account = 242,242.2; Voluntary account = 1,513.8; Housing account = 148,148.2; Total = 6.74 percent of GDP. - 2002: Number of AFOREs = 11; RCV account = 315,322.4; Voluntary account = 1,900.5; Housing account = 178,808.2; Total = 7.91 percent of GDP. - 2003: Number of AFOREs = 12; RCV account = 392,881.8; Voluntary account = 2,221.3; Housing account = 210,071.6; Total = 8.78 percent of GDP. - 2004: Number of AFOREs = 13; RCV account = 469,145.8; Voluntary account = 2,031.0; Housing account = 246,336.9; Total = 9.40 percent of GDP. - 2005: Number of AFOREs = 18; RCV account = 577,008.7; Voluntary account = 1,981.4; Housing account = 284,599.1. (Note: table shows these 2005 balances without an explicit Total/GDP percentage in the extracted content.) - Observations: - To assess potential private annuity demand, add reserves accumulated by IMSS for IV and RT to AFORE reserves. ### Decumulation Phase, Transitional and New Cohorts - Treatment by cohort: - New cohort (entrants after July 1997): - If disabled or dies: must buy an annuity from a private specialized annuity provider. - If retires at old age (65 years or older) or severance at old age (at least 60 years old): choice between programmed withdrawal (AFORE) or annuity (private provider). - If accumulated balance at retirement is less than needed to obtain an annuity ≥ PMG (Pensión Mínima Garantizada), the worker is required to select a programmed withdrawal; when resources are exhausted, Government pays the PMG. - Transitional cohort (at least one contribution pre-reform): - Can opt for old age retirement benefits under LSS-73. - If disabled or dies, originally required to buy an annuity from a private specialized annuity provider (changed in 2002; details referenced to Section IV). - Distribution of workers in 2004: - Transitional generation: 7,743,057 workers — 63 percent. - Actual (new) generation: 4,604,993 workers — 37 percent. - Total: 12,348,050 workers — 100 percent. - Implications: - Because LSS-73 old age benefits are on average more generous than LSS-97, it is expected that all retirees in the transitional cohort will elect LSS-73 benefits. - Short- to medium-term development of private annuity market depends heavily on accumulated funds in IV and RT lines. ### Market for Pension Annuities: Structure and Growth - Market evolution (1997–end-2005): - 1997: 13 insurance companies licensed to provide pension annuity benefits. - End-2005: 11 specialized pension annuity companies active. - Ownership composition at end-2005: - Seven companies were subsidiaries of foreign companies (F). - Four companies were part of financial groups (GF). - Pension annuity providers and 2005 market shares (Percentage of Market Premium): - Pensiones BBVA Bancomer, S.A. de C.V., Grupo Financiero BBVA Bancomer: 19.7 - Pensiones Inbursa, S.A., Grupo Financiero Inbursa: 17.0 - Profuturo GNP Pensiones, S.A. de C.V.: 15.6 - Pensiones Banamex, S.A. de C.V., Grupo Financiero Banamex: 10.5 - Pensiones Banorte Generali, S.A. de C.V., Grupo Financiero Banorte: 9.0 - HSBC Pensiones, SA.: 6.8 - Pensiones Comercial América, S.A. de C.V.: 6.6 - Metlife Pensiones México, S.A.: 6.1 - Principal Pensiones, S.A. de C.V.: 4.4 - Allianz Rentas Vitalicias, S.A.: 2.9 - Royal & SunAlliance Pensiones (México), S.A. de C.V.: 1.4 - Total: 100.0 - Market structure change and regulatory rationale: - 1997 LGISMS allowed life insurance companies to offer pension annuities for a five-year transition; after five years annuity operations had to be separated into specialized pension annuity companies. - Initially 50 percent of licensed companies were composite; after 2002 only specialized companies remained. - Rationale for specialization: - Avoid contagion from composite life and non-life insurers operating annuities. - Avoid implicit government guarantee associated with mandated benefit provision. - Counterpoint noted in the paper: - General life insurance companies would have been better placed to hedge longevity risk in annuitant portfolios with mortality risk in pure life portfolios. *Italic: Source: _cr07163 - References (extracted content from the IMF chapter on the Mexican pension annuity market, 1997–2005).* ### 15. Another way in which market changed over the year is reflected by the increased ### 15. Another way in which market changed over the year is reflected by the increased ### Market structure and concentration - Evolution of market structure (1997–2005) shows increased importance of foreign subsidiaries and decreased importance of local financial groups: - Foreign Subsidiaries: 1997: 4, 1998: 4, 1999: 5, 2000: 5, 2001: 5, 2002: 7, 2003: 6, 2004: 7, 2005: 7 - Local Financial Groups: 1997: 6, 1998: 5, 1999: 5, 2000: 5, 2001: 5, 2002: 3, 2003: 2, 2004: 2, 2005: 2 - Other: 1997: 3, 1998: 4, 1999: 4, 2000: 4, 2001: 4, 2002: 3, 2003: 3, 2004: 2, 2005: 2 - CR5 (percent) (participation of largest five providers in total gross premium written): 1997: 85.3, 1998: 77.0, 1999: 73.7, 2000: 71.4, 2001: 72.2, 2002: 72.0, 2003: 72.5, 2004: 72.9, 2005: 71.8 - Finding: Market concentration has been decreasing but remains fairly high. The CR5 decreased from 85 percent to 72 percent between 1997 and 2001 and remained stable since that year. ### Pension annuity market size, penetration, and business-line evolution - Pension annuity market described as "very small" due to the 1997 reform timing and low overall insurance penetration/density in Mexico: - Insurance penetration and density (2005): GPI/GDP = 1.7 percent of GDP; GPI/Pop. = MXN 1,389 per capita - Insurance sector business lines (Percent shares, 1997–2005): - Motor: 1997: 25.8, 1998: 25.5, 1999: 23.8, 2000: 25, 2001: 24.7, 2002: 24.2, 2003: 26.9, 2004: 24.4, 2005: 25.1 - Other P&C: 1997: 25.9, 1998: 20.4, 1999: 18, 2000: 16.8, 2001: 17.4, 2002: 18.5, 2003: 21.4, 2004: 20.3, 2005: 18.9 - Life: 1997: 35.2, 1998: 31.7, 1999: 36.1, 2000: 34.2, 2001: 33.7, 2002: 39.1, 2003: 36, 2004: 38.9, 2005: 37.9 - Accidents & Health: 1997: 9.8, 1998: 8.9, 1999: 9.4, 2000: 10.6, 2001: 10.9, 2002: 10.9, 2003: 12.9, 2004: 12.7, 2005: 14.8 - Pensions: 1997: 3.2, 1998: 13.5, 1999: 12.7, 2000: 13.2, 2001: 13.2, 2002: 7.3, 2003: 2.7, 2004: 3.7, 2005: 3.3 - Total GPI (MXN mln) 1/ (Real 2005 pesos): 1997: 78,639; 1998: 94,811; 1999: 101,001; 2000: 110,320; 2001: 125,699; 2002: 140,928; 2003: 128,063; 2004: 142,640; 2005: 138,377 - GPI/GDP (percent): 1997: 1.3, 1998: 1.5, 1999: 1.6, 2000: 1.6, 2001: 1.8, 2002: 2.0, 2003: 1.8, 2004: 1.9, 2005: 1.7 - GPI/Pop. (MXN): 1997: 839, 1998: 999, 1999: 1,051, 2000: 1,133, 2001: 1,273, 2002: 1,408, 2003: 1,263, 2004: 1,389, 2005: 1,331 - Pension annuity share of gross premium income: - Increased from 3.2 percent in 1997 to 13.2 percent in 2001, then decreased to 3.3 percent in December 2005. ### Pension annuity market growth (key variables, 1997–2005) - Eligible Insured (Annuitants eligible for IV and RT benefits): 1997: 6,068; 1998: 23,905; 1999: 24,782; 2000: 27,384; 2001: 30,778; 2002: 14,879; 2003: 5,998; 2004: 6,365; 2005: 7,917 - Annuitants (bought a pension annuity from private sector providers): 1997: 4,410; 1998: 23,457; 1999: 24,599; 2000: 27,188; 2001: 30,824; 2002: 15,798; 2003: 6,153; 2004: 6,187; 2005: 7,927 - GPI (MXN mln.) 1/ (Real 2005 pesos) for pension annuities: 1997: 2,522; 1998: 12,913; 1999: 13,389; 2000: 14,529; 2001: 16,641; 2002: 10,261; 2003: 3,497; 2004: 5,225; 2005: 4,504 - o/w IV (percent of pension annuity GPI): 1997: 96, 1998: 90, 1999: 87, 2000: 86, 2001: 86, 2002: 86, 2003: 86, 2004: 85, 2005: 85 - GPI/GDP (percent) for pension annuities: 1997: 0.08, 1998: 0.33, 1999: 0.29, 2000: 0.26, 2001: 0.28, 2002: 0.16, 2003: 0.05, 2004: 0.07, 2005: 0.05 - GPI/Pop. (MXN) for pension annuities: 1997: 26.9, 1998: 139.0, 1999: 139.3, 2000: 149.0, 2001: 168.3, 2002: 102.4, 2003: 34.4, 2004: 50.8, 2005: 43.6 - Finding: Number of eligible annuitants rose from 6,068 in 1997 to 30,778 in 2001; number of private-sector annuitants rose from 4,410 in 1997 to 30,824 in 2001; both then fell sharply to around 7,900 in 2005. - Notable spike: 2004 pensions premiums spike due to congressional approval of increased IV and RT pension annuities requiring retroactive payments as of March 2004 (resulted in transfer of almost $1,700 million MXN of single premium to annuity companies). ### Product design, pricing parameters, and regulation - Pension annuities from the 1997 reform are immediate premium level annuities indexed to inflation. - Product components: - (a) An annuity for retirement, unemployment at old age, disability, incapacity, worker’s compensation, orphans, widows and widowers; - (b) A survivor benefit for eventual beneficiaries existing when the right to a pension annuity is acquired. - Premium calculation: annuity and survivor benefit premiums calculated separately; survivor benefit uses death conditional probability of the pensioner. - Mortality tables and technical rate: - Two sets of gender-specific mortality tables for invalid (IV) pensioners: EMSSIM-97 (male), EMSSIH-97 (female) developed by IMSS. - Two sets for non-disabled pensioners: EMSSAM-97 (male), EMSSAH-97 (female) prepared by CNSF using CONAPO data. - A unique technical rate of 3.5 percent real is required. - Mortality tables are gender specific dynamic tables approved by the special committee under Art. 81 of the LSAR; methodology revised by Colegio Nacional de Actuarios de México (CONAC). - Indexation and premium adjustments: - Annuities are indexed every February to the inflation of the previous fiscal year; factors were developed to account for timing differences and monthly crediting of inflation. - Two mark-ups added to premiums: - 2 percent mark-up accumulated in the contingency reserve as an extra buffer for statistical deviations in mortality tables. - 1 percent mark-up to finance acquisition and administration costs. - Single pricing system requirement: - All companies must use a single pricing system (Sistema Único de Cotización (SUC)) capturing individual data and using prescribed regulatory pricing parameters, ensuring uniform basic IV and RT annuity pricing across providers. - Rationale: LSS-97 establishes defined benefits and grants pensioners choice of annuity company; SUC aims to eliminate unfair commercial practices and protect IMSS reserves. ### Additional benefits and reserving rules - Companies may offer additional benefits beyond the basic IV and RT annuities to compete, but: - Additional benefits cannot affect the price charged for the basic pension benefit; companies absorb costs for additional benefits, financed via spread between technical rate and expected asset returns. - Additional benefits must be registered in the CNSF before being offered; registered additional benefits are posted on CNSF website. - Two categories of additional benefits: - Additional economic benefits (increments on the basic pension) — reserves must be constituted considering a technical interest rate of 1 percent real. - Life insurance products for pensioner or beneficiaries — purchased from life insurers as single premium; amount registered as acquisition cost for the annuity company; life insurers constitute corresponding reserves. - The same mortality table used for basic benefits is used for additional economic benefits. ### Risks, pricing rigidity, and market implications - Regulatory pricing framework risks: - Fixed regulatory technical parameters may create an unbalanced basic annuity product, posing risks for providers and annuitants. - Risks for providers: - Technical profit margins can be quickly eroded if technical rates are not adjusted in response to market interest rates. - Examples of long-term real interest rate movements: - 10 yr Udibonos real rate: around 7.14 percent in October 1999 → 4.30 percent in December 2005 → 5.09 percent in June 2006. - Note: auctions average result reached the lowest real annual interest rate level (3.89 percent) in October 2003. Source: Banxico. - 20 and 30 yr PICs: 6.14 and 6.12 percent in October 2001 → 5.96 and 5.92 percent in November 2002. Source: Banxico and Banobras. - 20 and 30 yr CBICs: 6.16 and 6.21 percent in January 2003 → 4.59 and 4.68 percent in December 2005. Source: Banxico and Banobras. - The technical rate of 3.5 percent real has not been modified (see Section E). - Pension annuity companies hold large positions in UDIBONOS, PICs and CBICs with maturities between seven and 28 years. - Risks related to mortality tables: - Current mortality tables are experimental and based on population averages; annuity companies need to develop annuitant-specific mortality tables when sufficient experience is accumulated. - Pension policy changes introduced in 2002 altered demographic assumptions underlying pricing framework, causing IV annuitant pool to degenerate; CNSF prepared studies since 2000 but post-2002 trends require at least five years of market data to propose new tables. - Solvency and price transparency risk: - Fixed pricing rules may hide solvency differences across providers — the price of an annuity does not reflect the solvency situation of the offering company. - In a low long-term real interest rate environment without timely adjustment of the regulatory technical rate, companies’ inability to establish sound premium rates may expose companies and annuitants to undesirable levels of risk. ### Implications for annuitants and competitive considerations - For annuitants: - Fixed technical parameters do not guarantee good pricing in changing circumstances; while monthly benefit amounts are defined by law and unaffected, the cost of the annuity to the retiring worker is influenced by pricing rules. - Competition and product differentiation: - Companies compete by offering additional benefits, but these do not alter the regulated price for the basic annuity. - The Chilean experience (Box 1) indicates that deregulated pricing can still deliver good value to annuitants in a competitive, efficient market; noted observations from Chile include: - Annuities with longer expected durations have lower money’s worth ratios than shorter durations. - Larger premiums buy annuities of better value on average than smaller ones. - Nearly half of variation in individual money’s worth ratios is unexplained by individual characteristics, implying market search inefficiencies especially among lower-income retirees. - Wide spreads between highest and lowest annuities led to an electronic quotation system to improve transparency and outcomes for retirees; caps on broker commissions were also introduced controversially. - Chilean average money’s worth ratios for indexed annuities around 1.04 – 1.08 in recent years, higher than actuarially fair value of 1.0 and higher than international norms for indexed annuities (0.8 to 0.85 in some countries). - Note: The Chilean evidence does not imply Mexico must deregulate prices; rather, it offers lessons on competition, transparency, and market mechanisms that affect annuitant value. *Source: CNSF, IFS, WDI (content from the provided IMF document excerpt).* ### 33. Mexican pension annuity providers need to form a variety of technical provisions ### Mexican pension annuity providers need to form a variety of technical provisions ### Types of technical provisions and their calibration - Pension annuities are immediate premium annuities; the largest provision is the net mathematical (or premium) reserve (NPR) (reserva de riesgos en curso), formed with the immediate premium of IV and RT benefits and backing annuity provision using a technical rate of 3.5 percent in real terms. - Premium reserve for additional economic benefits must use a technical rate of 1 percent in real terms. - The NPR is the long plan equivalent of the unearned premium reserve (UPR) for P&C insurers. - The premium reserve for additional benefits is gradually “earned” when benefits are paid and represented between 95 percent and 97 percent of total reserves in the 1997 – 2005 period. - Contingency reserve (reserva de contingencia o de previsión): formed with 2 percent of the mathematical reserves for pensions and additional benefits; used to cover statistical deviations in expected loss ratio. - Special mathematical reserve for basic benefits (reserva matemática especial): used to cover possible systemic improvements in longevity of the annuitant pool for non invalid pensioners. - Reserve for fluctuation in investments (reserva para las fluctuaciones en inversiones): constituted by a portion of the excess investment rate of return of the market portfolio over the technical interest rate (3.5 percent real). - Loss reserve (reserva para obligaciones pendientes de cumplir): used for providers’ liabilities with their pensioners. - Annuity premiums include a 3 percent surcharge composed by 2 percent for mortality deviations and 1 percent for administrative and acquisition cost. ### Historical levels and composition of technical provisions (1997–2005) - Table 8 (Evolution of Separate Technical Provisions, 1997–2005, MXN mln., real 2005 pesos) — reported values: - NPR: 1997 = 2,332.4; 1998 = 14,774.8; 1999 = 27,417.3; 2000 = 40,728.7; 2001 = 55,791.0; 2002 = 63,875.1; 2003 = 65,429.0; 2004 = 68,592.6; 2005 = 70,624.7. - Pending NPR (Pending): 1997 = 51.5; 1998 = 37.2; 1999 = 55.3; 2000 = 60.8; 2001 = 43.2; 2002 = 10.6; 2003 = 7.20; 2004 = 10.8; 2005 = 7.5. - Contingency: 1997 = 46.7; 1998 = 295.0; 1999 = 548.3; 2000 = 814.5; 2001 = 1,115.0; 2002 = 1,277.0; 2003 = 1,308.0; 2004 = 1,371.0; 2005 = 1,412.0. - Special Mat.: 1997 = 0.3; 1998 = 9.6; 1999 = 32.7; 2000 = 82.1; 2001 = 139.5; 2002 = 226.2; 2003 = 315.5; 2004 = 445.0; 2005 = 574.4. - Inv. Fluct.: 1997 = 1.8; 1998 = 33.7; 1999 = 156.2; 2000 = 234.1; 2001 = 333.6; 2002 = 628.0; 2003 = 744.9; 2004 = 747.1; 2005 = 817.8. - Loss Reserves.: 1997 = 7.4; 1998 = 9.7; 1999 = 15.2; 2000 = 33.8; 2001 = 70.6; 2002 = 97.1; 2003 = 133.2; 2004 = 167.5; 2005 = 203.6. - NPR/Total (percent): 1997 = 96; 1998 = 97; 1999 = 97; 2000 = 97; 2001 = 97; 2002 = 97; 2003 = 96; 2004 = 96; 2005 = 96. - Aggregate technical provisions: - Between 1997 and 2001 technical provisions increased from MXN 2,440 million to MXN 57,493 million. - Total reached MXN 73,640.41 in December 2005 (as reported in the figure caption). - Figure 1 (Evolution of Total Technical Provisions, Real 2005 MXN mln.) shows rapid growth until 2001 and much lower growth afterwards. ### Regulatory and actuarial valuation changes - A major change introduced in 2002 amended Article 47 of the Insurance Law (LGISMS): valuation of technical provisions must be conducted on the basis of actuarial standards requiring an actuarial valuation of future risks. - For annuities, valuation of technical provisions was established in 1997 to follow actuarial methodology issued by the Ministry of Finance (SHCP (2000)). - For inflation-indexed annuities, the methodology for the net premium reserve includes additional provisions to avoid revaluation due to changes in inflation. - Since 2004, valuation of the technical provision must be reviewed and signed by an actuary certified by the National College of Actuaries (CONAC) or approved by the CNSF. - CONAC developed actuarial practice standards; CNSF published relevant standards in Circulares S-10.1.8 and F-6.6.4. - The new rules (CNSF 2003, 2003a, 2003b, 2004, 2005) required companies to file an actuarial methodology for valuation of adequacy of reserves and included minimum requirement levels for technical provisions; methodologies must estimate technical provisions as the expected value of future claims derived from all policies in force from the valuation date until policy expiration. ### Formula and operational mechanics for NPR accumulation - The NPR reserve at the end of every month is given by the formula (text as provided): - mV is the NPR reserve at the end of month m; V m1− is the NPR reserve at the end of the previous month; mGPI is the gross premium income of the policies issued during the month m; C are the claims paid during the month m; mINPCΔ is the change in the price index during the month m; and i is the technical rate of 3.5 percent real that companies need to use in pricing their product. - The difference RMA mmm V V = − − )( 1 is the monthly minimum accumulation rate (rendimiento mínimo acreditable) used to form the UPR reserve. - The formula uses ½ for premiums and claims, implying uniform distribution within the month. ### Evolution of assets, financing structure and notable movements - Total assets closely follow liabilities; average total assets in the last nine years have been 90 percent financed by liabilities and only 10 percent by paid up capital. - Assets grew rapidly until 2001 and less rapidly afterward; in the first five years since the reform, total assets reached MXN 60,000 million while in the last four years they increased about only MXN 21,000 million. - Figure 2 (Evolution and Financing of Total Assets, Real 2005 MXN mln.) shows liabilities and paid up capital series; a drastic increase in capital in 2003 was due to one company transferring excess resources to capital participation in subsidiaries; in 2004 that company registered another increment in capital; in 2005 the same company disinvested around MXN 7 billion. Capital excess resources were stocks not listed on the Mexican stock exchange market and had no effect on solvency of the annuity market. ### Investment regulation for technical provisions: admitted assets and constraints - Regulatory framework: RIRT (CNSF 2000) and Circulares S-11.1 to S-11.6, expanding Articles 56, 57, 58, 59, 81 (Cap II, III and IV), 82 (Cap XIV), 91 and 92 of the Insurance Law (LGISMS). - Investment rules cover: formation and operations of investment committee, admitted assets list, rates for private companies’ values, currency and duration matching, custody requirements, investment administration and investment limits per issuer or security type. - Admitted assets (CNSF 2006a): - Securities registered in the SIC of the Mexican Stock Exchange. - Investments in equity and fixed income mutual funds with portfolios that include foreign securities issued by Central Banks or governments of the EU or IOSCO Technical Committee countries, or securities registered/authorized/regulated for general public sale by CNBV or equivalent. - Securities denominated in national currency issued by international financial institutions in which Mexico is a member. - Private equity investment funds in Equity Investment Societies (Sociedades de Inversión de Capitales, SINCAS), and trusts for firm capitalization. ### Investment limits (by type and issuer) - Table 9 (Investment Limits by Type of Security for Technical Provisions) — limits (in percent): - (1) Papers issued or backed by the Federal Government: 100 - (2) Papers issued or backed by credit institutions or international financial bodies: 60 - (3) Papers issued by SINCAS: 1 - (4) Papers issued or backed by institutions different from those in (1), (2), and (3): 30 - (5) Papers issued or backed by institutions different from those in (1), (2), and (3) exchange at the Mexican Stock Exchange: 5 - (6) Discount and rediscount operations: 5 - (7) Credit with pledge bonds or papers guarantee: 5 - (8) Credit with mortgage guarantee: 5 - (9) Real state properties: 25 - (10) Repurchasing: 30 - (11) Foreign debt papers: 2 - Table 10 (Investment Limits by Type of Issuer for Technical Provisions) — limits (in percent): - (1) Federal Government: 100.0 - (2) Banks or international financial bodies: 18.0 - (3) SINCAS: 0.5 - (4) Institutions different from those in (1), (2), and (3): 7.0 - (5) Groups and institutions in the same economic sector (shares) 1/ 2/: 10.0 - (6) Financial institution (shares) 1/: 5.0 - (7) Stock and papers issued inter-related groups: 18.0 - (8) Foreign securities belonging to investment funds: 1.0 - (9) Securities registered in the SIC (Sistema Internacional de Cotizaciones) of the Mexican Stock Exchange: 1.0 - Notes: - 1/ Constitute standard risk for mutual insurance companies. - 2/ For manufacturing, this limit can reach up to 20 percent. ### Liquidity, custody, matching, and credit risk rules - Liquidity requirements: - Companies must cover short term technical reserves with equivalent assets. - Short term obligations are obtained by assigning a percentage short term requirement to each type of reserve; total must be covered with investments with maturities of a year or less or equity shares classified as highly traded. - Historically, shares of reserves to be invested in liquid instruments varied from 100 percent for the loss reserve to 6 percent for the NPR, the contingency and the special mathematical reserves. - Starting with 2006, liquidity requirements must be calculated according to the calculation derived from the asset–matching analysis for pensions. - Liquidity requirements for life operations must be calculated for each type of life insurance policies issued, considering proportion of technical provisions from short term vs long term policies. - Custody and investment administration: - All titles and securities issued in the country must be administered by credit institutions or securities firms and kept in custody by depository institutions for securities. - For investments in foreign currency operated outside the country, only Mexican Financial Entities or subsidiaries may act as intermediaries; custodians may be authorized depositary entities in the applicable foreign jurisdiction. - Matching rules: - Solvency Margin Rules amended to include matching of assets and liabilities for life insurance starting January, 2006. - For pension annuity companies, matching rules established in 2000 include duration, inflation and currency matching of liabilities and admitted assets. - Obligations must be covered by assets in the same currency (indexed to inflation) with specific exceptions: - Before 1999, investment of technical reserve only in inflation indexed instruments was prescribed. After 1999, companies were allowed to invest in nominal instruments with yield higher than inflation if they acquire derivatives to cover inflation risk. - After 2000 companies were allowed to invest up to 10 percent of technical reserves in foreign currency instruments issued or guaranteed by the Federal Government. - Institutions may cover obligations in foreign currency with equivalent in local currency by acquiring derivative products exclusively for covering exchange rate risk with authorized intermediaries and recognized markets by the central bank. - Credit risk rules: - All private debt securities and titles in investment funds used to cover technical reserves must have a minimum acceptable rating determined by a rating agency approved by the Banking and Securities National Commission (CNBV). ### Investment composition and risk exposure evolution (1997–2005) - Table 11 (Investment of Technical Provisions, 1997–2005, constant prices as of December 2005; total MXN mln.): - Other Inv. (percent): 1997 = 1.5; 1998 = 1.5; 1999 = 4.7; 2000 = 4.9; 2001 = 6.8; 2002 = 7.8; 2003 = 10.6; 2004 = 15.3; 2005 = 11.8. - Private Sec. (percent): 1997 = 20.0; 1998 = 16.3; 1999 = 10.4; 2000 = 14.4; 2001 = 10.4; 2002 = 11.6; 2003 = 15.7; 2004 = 16.2; 2005 = 11.9. - Government Sec. (percent): 1997 = 78.5; 1998 = 82.2; 1999 = 84.9; 2000 = 80.7; 2001 = 82.8; 2002 = 80.6; 2003 = 73.7; 2004 = 68.5; 2005 = 76.3. - Total (MXN mln.) /1: 1997 = 3,602; 1998 = 16,191; 1999 = 29,713; 2000 = 43,743; 2001 = 60,039; 2002 = 69,061; 2003 = 78,112; 2004 = 84,112; 2005 = 79,852. - Table 12 (Inflation and Currency Risk in Technical Provisions, 1997–2005, percent): - Local nominal: 1997 = 36.1; 1998 = 40.7; 1999 = 14.0; 2000 = 4.0; 2001 = 5.1; 2002 = 4.8; 2003 = 0.2; 2004 = 0.5; 2005 = 0.4. - Foreign nominal: 1997 = 3.9; 1998 = 1.3; 1999 = 1.6; 2000 = 2.6; 2001 = 2.4; 2002 = 2.3; 2003 = 2.2; 2004 = 3.6; 2005 = 0.8. - Local real: 1997 = 60.0; 1998 = 58.0; 1999 = 84.4; 2000 = 93.1; 2001 = 92.2; 2002 = 92.6; 2003 = 97.4; 2004 = 94.2; 2005 = 91.3. - Local and foreign nominal: 1997 = 0.0; 1998 = 0.0; 1999 = 0.0; 2000 = 0.4; 2001 = 0.3; 2002 = 0.3; 2003 = 0.3; 2004 = 1.7; 2005 = 7.6. - Findings: - Most technical provisions are invested in government bonds. - Companies are required to fully hedge inflation and interest rate risk; only a very small currency mismatch is allowed by regulation between assets and liabilities. - Because private sector issuers of inflation linked long term bonds are essentially non existent, hedging is done by purchasing inflation linked instruments with long maturities mostly issued by the Federal Government. - Investment of technical reserves shifted over time from nominal and foreign assets to real and domestic assets, considerably reducing exposure to inflation and currency risk. *Source: CNSF (data and regulatory citations as presented in the chapter).* ### 48. For accounting purposes investment portfolios can be classified in two types: (a) for ### _cr07163 - 48. For accounting purposes investment portfolios can be classified in two types: (a) for ### Accounting classification and valuation - Investment portfolios are classified for accounting purposes as: - (a) for sale — used to finance operations and short term liabilities; assets are marked to market. - (b) to hold until maturity — used to back longer term liabilities; valuation based on a constant discount rate; assets are valued at maturity. - It is natural to classify in these circumstances all reserves in at least the UPR as held to maturity. - On average, 90 percent of the portfolio of pension annuity companies is classified as held to maturity. - Liabilities regarding basic benefits are valued at the fixed technical interest rate defined in regulations which is currently 3.5 percent real. ### Maturity profile and mismatch - Collected data show the evolution of average maturity for government and private sector bond portfolios (Tables 13 and 14; Source: CNSF). - Key reported totals (Average Maturity of Government Bond Portfolio / Total): - 1997: 6.14 - 1998: 7.51 - 1999: 6.73 - 2000: 13.07 - 2001: 14.71 - 2002: 16.36 - 2003: 14.76 - 2004: 14.88 - 2005: 14.53 - Reported "Total bond portfolio" row (duplicated across tables) exactly: - 1997: 5.80 - 1998: 7.30 - 1999: 6.50 - 2000: 11.80 - 2001: 13.70 - 2002: 15.00 - 2003: 13.40 - 2004: 13.60 - 2005: 14.20 - Average maturity of the whole bond portfolio increased over the last nine years from less than six years to more than 14 years. - The increase is more noticeable in the year 2000 when asset-liability matching rules were introduced for solvency margin calculation. - Average maturity of liabilities, estimated with expected longevity, is 37 years — creating a maturity mismatch with assets. ### Duration and interest rate risk management - Duration (modified, or Macaulay) is used to measure sensitivity to marginal changes in interest rates of option-free bond portfolios; defined as the approximate percentage change in price for a 100 basis point parallel shift in the yield curve. - Pension annuity companies hold large positions in government bonds UDIBONOS, PICs, and CBICs with maturities between seven and 28 years; these are fixed coupon non-callable bonds (no embedded options). Measuring interest sensitivity of CBICs with duration is appropriate. - Liability cash flows of pension annuity companies are not interest rate sensitive; duration is a good measure for liabilities. - Limitations of duration: - Accurate only when interest rate risk stems from parallel shifts in the yield curve. - Accurate only for small changes in interest rates. - Policy implication: CNSF may want to extend required metrics to measure interest rate sensitivity beyond duration. ### Extended metrics: convexity and key-rate (partial) duration - Convexity: - A second-order term that measures the change in price from the duration estimate for a small change in rates. - For positive-duration, no-embedded-option instruments, positive convexity implies duration increases when interest rates fall and decreases when rates rise. - Duration plus convexity help measure asymmetric percentage price changes for large interest rate changes. - Typical ALM requirement: convexity(assets) > convexity(liabilities) so assets lose less value than liabilities when rates rise and gain more when rates fall. - Key rate / partial duration: - Addresses non-parallel yield curve movements (twists, curvature/butterfly shifts). - Key rate duration measures local sensitivity to a shift in a portion of the yield curve; partial duration shocks the yield curve (key rate shocks the spot rate curve). - Matching partial durations of assets and liabilities provides protection against nonparallel shifts. - Constructed key rate shifts sum to a parallel shift; sum of key rate durations equals effective duration for fixed cash-flow instruments. ### Supervisory and company policy considerations - Full duration matching immunizes assets and liabilities but may conflict with: - Surplus management policies (companies may match in dollar duration to immunize surplus). - Yield enhancement tactics (companies may shorten asset duration if they expect rising rates). - Supervisory implication: Diversity of risk metrics is justified for prudential rules to accommodate tactical business decisions while maintaining oversight. ### Market performance — underwriting and investment interplay - Company performance determined jointly by underwriting and reserve investment performance. - Underwriting performance ratios used: loss ratio (LR), operational expense ratio (OR), underwriting expense ratio (UR), combined ratio (CR = LR + OR + UR). - Observed underwriting trends (1997–2005): - Average LR in 1997 exceeded 140 percent. - Underwriting performance deteriorated progressively until 2002 when CR exceeded 100. - Deterioration driven by a drastic fall in gross premium income since 2002 and delayed cost structure adjustments. - Table 15 — Evolution of Average Underwriting Costs, MXN Million, 2001–2005 (Note: 1/ Real 2005 pesos; Source: CNSF): - GPI: - 2001: 16,640.59 - 2002: 10,261.26 - 2003: 3,501.71 - 2004: 5,224.7 - 2005: 4,504.33 - Pure Underwriting: - 2001: 403.90 (Percent 2.4) - 2002: 288.70 (Percent 2.8) - 2003: 173.96 (Percent 5.0) - 2004: 162.5 (Percent 3.1) - 2005: 203.26 (Percent 4.5) - Additional Benefits: - 2001: 1,178.58 (Percent 7.1) - 2002: 695.98 (Percent 6.8) - 2003: 332.21 (Percent 9.5) - 2004: 156.14 (Percent 3.0) - 2005: 177.67 (Percent 3.9) - Total Underwriting: - 2001: 1,152.48 (Percent 9.5) - 2002: 984.68 (Percent 9.6) - 2003: 506.18 (Percent 14.5) - 2004: 318.64 (Percent 6.1) - 2005: 380.93 (Percent 8.5) - Observations: - Gross premium income in 2003 was 34 percent of gross premium income of 2002. - Pure underwriting costs in 2003 were 60 percent of the value in 2002. - Premiums for additional benefits in 2003 decreased to 48 percent of 2002 values. - By 2005 the overall underwriting expense ratio was 8.5 percent, lower than in 2001. ### Investment performance and solvency linkage - Combined ratios often exceed 100 percent; profitability thus depends on investment performance. - Example international data (Table 16 — Evolution of Average Non-Life Loss Ratios in Select Countries; Source: Holzeu Lechner (2006)): - Loss ratio: - USA (‘94 – ‘04): 78.7 - CAN (‘94 – ‘04): 73.3 - GBR (‘94 – ‘04): 73.0 - DEU (‘94 – ‘04): 71.3 - FRA (‘95 – ‘04): 80.8 - JPN (‘96 – ‘04): 61.2 - Expense ratio (comparable to UR + OR for Mexico): - USA: 26.4 - CAN: 29.8 - GBR: 31.2 - DEU: 27.5 - FRA: 23.3 - JPN: 38.1 - Combined ratio: - USA: 106.2 - CAN: 103.1 - GBR: 104.2 - DEU: 99.9 - FRA: 105.2 - JPN: 99.4 - Investment performance: - USA: 16.2 - CAN: 13.8 - GBR: 16.8 - DEU: 15.4 - FRA: 13.4 - JPN: 4.7 - Profit margin (pre tax): - USA: 9.2 - CAN: 10.4 - GBR: 12.8 - DEU: 12.7 - FRA: 7.0 - JPN: 4.4 - Mexican context: - Pension annuity companies invest reserves mainly in Government debt. - As of November 22 2005, JP Morgan EMBI+ for Mexico was 112 basis points. - After a record investment return performance of 20 percent nominal in 1998, average nominal rates over the 2002 and 2005 period decreased to 6.1 percent. - Implication: Deterioration in investment returns contributed to weakened overall company profitability and solvency metrics. *Source: CNSF (data) and _cr07163 PDF content.* ### 65. Pension annuity companies have large positions of Government bonds known as ### _cr07163 - 65. Pension annuity companies have large positions of Government bonds known as ### Investment positions and long-term real interest rates - Pension annuity companies hold large positions of Government bonds: UDIBONOS (Bonos De Desarrollo Del Gobierno Federal Denominados En Unidades De Inversión), PICs (Pagarés de Indemnización Carretera con aval del Gobierno Federal) and CBICs (Certificados Bursátiles de Indemnización Carretera con aval del Gobierno Federal). - Reported long-term real interest rate movements: - 10 yr Udibonos: 7.14 percent in October 1999; 4.30 in December 2005; 5.09 percent in June 2006. - PICs: around 6.13 percent in October 2001; around 5.94 percent in November 2002. - 20 yr CBICs: 6.16 percent in January 2003; 4.59 percent in December 2005. - 30 yr CBICs: 6.21 percent in January 2003; 4.68 percent in December 2005. - Effect on business: - Decrease in yield on fixed income paper is lowering profit margins, especially for new business. - This decrease has so far not translated into lowered regulatory solvency margins. ### Solvency framework and Minimum Guarantee Capital (CMG) - Solvency rules under the LGISMS require companies to establish a Minimum Guarantee Capital (Capital Mínimo de Garantía, CMG). - CMG purposes: - Cover deviations from expected claims. - Cover breakdowns in payments due to insolvent reinsurers. - Cover adverse fluctuations in asset valuation and asset-liability mismatches. - CNSF methodology: - Incorporates requirements to cover credit, inflation and interest rate risk in investment of technical provisions. - Establishes an asset-liability matching system with weights to measure impact of mismatches on capital. - Weights depend on availability of investment instruments of different maturities, volume and frequency of issues in the primary market, and availability and liquidity in the secondary market. - CNSF regularly updates and communicates the weighting system to industry as market conditions require. - CNSF observation: Ideally all weights should be equal to 1. If so, companies would need to set aside an extra amount of capital equivalent to 25 percent of current technical provisions (or MXN 18,410 million). ### Weighting system for asset-liability mismatches (as of March 2006) - Year of projection k and weight Wk: - k <= 5 : 1 - 5 < k <= 6 : 0.96 - 6 < k <= 7 : 0.94 - 7 < k <= 8 : 0.85 - 8 < k <= 9 : 0.78 - 9 < k <= 12 : 0.51 - 12 < k <= 14 : 0.46 - 14 < k <= 15 : 0.42 - 15 < k <= 17 : 0.37 - 17 < k <= 18 : 0.30 - 18 < k <= 19 : 0.25 - 19 < k <= 20 : 0.24 - 20 < k <= 21 : 0.23 - 21 < k <= 23 : 0.20 - 23 < k <= 24 : 0.19 - 24 < k <= 25 : 0.17 - 25 < k <= 27 : 0.13 - 27 < k <= 28 : 0.08 - 28 < k <= 29 : 0.02 - 29 < k <= 30 : 0.01 - k > 30 : 0.00 ### Assessment of the weighting approach and ALM - The CNSF methodology follows a regulatory maturity gap approach to asset-liability management. - Concerns and observations: - Unclear how this methodology complements or substitutes an economic duration approach that companies use for measuring and monitoring risks. - Regulations do not reference the impact of an economic ALM framework based on duration metrics on capital. - A maturity gap analysis that assigns zero or near-zero weights to long-term mismatches may underestimate interest rate risk from yield curve movements in long maturities. - Recommendation/observation: there may be a need to complement the regulatory maturity gap with an economic maturity gap analysis using unit weights at all maturities. - Despite limitations, the weighted maturity gap approach in Mexico is a marked improvement over ad hoc fixed paid up capital rules, but it is no substitute for a full fledged ALM framework for ongoing monitoring and control of asset and liability risks. ### Solvency and profitability indicators (1997–2005) - Pension annuities regulatory solvency position and profitability improved over time according to CNSF data. - Table 18 indicators (values preserved): - Solvency Margin (MXN million): - 1997: 667.2 - 1998: 292.5 - 1999: 811.3 - 2000: 862.0 - 2001: 570.8 - 2002: 829.1 - 2003: 2,176.8 - 2004: 1,761.2 - 2005: 2,927.5 - ROE (percent): - 1997: -8 - 1998: -29.2 - 1999: -20.2 - 2000: -19.8 - 2001: -6.4 - 2002: 15.6 - 2003: 22.5 - 2004: 26.1 - 2005: 24.3 - ROA (percent): - 1997: -2.9 - 1998: -3 - 1999: -1.2 - 2000: -0.9 - 2001: -0.3 - 2002: 0.7 - 2003: 2.9 - 2004: 4.0 - 2005: 1.9 ### Puzzling divergence: premiums, underwriting, yields vs. solvency margin - Observed contradictions: - Since 2002: sudden decrease in premiums, worsening underwriting profitability, and decrease in investment yields. - Yet regulatory solvency margins improved. - Possible explanations offered: - Asset size of the core portfolio is still growing quickly enough to compensate for reduced new business and other negative trends. - Companies are still earning high ROE and ROA on assets invested at historically higher yields. - Alternatively, economic solvency may have deteriorated in the medium term but regulatory solvency does not disclose this because of low weights on long-term mismatches. - Policy implication: - Adoption of solvency margins based on economic capital, at least for prudential purposes, would enable CNSF to better monitor economic solvency and take remedial actions. ### Pension annuity market development and pension law reforms - Two distinct periods in annuity market evolution: - 1997–2002: rapid growth of annuity market. - Post-2002: drastic drop in gross premium income. - Key changes affecting disability (IV) benefits and market growth: - Under LSS-73: disability benefits based on nominal average base salary of last 500 weeks; benefits originally indexed to minimum salary in the Federal District; eligibility for pension increments and family allowances. - 1997 reform (LSS-97): IV benefit defined as 35 percent of average individual base salary during last 500 weeks indexed to INPC; minimum IV pension defined as minimum salary in Federal District as of July 1, 1997 indexed to INPC; IMSS to purchase IV immediate premium annuities with accumulated reserve from a 2.5 percent contribution rate from private annuity providers. - Rationale: promote private sector annuity market development before new entrants retire. - Subsequent reforms and effects: - 2001: indexation base of LSS-73 benefits changed from minimum salary to INPC, increasing relative generosity of 1973 benefits versus 1997 benefits (since average wage growth after 1997 was lower than inflation during the referenced period). - 2002: reform allowed individuals who retired under LSS-73 to withdraw as a lump sum the balance of the 2 percent accumulated contributions for old age pension in the RCV subaccount, increasing relative generosity of LSS-73 provisions. - IMSS was given authority to interpret LSS-97 provisions and extended the option of electing LSS-73 benefits also to IV benefits. - Market outcome: - These reforms increased the relative attractiveness of LSS-73 benefits and changed growth prospects, contributing to decreased premium growth since 2002. ### Disability benefit comparisons and behavioral puzzle - Table 19 summary (preserving presented values where provided): - LSS-73 vs LSS-97 differences include salary base, basic pension, minimum pension (MXN 1,402 as of July 2004 for LSS-73; MXN 1,551 as of July 2004 for LSS-97), pension indexation INPC, SAR-92 lump sum treatments, and RCV subaccount lump sum options after SAR-02 reform. - Table 20 monthly disability benefit comparison (MXN) indicates: - Comparison suggests that on average the large majority of individuals would prefer LSS-97 benefits. - Only individuals with more than 15 years of contributions and more than twice the minimum salary would prefer LSS-73 (noting average salary in Mexico is between four and five times the minimum salary). - Despite LSS-97 generally being more generous, premiums fell and insured individuals often chose LSS-73 options. Reasons attributed to this "seeming irrational behavior" include informational and administrative factors. ### Informational, administrative, and eligibility policy drivers of premium decline - Possible causes for 28 percent annualized average decrease in premiums since 2001: - IMSS may not be informing insured individuals about available IV benefit options; Article IV of transition provisions of LSS-97 and IMSS’ position may not require the IMSS to compare benefits unless individuals request it. - LSS-97 contains ambiguities ("badly written" in several articles) that limit clarity on obligations to inform; only jurisprudence can establish correct interpretations. - IMSS increased use of temporary disability pensions and tightened eligibility rules in 2004; IMSS supports rehabilitation before declaring invalidity. - Consequences of IMSS policies: - (a) Number of potential private-sector annuitants is further reduced. - (b) The pool of annuitants covered by pension annuity providers is deteriorating. *Source: IMF staff summary of CNSF and legislation material in _cr07163 (excerpts provided).* ### 86. The decrease in the long term interest rate and the lower number of new annuitants ### _cr07163 - 86. The decrease in the long term interest rate and the lower number of new annuitants ### Impact on annuity profitability and underwriting performance - The decrease in the long term interest rate and the lower number of new annuitants are reducing profit margins of annuity companies, in part because during the early years of marketing of a new product, companies tend to underprice in an effort to gain market share. - The lower profitability of the initial products is recuperated through the yield on income paper and the business growth. - Evidence of average deterioration of underwriting performance is presented in this paper. IMSS’ behavior has reduced the opportunity of higher profits for insurance companies by reducing the scope for business growth. - The deterioration of the pool of annuitants also impacts profitability: - Insurance companies make money by pooling different types and charging a price that on average generates reasonable profits. - If ex post variation is only stochastic, the pool is stable. IMSS’ behavior has introduced a systemic element in the pool of annuitants. - Use of temporary pensions delays entry of disabled individuals in the pool. Since seriously disabled individuals tend to die early, the delayed entry implies that, on average, individuals insured by private sector companies tend to live longer than expected. - Even if initial technical parameters required by regulation were correct, delayed entry makes the product non profitable or less profitable. ### Mechanics and fiscal incidence of choices under LSS-73 and LSS-97 (transitional cohort) - If an individual elects benefits under LSS-73 (Case E): - The individual withdraws the cash balance in the SAR-92 account and the old age cash balance in the RCV subaccount. - The rest of the cash balances in the housing and RCV subaccounts are collected by the Federal Government that will grant the annuity or the minimum pension guarantee (PMG). - The IMSS is not required to transfer to the Federal Government the portion of the reserve accumulated by the 2.5 percent contribution of the insured individual declared invalid. - If an individual elects benefits under LSS-97, four cases arise: - Case A (PB>PMG, CB>MC): IMSS buys PB; insured collects SAR-92; IMSS collects CB in housing and RCV subaccounts; IMSS retains the difference with the MC. - Case B (PB>PMG, CBMC): IMSS buys PMG; insured collects SAR-92; IMSS collects CB of housing and RCV subaccounts; IMSS retains the difference with the MC. - Case D (PB