## _cr14320

## Source details

**Canonical URL:** [_cr14320](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14320.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14320.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14320.pdf.json)

---

### Overview and purpose
- Examines Mexico’s recently adopted energy reform and its implications for hydrocarbons production, investment costs, and foreign direct investment (FDI).
- Paper tasks:
  - Discuss the nature of the reforms and the problems they address.
  - Present illustrative production scenarios for crude oil and natural gas.
  - Estimate commensurate investment costs and FDI associated with each scenario.
  - Examine natural gas and electricity distribution markets.
  - Draw key messages and policy implications.

### Current energy-sector challenges
- Crude oil:
  - Peak production in 2004: 3.4 million barrels per day (mmbd).
  - Production fallen to 2.5 mmbd.
  - PEMEX has about a decade’s worth of proven crude oil reserves under current estimates.
  - If reserves-to-production ratio held constant at the 5-year average, production could fall to about 1.5 mmbd and Mexico could become a net oil importer before the end of the decade.
- Natural gas:
  - Production fell to 6.4 billion cubic feet per day (bcfd) from a peak of 7.0 bcfd in 2010.
  - Mexico has been a natural gas importer since 2002; imports grew significantly since 2009.
  - Imports primarily pipeline from the US and, at the margin, LNG from Qatar, Peru, and Nigeria.
  - Domestic delivery system experienced critical shortages with negative impacts on industrial production (notably in 2013).
- Electricity:
  - Mexican consumers pay significantly higher electricity prices than northern neighbors.
  - Service efficiency weaker than many OECD countries; transmission and distribution system relatively old, little investment, and relatively high losses.
  - New generating capacity largely natural-gas based; diesel and fuel oil remain used at the margin.

### Scope and principles of the reform
- Timeline:
  - Constitutional amendments approved in December 2013; secondary legislation passed in August 2014.
- Reform principles:
  - Nation ownership of hydrocarbons in the ground.
  - Promotion of open and competitive markets across upstream, midstream, and downstream operations.
  - Strengthening regulatory framework and institutions; transformation of PEMEX and CFE.
  - Transparency and accountability, industrial safety, and environmental protection.
- Key practical changes:
  - Market opening and auctions:
    - Round 0 assigned over 80 percent of Mexico’s proven and probable oil reserves to PEMEX.
    - Round 1 auctions for remaining E&P blocks; bidding process expected to be completed by the second half of next year (authorities’ timeline in the SIP).
    - Contracts to include profit- and production-sharing as well as licenses.
    - Electric generation market opened to allow IPPs and self-generators to sell directly.
    - Starting in 2018, domestic gasoline prices to become fully market-determined; PEMEX’s monopsony on gasoline imports to disappear.
  - Transformation of state enterprises:
    - PEMEX and CFE reconstituted as state productive enterprises with greater autonomy.
    - Fiscal take from PEMEX to be lowered to 65 percent over the next 5 years as new fiscal regimes take hold.
    - PEMEX allowed joint ventures and farmouts; CFE permitted to contract private parties for natural gas supply and T&D projects.
  - Regulatory strengthening and institutions:
    - Enhanced roles for SENER and CNH; CNH to conduct transparent auctions and manage contracts.
    - Creation of independent system operators: CENACE and CENEGAS.
    - CRE to set tariffs for transmission, distribution, and ancillary services.
  - Domestic content and participation rules:
    - Domestic content rules to gradually rise to 35 percent by 2025.
    - Minimum participation rule of 20 percent for PEMEX in deepwater trans-boundary projects.
  - Fiscal and transparency mechanisms:
    - Creation of the Mexican Oil Stabilization Fund, managed by the central bank, to administer proceeds and payments from assignments and contracts.

### Resource endowments and potential
- PEMEX-reported reserves (as of end-2013):
  - Proven reserves (≥90 percent probability): 10 billion barrels (bbl) of crude oil and 13.6 trillion cubic feet (tcf) of natural gas.
  - Probable and possible reserves: 21 bbl (oil) and 38.5 tcf (gas).
- Additional potential (PEMEX estimates):
  - Deepwater Gulf of Mexico (GoM): 27.1 billion barrels of oil equivalent (bboe).
  - Shale deposits (northern Mexico): 60.2 bboe.
- U.S. EIA rankings cited:
  - Mexico ranked 8th with 13 bbl of technically recoverable shale oil resources and 6th with 545 tcf of technically recoverable natural gas.
- Caveat: Deepwater and shale potential require substantially more exploratory drilling to better quantify recoverable volumes.

### Timing and development pathways
- Short- to medium-term improvements expected from:
  - Developing conventional fields.
  - Secondary and enhanced recovery in existing producing fields.
  - PEMEX contracting with private firms and farmouts to shorten timelines and access technology.
- Authorities indicated about 70 percent of blocks in Round 1 will be 2P (probable) reserves.
- Typical timelines for deepwater and shale (based on Goldman Sachs experience cited):
  - Exploration: 2016–2018.
  - Decision to commercially develop: 1–2 years after discovery.
  - Small production: 5–10 years after contracts are won.
  - Robust production: after that initial decade.

### Baseline production scenarios and key projections
- Crude oil baseline:
  - 2013 production: 2.5 mmbd.
  - 2014 projection: 2.35 mmbd.
  - 2015 projection: 2.4 mmbd.
  - Production reaches 3.0 mmbd by 2019 and 3.5 mmbd by 2025.
  - Proven R/P 5-year historical average used: 10.9 years.
  - Required reserve replacement ratios (RRR) to meet baseline:
    - Average 159 percent between 2016 and 2019.
    - Average 128 percent between 2020 and 2025.
- Natural gas baseline:
  - Initial baseline assumes 6.5 bcfd rising to:
    - 8.0 bcfd by 2018.
    - 10.4 bcfd by 2025.
  - Proven R/P 5-year historical average used: 5.5 years.
  - Required RRR for natural gas:
    - Average 126 percent between 2016 and 2019.
    - Average 120 percent thereafter.

### Downside scenario (crude oil and natural gas)
- Crude oil downside assumptions:
  - PEMEX production in 2015 stays at the 2014 projected level of 2.35 mmbd, and this level continues between 2016 and 2025.
  - From 2016 to 2019, additional production above 2.35 mmbd comes from half of the 2P reserves in the Round 1 blocks announced on August 13, 2014 spread over the 4 years.
  - Production increases to 2.82 mmbd by 2019.
  - From 2020 to 2025, conventional, shale, and deepwater contribute following typical schedule; production rises to 3.33 mmbd by 2025.
  - Proven R/P stays constant at the historical 5-year average of 10.9 years.
  - Scenario equivalence: government achieves its production goals with a 2-year delay (3.0 mmbd in 2021 vs. 2019 and 3.5 mmbd in 2027 vs. 2025).
  - Required RRR under this scenario:
    - Average 149 percent between 2015 and 2019.
    - Average 130 percent from 2019 to 2025.
- Natural gas downside assumptions:
  - Natural gas production stays constant at the projected 2014 level of 6.5 bcfd between 2015 and 2018.
  - No additions to production in the first few years; additions in this period are only in crude oil.
  - Natural gas production increases from 2019 to 2025, reaching 9.5 bcfd in 2025.
  - Shale gas contributes starting in 2021.
  - Proven R/P assumed constant at historical 5-year average of 5.5 years.
  - Government’s gas goals under this scenario:
    - 8.0 bcfd reached only after 2021.
    - 10.4 bcfd in 2027.
  - RRR of natural gas under this scenario:
    - 100 percent in the first period.
    - 129 percent in the latter period.

### Investment needs and cost assumptions
- Cost assumptions:
  - Exploration of new conventional fields: about $20 per barrel.
  - Advanced recovery techniques: between 15 to 25 dollars per barrel (assumed same for crude oil and natural gas, particularly associated gas).
  - Shale development cost: about $11 to $20 per boe on average.
  - Deep water development cost: $9 to $20 per boe.
- Investment requirements (baseline given costs and RRRs):
  - Phase 1 (2015–2019): estimated to require investment of about $40 billion per year.
  - Phase 2 (2020–2025): estimated to require about $50 billion per year.
  - These investments include roughly $25 billion per year already estimated in PEMEX E&P capital expenditure plan.
  - Investment schedule shows a hump in 2021 reflecting timing for shale and deepwater investment.
  - Downside scenario: average annual investment needs are about 10 percent less per year.

### Foreign direct investment (FDI) expectations and domestic share
- 2016 to 2019:
  - PEMEX likely to lead in conventional fields and recovery, with private partners.
  - Foreign share of investment likely less than 30 percent.
  - Starting in 2016 estimated related-FDI will increase by about $10 to $15 billion from current levels.
- 2020 to 2025:
  - Domestic content requirement assumed to increase from 25 percent to target of 35 percent by 2025.
  - For deepwater, minimum 20 percent PEMEX participation in trans-boundary projects.
  - Estimated FDI increases:
    - About $10 to $15 billion from current levels between 2016 and 2019.
    - About $20 to $30 billion between 2020 and 2025.
  - Downside scenario: FDI would on average be about 20 percent less annually over the 10 years.
- Comparison to other estimates:
  - Analysts’ estimates of annual FDI from energy reform range from less than US$10 billion per year to $30 billion or more; report’s estimates align with the lower end.
  - Estimates cover only investments into development of oil and natural gas fields, not the wider scope of the energy reform.

### Natural gas imports, transport, and infrastructure
- Demand and supply trends:
  - Since 2010, natural gas demand grew at 6.5 percent per year on average in real terms, versus a 2.4 percent average decrease in production.
  - Instituto Mexicano del Petroleo projects demand to grow at an average annual rate of 3.9 percent between 2015 and 2025.
  - Baseline production growth: 4.3 percent per year.
  - Baseline implication by 2025: imports fall 9 percent compared with 2013, but remain at 2.9 bcfd.
  - Downside implication by 2025: imports increase by 24 percent relative to 2013.
- Import composition and prices:
  - Majority of imports via pipeline from the U.S.; LNG imports peaked in 2010 at 0.55 bcfd.
  - In 2013 LNG’s share of natural gas imports peaked at 26 percent.
  - 2013 prices: pipeline gas from the U.S. cost $3.9 per thousand cubic feet; LNG cost $13.3 per thousand cubic feet.
- Pipeline capacity and implications:
  - U.S. pipeline gas prices have fallen over 50 percent since 2008 due in part to unconventional supply.
  - The U.S. exported 1.8 bcfd to Mexico in 2013, with about 5.2 bcfd in additional pipeline capacity being built between 2013 and 2016; delays would postpone Mexico’s ability to take advantage of lower-priced U.S. gas.
- Institutional and infrastructure reforms:
  - Cenegas will manage natural gas traffic to make allocations more efficient.
  - Transport and storage markets open to private participants.
  - SENER plans to build domestic pipeline infrastructure to connect to U.S. pipelines and expand transport within Mexico.
  - Reforms aim to prevent strains experienced in 2013 (critical alerts) that affected industrial users and electricity generators.

### Electricity sector implications and reforms
- Link between gas and electricity prices:
  - Lower natural gas prices will help lower electricity prices; largest natural gas users are electricity generators.
  - Natural gas combined cycle plants have been main choice for new generation since the 1990s.
  - Private self-generators projected by SENER to have fastest growth in natural gas demand over the next decade, displacing diesel and fuel oil.
- Market liberalization and grid operations:
  - Generation market opened to competition; IPPs and self-generators can sell directly to the market.
  - Authorities considering expanding the scope of qualified users who can buy electricity in the wholesale market.
  - Grid condition:
    - Just under half of transmission lines more than 30 years old; less than 10 percent added in the last five years.
    - CFE plans to expand system by about 1 percent per year over the next decade.
    - Cenace to be transformed into an independent system operator to dispatch based on efficiency or emergency needs.
    - CRE will set tariffs on transmission, distribution, and ancillary services and can provide incentives to reduce transmission losses.
- CFE institutional changes:
  - Conversion into a state productive enterprise to give more operational and budgetary autonomy and expanded ability to contract with third parties.
  - CFE tasked with adopting international standards for management of state enterprises.

### Manufacturing impacts: elasticities and scenario outcomes
- Estimated elasticities of manufacturing output to energy prices:
  - Electricity price elasticities range from -0.11 to -0.28 (up to -0.28).
  - Natural gas elasticities range from -0.04 to -0.07.
  - Metals and transportation equipment show electricity-price elasticity of -0.51.
  - Chemical sector elasticity about -0.1 and not statistically significant.
- Fuel-substitution potential and price effects:
  - Fuel oil represents 18.1 percent of total generation; natural gas is about 71 percent cheaper than fuel oil.
  - Immediate substitution of fuel oil for natural gas would imply, ceteris paribus, a reduction of about 13 percent in electricity prices (0.181 * -0.71).
  - Given planned pipeline investments and plant conversions, a 13 percent reduction seems feasible within five years.
  - Assuming convergence to U.S. prices as a limiting case, combined gains could lead to a reduction of about 50 percent in commercial and industrial electricity prices in total.
- Scenario impacts on manufacturing and GDP:
  - Scenario 1 — 13 percent electricity-price reduction (fuel oil → natural gas):
    - Increase in manufacturing output: 1.4 percent to 3.6 percent.
    - Increase in overall GDP: 0.2 percent to 0.6 percent.
    - Timing: gains plausible over 2016–2019.
  - Scenario 2 — Convergence to U.S. levels:
    - Increase in manufacturing output: 5.5 percent to 14.0 percent.
    - Increase in overall GDP: 0.9 percent to 2.2 percent.
    - Timing: more uncertain; depends on reducing energy losses in T&D.
- Micro- and indirect channels:
  - Panel VAR analysis shows allowing unit labor costs to respond endogenously amplifies output responses; by the fifth year cumulative effect about 50 percent larger than impact effect.
  - No statistically significant direct sectoral or regional spillovers detected under the employed methodology.

### Empirical approach (VAR and regression evidence)
- Panel regression of real manufacturing GDP by industry i, state j, year t includes lagged dependent variable, lagged changes in electricity (EL), natural gas (NG), and oil derivatives (OD) prices (deflated by CPI), plus controls (unit labor costs, real effective exchange rate, cost of capital, U.S. industrial production), and fixed effects at sector-state pair level.
- Panel VAR specification:
  - Variables: manufacturing output, real interest rates, U.S. industrial production, unit labor costs, electricity prices.
  - Identification: Choleski ordering; impulse responses for a 1% shock to electricity prices over 5 years; 90% confidence intervals via parametric bootstrap.
  - Variants: subsector spillovers (13x1 vector) and regional spillovers (5x1 vector).
- Selected energy consumption statistics (industrial sector, Petajoules):
  - Industrial sector total: 1,381.1 (2010), 1,492.3 (2011), 1,530.6 (2012).
  - Shares (2010, 2011, 2012): Electricity: 34.2% ; 33.6% ; 34.5%; Natural Gas: 35.3% ; 35.2% ; 35.8%; Oil derivatives: 17.0% ; 15.5% ; 15.6%.
  - Auto industry energy use (Petajoules): 10.5 (2010), 12.7 (2011), 14.4 (2012); shares show high electricity reliance (e.g., 68.3% in 2010).
- Regression sample: 1998–2012 (annual; N 3,584) with various specifications; selected coefficient magnitudes preserved in source tables.

### Key conclusions and policy implications (energy and manufacturing)
- Reform potential:
  - Comprehensive reform can support faster growth, better living standards, and greater energy security.
  - Short-run objective: overcome risk of falling hydrocarbons production and improve fiscal outlook.
  - Medium- to long-run objective: enable exploitation of shale and deepwater potential and reduce domestic energy costs and improve services.
- Short-run priorities:
  - Prioritize development of conventional and enhanced-recovery projects for near-term production gains; Round 1 focuses on auctioning 2P fields.
  - PEMEX has more freedom to partner with third parties to increase investment and import technologies.
- Institutional and implementation priorities:
  - Ensure transparency and perceived fairness in Round 1 auctions to maintain investor and political confidence.
  - Implement and operationalize CNH, CENACE, CENEGAS, and CRE to manage auctions, contract administration, and distribution system efficiency.
  - Use domestic content rules and PEMEX participation requirements to balance technology transfer, domestic industry development, and attractiveness to foreign investors.
  - Leverage the Mexican Oil Stabilization Fund to increase transparency of oil-related revenues and potentially smooth fiscal impacts.
- Infrastructure and market access:
  - Increase gas pipeline capacity to allow larger natural gas imports from the U.S.
  - Convert existing fuel oil–based plants to natural gas to realize substitution gains.
  - Open transport and storage markets to private participants while ensuring open access to pipeline capacity.
  - Independent system operators for gas and electricity can reduce critical alerts and enhance service delivery.
- Risks and caveats:
  - Patience needed for shale and deepwater given long lead times.
  - Delays or problems that dampen investor confidence could reduce exports and fiscal revenues (stylized downside scenarios illustrate lower production, investment, and FDI).
  - Reducing distribution losses will be challenging; efficiency gains depend on incentives and performance of new administrators.

### Gross capital inflows and investor behavior (portfolio flows)
- Recent patterns:
  - Gross capital inflows (particularly portfolio inflows) to Mexico have increased significantly in recent years.
  - Mexico has strong links to the United States, its largest trading partner and main source of portfolio and FDI.
  - Inclusion in Citigroup’s World Government Bond Index (WGBI) in 2010 attracted new groups of foreign investors.
  - Share of non-resident holdings of domestic sovereign debt reached 36 percent in April 2014.
  - Pension, insurance, and mutual funds now account for about half of the financial system (more than 40 percent of GDP).
- Volatility and risks:
  - Portfolio inflows have become more volatile with episodes around the GFC, 2011 European crisis, and May 2013 Fed tapering announcement.
  - Excessive volatility can amplify cycles and financial vulnerabilities.
- Mutual fund micro-evidence:
  - EPFR-based fund-level analysis (monthly/weekly) shows foreign mutual funds more likely to herd and engage in positive-feedback trading during stress.
  - During September 2008 after Lehman collapse, around 75 and 95 percent of equity and bond funds active in Mexico were net sellers, respectively.
  - During May 2013 tapering announcement: about 50 percent of domestic mutual funds were selling, while net sellers among foreign mutual funds rose above 70 percent.
  - Fund-flow regressions (sample 2011 January – 2014 March) show foreign bond funds: Lagged Mex 3-month bond yield coefficient 0.00725***; Lagged bond yield * Stress 0.0171***; Stress -0.0677***; R-squared 0.178; Number of Observations 3422.
- Sovereign bond volatility and investor shares:
  - Foreign participation amplifies the impact of global financial shocks on Mexican sovereign bond yield volatility (OLS and MGARCH results).
  - MGARCH: higher foreign participation increases sensitivity of 10-year Mbono yield volatility to a VIX shock.
  - Domestic investor roles mixed: pension funds and banks may dampen some VIX shocks in some specifications, but effects depend on shock type and are not uniformly robust.
- Policy implications (capital flows):
  - Mexico remains vulnerable to global financial shocks due to integration and volatile gross portfolio inflows.
  - Policy credibility, exchange rate flexibility, high reserves, and FCL arrangements are important buffers.
  - Developing a deep and diverse domestic investor base is desirable; domestic investors can play a stabilizing role.

*Source: IMF staff report text in content unit _cr14320.*

### References _______________________________________________________________________________ 18

### References

### Overview and purpose
- This Selected Issues Paper (SIP) examines Mexico’s recently adopted energy reform and its implications for hydrocarbons production, investment costs, and foreign direct investment (FDI).
- The SIP:
  - Discusses the nature of the reforms and the problems they address.
  - Presents illustrative production scenarios for crude oil and natural gas.
  - Estimates commensurate investment costs and FDI associated with each scenario.
  - Examines natural gas and electricity distribution markets.
  - Concludes with key messages and policy implications.

### Current energy-sector challenges
- Crude oil:
  - Peak production in 2004: 3.4 million barrels per day (mmbd).
  - Production has fallen to 2.5 mmbd.
  - PEMEX has about a decade’s worth of proven crude oil reserves under current estimates.
  - Without significant reserve additions and if reserves-to-production ratio is held constant at the 5-year average, production could fall to about 1.5 mmbd and Mexico could become a net oil importer before the end of the decade.
- Natural gas:
  - Production fell to 6.4 billion cubic feet per day (bcfd) from a peak of 7.0 bcfd in 2010.
  - Mexico has been a natural gas importer since 2002; imports have grown significantly since 2009.
  - Imports are primarily pipeline from the US and, at the margin, higher-priced LNG from Qatar, Peru, and Nigeria.
  - The domestic delivery system has experienced critical shortages with negative impacts on industrial production (notably in 2013).
- Electricity:
  - Mexican consumers pay significantly higher electricity prices than northern neighbors.
  - Service efficiency is weaker than many other OECD countries.
  - Transmission and distribution system is relatively old, receives little investment, and experiences relatively high losses.
  - New generating capacity has been largely natural-gas based, but diesel and fuel oil generation remain used at the margin.

### Scope and principles of the reform
- Constitutional amendments approved in December 2013; secondary legislation passed in August 2014.
- Reform principles include:
  - Nation ownership of hydrocarbons in the ground.
  - Promotion of open and competitive markets across upstream, midstream, and downstream operations.
  - Strengthening regulatory framework and institutions; transformation of PEMEX and CFE.
  - Transparency and accountability, industrial safety, and environmental protection.
- Key practical changes:
  - Market opening and auctions:
    - Round 0 assigned over 80 percent of Mexico’s proven and probable oil reserves to PEMEX.
    - Round 1 auctions to begin for remaining E&P blocks; bidding process expected to be completed by the second half of next year (per authorities’ timeline in the SIP).
    - Contracts to include profit- and production-sharing as well as licenses; flexibility intended to attract foreign firms.
    - Electric generation market opened to allow IPPs and self-generators to sell directly to the market.
    - Starting in 2018, domestic gasoline prices to become fully market-determined; PEMEX’s monopsony on gasoline imports to disappear.
  - Transformation of state enterprises:
    - PEMEX and CFE reconstituted as state productive enterprises with greater autonomy.
    - Fiscal take from PEMEX to be lowered to 65 percent over the next 5 years as new fiscal regimes take hold.
    - PEMEX allowed to enter joint ventures and farmouts; CFE permitted to contract private parties for natural gas supply and T&D projects.
  - Regulatory strengthening:
    - Enhanced roles for the Ministry of Energy (SENER) and the National Hydrocarbons Commission (CNH); CNH to conduct transparent auctions and manage contracts.
    - Creation of independent system operators: CENACE and CENEGAS.
    - The Energy Regulatory Commission (CRE) to set tariffs for transmission, distribution, and ancillary services.
  - Domestic content and participation rules:
    - Domestic content rules to gradually rise to 35 percent by 2025.
    - Minimum participation rule of 20 percent for PEMEX in deepwater trans-boundary projects to facilitate know-how transfer.
  - Fiscal and transparency mechanisms:
    - Creation of the Mexican Oil Stabilization Fund, managed by the central bank, to administer proceeds and payments from assignments and contracts.

### Resource endowments and potential
- PEMEX-reported reserves (as of end-2013):
  - Proven reserves (≥90 percent probability): 10 billion barrels (bbl) of crude oil and 13.6 trillion cubic feet (tcf) of natural gas.
  - Probable and possible reserves: 21 bbl (oil) and 38.5 tcf (gas).
- Additional potential:
  - Deepwater Gulf of Mexico (GoM): 27.1 billion barrels of oil equivalent (bboe) (PEMEX estimate).
  - Shale deposits (northern Mexico): 60.2 bboe (PEMEX estimate).
  - U.S. Energy Information Agency rankings cited: Mexico ranked 8th with 13 bbl of technically recoverable shale oil resources and 6th with 545 tcf of technically recoverable natural gas.
- Caveat: Deepwater and shale potential require substantially more exploratory drilling to better quantify recoverable volumes.

### Timing and development pathways
- Short- to medium-term production improvements expected primarily from:
  - Developing conventional fields.
  - Secondary and enhanced recovery in existing producing fields.
  - PEMEX contracting with private firms and farmouts to shorten timelines and access technology.
- Authorities indicated about 70 percent of blocks in Round 1 will be 2P (probable) reserves, more ready for extraction.
- Typical timelines for deepwater and shale (based on Goldman Sachs experience cited):
  - Exploration: 2016–2018.
  - Decision to commercially develop: 1–2 years after discovery.
  - Small production: 5–10 years after contracts are won.
  - Robust production: after that initial decade.

### Baseline production scenarios and key projections
- Crude oil baseline assumptions and projections:
  - 2013 production: 2.5 mmbd.
  - 2014 projection: 2.35 mmbd.
  - 2015 projection: 2.4 mmbd.
  - Production reaches 3.0 mmbd by 2019 and 3.5 mmbd by 2025.
  - Proven reserves-to-production (R/P) 5-year historical average used: 10.9 years.
  - Required reserve replacement ratios (RRR) to meet baseline:
    - Average 159 percent between 2016 and 2019.
    - Average 128 percent between 2020 and 2025.
- Natural gas baseline assumptions and projections:
  - Historical production noted as 6.4 bcfd (from earlier text) and the baseline assumes 6.5 bcfd initial rising to:
    - 8.0 bcfd by 2018.
    - 10.4 bcfd by 2025.
  - Proven R/P 5-year historical average used: 5.5 years.
  - Required RRR for natural gas:
    - Average 126 percent between 2016 and 2019.
    - Average 120 percent thereafter.

### Analytical approach and key questions
- Analysis asks:
  - Are there enough potential reserves given current geological estimates?
  - What is the timeline for production by type (conventional, enhanced recovery, deepwater, shale)?
  - What proven reserve replacement ratios (RRRs) would be required to meet production targets, and how do those RRRs compare to historical trends?
  - What investment costs and FDI would be implied to attain those RRRs, given domestic content rules?

### Policy implications and implementation priorities (drawn from analysis)
- Ensure transparency and perceived fairness in Round 1 auctions to maintain investor and political confidence.
- Prioritize development of conventional and enhanced-recovery projects for near-term production gains, while progressing regulatory and contractual reforms to unlock deeper, longer-term deepwater and shale potential.
- Implement and operationalize new regulatory institutions (CNH, CENACE, CENEGAS, CRE) to manage auctions, contract administration, and distribution system efficiency.
- Use domestic content rules and PEMEX participation requirements to balance technology transfer, domestic industry development, and attractiveness to foreign investors.
- Leverage the Mexican Oil Stabilization Fund to increase transparency of oil-related revenues and to potentially smooth fiscal impacts.

* _cr14320 - References _______________________________________________________________________________ 18_*

### 14.      Between 2015 and 2019, additions to proven reserves will likely come from the

### _cr14320 - 14.      Between 2015 and 2019, additions to proven reserves will likely come from the

### Reserves, production outlook, and downside scenario
- Between 2015 and 2019, additions to proven reserves will likely come from existing fields and 2P reserves, produced by Pemex and its partners or new entrants; oil is likely to come from conventional fields and the application of secondary and enhanced recovery techniques on existing fields.
- Between 2020 and 2025, shale and deep water sources are likely to come into play, developed and produced by firms winning fields from federal government auctions.
- Downside scenario assumptions:
  - Pemex production in 2015 stays at the 2014 projected level of 2.35 mmbd, and this level continues between 2016 and 2025.
  - From 2016 to 2019, additional production above 2.35 mmbd comes from half of the 2P reserves in the Round 1 blocks announced on August 13, 2014 spread over the 4 years.
  - Production increases to 2.82 mmbd by 2019.
  - From 2020 to 2025, conventional, shale, and deepwater contribute to production following the typical production schedule shown in Figure 4; production rises to 3.33 mmbd by 2025 (Figure 2).
  - Proven reserves to production ratio stays constant at the historical 5 year average of 10.9 years.
  - Scenario equivalence: government achieves its production goals with a 2-year delay (3.0 mmbd in 2021 vs. 2019 and 3.5 mmbd in 2027 vs. 2025).
  - Required RRR for crude oil under this scenario:
    - Average 149 percent between 2015 and 2019.
    - Average 130 percent from 2019 to 2025 (Figure 2).
- Natural gas in the downside scenario:
  - Natural gas production stays constant at the projected 2014 level of 6.5 bcfd between 2015 and 2018.
  - No additions to production in the first few years; additions to proven reserves in this period are only in crude oil.
  - Natural gas production increases from 2019 to 2025, reaching 9.5 bcfd in the last year.
  - Shale gas contributes to production starting in 2021, consistent with the longer end of the 5 to 10 year range between auction and start of production.
  - Proven reserves to production ratio for gas assumed constant at the historical 5-year average of 5.5 years.
  - Government’s gas goals under this scenario:
    - 8.0 bcfd reached only after 2021.
    - 10.4 bcfd in 2027 (Figure 2).
  - RRR of natural gas under this scenario:
    - 100 percent in the first period.
    - 129 percent in the latter period (Figure 2).

### Investment needs and cost assumptions
- Cost assumptions:
  - Exploration of new conventional fields: about $20 per barrel (EIA, South America and the U.S.).
  - Advanced recovery techniques: between 15 to 25 dollars per barrel (industry analysts). Assumed same for crude oil and natural gas (particularly for associated natural gas).
  - Shale development cost: about $11 to $20 per boe on average (Goldman Sachs-based cost curves).
  - Deep water development cost: $9 to $20 per boe (Goldman Sachs-based cost curves).
- Investment requirements (baseline assumptions given above costs and RRRs):
  - Phase 1 (2015–2019): estimated to require investment of about $40 billion per year.
  - Phase 2 (2020–2025): estimated to require about $50 billion per year.
  - These investments include the roughly $25 billion per year already estimated in the capital expenditure plan for E&P of PEMEX.
  - The investment schedule shows a hump in 2021 reflecting timing of increased investment for shale and deepwater per the cost curves used.
  - Downside scenario: average annual investment needs are about 10 percent less per year.

### Foreign direct investment (FDI) expectations and domestic share
- Phase differentiation and FDI share:
  - 2016 to 2019:
    - PEMEX likely to lead in conventional fields and recovery, with private partners or contracts with private firms.
    - Share of investment costs borne by foreigners likely smaller than in phase 2; assumed less than 30 percent of investment from foreigners.
    - Starting in 2016 estimated related-FDI will increase by about $10 to $15 billion from current levels.
  - 2020 to 2025:
    - Domestic content requirement assumed to increase from 25 percent to target of 35 percent by 2025 as written in the law.
    - For deepwater, a minimum of 20 percent participation by Pemex is required in trans-boundary projects.
    - Estimated FDI increases:
      - About $10 to $15 billion from current levels between 2016 and 2019.
      - About $20 to $30 billion between 2020 and 2025.
    - Downside scenario: FDI would on average be about 20 percent less annually over the 10 years.
- Comparison to other estimates:
  - Analysts’ estimates of annual FDI from energy reform range from less than US$10 billion per year to $30 billion or more; the report’s estimates align more with the lower end.
  - Note: estimates cover only investments into development of oil and natural gas fields, not the wider scope of the energy reform.

### Natural gas imports, transport, and infrastructure
- Demand and supply trends:
  - Since 2010, natural gas demand has been growing at 6.5 percent per year on average in real terms, versus a 2.4 percent average decrease in production.
  - Instituto Mexicano del Petroleo projects demand to grow at an average annual rate of 3.9 percent between 2015 and 2025.
  - Baseline scenario production growth: 4.3 percent per year.
  - Baseline implication by 2025: imports fall 9 percent compared with 2013, but remain at 2.9 bcfd.
  - Downside implication by 2025: imports increase by 24 percent relative to 2013.
- Import composition and prices:
  - Majority of Mexico’s natural gas imports are via pipeline from the U.S.; LNG imports have grown since 2006, peaking in 2010 at 0.55 bcfd.
  - In 2013 LNG’s share of natural gas imports peaked at 26 percent.
  - 2013 prices: pipeline gas from the U.S. cost $3.9 per thousand cubic feet; LNG cost $13.3 per thousand cubic feet.
- Pipeline capacity and implications:
  - U.S. pipeline gas prices have fallen over 50 percent since 2008 due in part to unconventional supply (Bakken and Eagle Ford).
  - The U.S. exported 1.8 bcfd to Mexico in 2013, with about 5.2 bcfd in additional pipeline capacity being built between 2013 and 2016 (industry analysts); delays would postpone Mexico’s ability to take advantage of lower-priced U.S. gas.
- Institutional and infrastructure reforms:
  - Reforms change pipeline governance: an independent regulatory body, Cenegas, will manage natural gas traffic to make allocations more efficient.
  - Transport and storage markets open to private participants.
  - SENER plans to build out domestic pipeline infrastructure to connect to U.S. pipelines and expand transport within Mexico.
  - Reforms aim to prevent strains experienced in 2013 (critical alerts) that affected industrial users and electricity generators.

### Electricity sector implications and reforms
- Natural gas and electricity prices:
  - Lower natural gas prices will help lower electricity prices, improving firms’ cost structure.
  - Largest natural gas users are electricity generators; natural gas combined cycle plants have been main choice for new generation since the 1990s.
  - Private self-generators projected by SENER to have fastest growth in natural gas demand over the next decade, displacing diesel and fuel oil.
- Market liberalization and grid operations:
  - Energy reforms further open generation market to competition.
  - Previously IPPs had to sell electricity to CFE and self-generators could not sell excess power to the market; reforms change this to allow more market participation.
  - Authorities are considering expanding the scope of qualified users who can buy electricity in the wholesale market.
  - Grid improvements and operations:
    - Just under half of transmission lines more than 30 years old; less than 10 percent added in the last five years.
    - CFE plans to expand system by about 1 percent per year over the next decade.
    - Prior management of grid traffic by an entity within CFE raised conflict-of-interest concerns; Cenace will be transformed into an independent system operator to dispatch based on efficiency or emergency needs.
    - CRE will set tariffs on transmission, distribution, and ancillary services and can provide incentives to reduce transmission losses.
- CFE institutional changes:
  - Conversion of CFE into a state productive enterprise will give it more operational and budgetary autonomy and expanded ability to contract with third parties for transmission and distribution services.
  - CFE tasked with adopting international standards for management of state enterprises to improve efficiency and lower costs.

### Key conclusions and policy implications
- Energy reform potential:
  - Reform is comprehensive with potential to support faster growth, better living standards, and greater energy security.
  - Short-run objective: overcome risk of falling hydrocarbons production and improve fiscal outlook.
  - Medium- to long-run objective: enable exploitation of shale and deepwater potential and reduce domestic energy costs and improve services.
- Short-run priorities:
  - Improvements in recovery and development of existing fields are crucial; Round 1 focuses on auctioning 2P fields that could yield hydrocarbons quickly.
  - Pemex has more freedom to partner with third parties to increase investment and import technologies.
- Implementation risks:
  - Legislative hurdles have been tackled, but implementation risks remain; Round 1 is critical and many regulatory and institutional changes are still required.
  - Delays or problems that dampen investor confidence could reduce exports and fiscal revenues (stylized downside scenarios illustrate lower production and lower investment and FDI).
- Managing expectations and infrastructure focus:
  - Patience is needed for shale and deep water given long lead times before meaningful production.
  - Pipelines and the electricity grid are essential for growth; planned natural gas pipeline projects will lower dependence on LNG, diesel, and fuel oil.
  - Independent system operators for natural gas and electricity can reduce critical alerts and enhance service delivery.
- Institutional reforms for state enterprises:
  - Transforming Pemex and CFE into productive state enterprises and reducing their fiscal burdens are first steps toward improving efficiency, costs, and ability to invest in infrastructure.

*Source: IMF staff report text in content unit _cr14320 - 14.      Between 2015 and 2019, additions to proven reserves will likely come from the*

### 2.      Nevertheless, the cost and quality of electricity, and more recently temporary

### _cr14320 - 2.      Nevertheless, the cost and quality of electricity, and more recently temporary

### Background and context
- Mexico ranks consistently low on global competitiveness indicators regarding the costs and quality of electricity.
- Limited investment in infrastructure has led natural gas pipelines to operate close to maximum capacity and unable to accommodate rising demand, including from the manufacturing sector (Bank of Mexico, 2013).
- The recent energy reform opens the hydrocarbons and electricity sectors to private participation in different modalities, aiming to attract investment, improve infrastructure and efficiency, and increase the availability of natural gas—changes that are expected to lower electricity prices and boost Mexican manufacturing.

### Stylized facts on manufacturing since NAFTA
- Car production in Mexico tripled since NAFTA; Mexican cars account for about a third of U.S. auto imports and Mexico supplies one third of all U.S. imports of auto-parts.
- Manufacturing growth was uneven across sectors and regions; transportation equipment was an outlier with meaningful contributions from total factor productivity (almost a fifth of overall growth).
- The rise of China after WTO accession in 2001 crowded out some Mexican exports in U.S. markets; proximity to the U.S. market regained importance around 2005 as oil prices rose.

### Estimated sensitivity of manufacturing output to energy prices (elasticities)
- Regression framework: panel regression of real manufacturing GDP by industry i, state j, year t; RHS includes lagged dependent variable, lagged changes in electricity (EL), natural gas (NG), and oil derivatives (OD) prices (deflated by CPI), plus controls (unit labor costs, real effective exchange rate, cost of capital, U.S. industrial production, etc.) and fixed effects at the sector-state pair level.
- Key elasticity estimates:
  - Electricity price elasticities range from -0.11 to -0.28 (up to -0.28).
  - Natural gas elasticities range from -0.04 to -0.07.
  - Oil derivatives coefficient is positive in some specifications but weakens with additional controls; statistical significance declines in richer specifications.
- Subsector heterogeneity:
  - Metals and transportation equipment (including autos) show a highly statistically significant electricity-price elasticity of -0.51.
  - Chemical sector elasticity about -0.1 and not statistically significant.
- Additional empirical notes:
  - About 18 percent of national natural gas demand comes from the industrial sector, including manufacturing.
  - LPG represents less than 3 percent of energy inputs in the industrial sector, and the estimated LPG elasticity (~ -0.09) should be interpreted with caution as it may capture demand effects.

### Potential for reducing electricity generation costs and price outcomes
- Fuel substitution potential:
  - Mexico relies more on oil derivatives (mostly fuel oil) for generation relative to the United States, which relies more on coal.
  - Under the current Mexican pricing mechanism, an immediate substitution of fuel oil for natural gas—given fuel oil represents 18.1 percent of total generation and natural gas is about 71 percent cheaper than fuel oil—would imply, ceteris paribus, a reduction of about 13 percent in electricity prices (0.181 * -0.71).
  - Given existing plans for pipeline investments and conversion of fuel oil plants to operate with gas, these 13 percent reductions seem feasible within five years.
- Competition and efficiency gains:
  - Attracting private investment and enabling private operation/maintenance of transmission and distribution lines (ownership remaining public) can yield efficiency gains.
  - Assuming convergence to U.S. prices as a limiting case, these combined gains could lead to a reduction of about 50 percent in commercial and industrial electricity prices in total.
- Natural gas supply and infrastructure:
  - Increasing supply of natural gas for electricity generation will come first from additional imports from the United States.
  - Shortfalls of natural gas (caused by lack of spare pipeline capacity) affected manufacturing; Bank of Mexico estimates that without these shortfalls GDP growth in 2013Q2 in annual terms would have been -0.4 and not -0.7 percent.
  - Notable pipeline project details cited: a large pipeline project of 1,200 Km from Texas to Guanajuato with capacity to transport 2,100 million of cubic feet per day by 2016 (Los Ramones II-Norte is part of this project), and plans to convert 7 fuel oil-based electricity generation plants to operate with natural gas by 2016.

### Quantified real effects on manufacturing and GDP under scenarios
- Elasticity-derived impacts (using estimated elasticities and two scenarios for electricity price reductions):

  Scenario 1 — Substitution of fuel oil for natural gas (assumes a 13 percent reduction in electricity prices):
  - Increase in manufacturing output: 1.4 percent to 3.6 percent.
  - Increase in overall GDP: 0.2 percent to 0.6 percent.
  - Timing: gains under scenario 1 are reasonable to expect to materialize over 2016–2019.

  Scenario 2 — Convergence to U.S. levels (assumes convergence of industrial/commercial electricity prices to U.S. levels):
  - Increase in manufacturing output: 5.5 percent to 14.0 percent.
  - Increase in overall GDP: 0.9 percent to 2.2 percent.
  - Timing: convergence timing is more uncertain because it requires reducing energy losses in transmission and distribution.

- Interpretation:
  - Manufacturing output could increase up to 14 percent if prices converged to U.S. levels.
  - For the economy as a whole, the manufacturing impulses imply increases in real GDP of up to 0.6 percent (scenario 1) and up to 2.2 percent (scenario 2), given today’s share of manufacturing in the economy.
  - These level effects would materialize over the horizon required for electricity prices to achieve the stated reductions.

### Indirect effects and spillovers
- Methodology: panel vector autoregression (VAR) to allow endogenous responses (e.g., unit labor costs) and to assess spillovers across subsectors and regions.
- Potential indirect channels:
  - Unit labor costs may react to energy-price changes (potential substitution effect increasing labor demand or labor-supply channels tied to cost-of-living changes).
  - Subsector interdependencies: one subsector may be affected directly by energy prices and indirectly via demand from other subsectors.
  - Regional spillovers: changes in electricity prices may propagate across regional manufacturing clusters.
- Focus: electricity prices are the primary energy input with the largest direct and potential indirect impacts on manufacturing output.

### Concluding implications for policy and reform
- The energy reform can alleviate bottlenecks by attracting investment to expand pipeline capacity and lower generation costs, thereby improving competitiveness of manufacturing through lower energy input costs.
- Short-run and medium-run gains are plausible: a 13 percent electricity price reduction seems feasible within five years given pipeline investments and plant conversions already planned; additional efficiency and competition gains under reform could yield larger, longer-run price declines (benchmark up to 50 percent relative to U.S. prices).
- Lower electricity prices are likely to yield heterogeneous gains across manufacturing subsectors and regions, with metals and transportation equipment likely to benefit most.
- Monitoring and policies to ensure open access to pipeline capacity and to reduce distribution/transmission losses will be important to realize and sustain these gains.

*Prepared by IMF staff; source content: IMF staff report section on Mexico’s energy reform and manufacturing, including empirical estimates and scenario analysis.*

### Appendix A provides technical details about the VAR specification. A more detailed technical discussion and several

### Appendix A provides technical details about the VAR specification.

### VAR specification and identification
- Panel VAR includes variables: manufacturing output, real interest rates, U.S. industrial production, unit labor costs, and electricity prices.
- Manufacturing output is measured at each state (i), subsector (j), and year (t).
- Real interest rates correspond to the ex-post real yield on Moody’s U.S. corporate bonds.
- Identification is achieved through Choleski ordering with baseline specification including fixed effects at the state-subsector level.
- Impulse response functions are reported for a shock of 1% to electricity prices and are calculated for a period of 5 years.
- 90% confidence interval bands are based on parametric bootstrapped standard errors.
- Variants estimated:
  - Subsector spillovers: 13x1 vector containing manufacturing output for each subsector plus output growth in non-manufacturing sectors; total manufacturing unit labor costs at national level; state fixed effects.
  - Regional spillovers: 5x1 vector containing manufacturing output for each economic region of Mexico; fixed effects at subsector level.
- In all models, the real interest rate and U.S. industrial production are assumed to be exogenous from other variables.
- Impulse response functions without spillovers are calculated by restricting coefficients of output in each sector on output in other sectors to zero.

### Main empirical impulse-response findings
- Shock specification and confidence:
  - Shock: 1% to electricity prices.
  - Horizon: 5 years.
  - Confidence bands: 90% (parametric bootstrap).
- Amplification via unit labor costs:
  - Allowing unit labor costs to respond endogenously to changes in electricity prices amplifies the response of output.
  - Direct (impact) effect measured in period 2; indirect effect operates through unit labor costs.
  - By the fifth year, the cumulative effect of an electricity price shock on manufacturing output is about 50 percent larger than the impact effect.
- Spillovers:
  - Under the employed methodology, no statistically significant evidence of regional and sectoral direct spillovers is found.
  - Appendix Figure 1 compares impulse responses with (dotted purple) and without (solid blue) direct spillovers; differences remain within the 90-percent confidence bands (dashed red), implying non-significance.
  - Similar conclusion for regional spillovers.
- Possible reasons for weak statistical evidence of direct spillovers:
  - Substantial cross-sectional and time variation in spillover coefficients.
  - Some interdependence may occur simultaneously, so the impact effect already captures amplification from spillovers.
  - Spillovers through direct interdependence may appear weak given uneven historical performance across manufacturing subsectors and regions.
- Model specifics and robustness:
  - Impulse response functions are reported along with bootstrapped confidence intervals.
  - Appendix Figures 1 and 2 present subsector and regional impulse response functions respectively; all variables are expressed as yearly differences.

### Energy consumption and regression evidence (selected statistics)
- Table 1. Energy Consumption (in Petajoules) of the Industrial Sector:
  - Industrial sector total: 1,381.1 (2010), 1,492.3 (2011), 1,530.6 (2012).
  - Shares (2010, 2011, 2012):
    - Electricity: 34.2% ; 33.6% ; 34.5%
    - Natural Gas: 35.3% ; 35.2% ; 35.8%
    - Oil derivatives: 17.0% ; 15.5% ; 15.6%
    - Others: 13.5% ; 15.8% ; 14.0%
  - Auto industry energy use (Petajoules): 10.5 (2010), 12.7 (2011), 14.4 (2012).
    - Auto industry shares (2010, 2011, 2012):
      - Electricity: 68.3% ; 60.0% ; 60.4%
      - Natural Gas: 19.9% ; 28.1% ; 28.7%
      - Oil derivatives: 11.8% ; 11.9% ; 10.9%
      - Others: 0.0% ; 0.0% ; 0.0%

- Table 2. Estimates of Elasticities of Manufacturing Output to Energy Prices (sample: 1998–2012, annual; N 3,584)
  - Reported example coefficients (selected, preserve exact values as in table):
    - Coefficient examples across columns: 0.036; 0.029; 0.030; 0.031; 0.005; 0.005; 0.005 (with parenthesis standard errors (0.040) (0.040) (0.041) (0.041) (0.041) (0.041) (0.041))
    - Negative coefficients reported: -0.114 -0.141 -0.141 -0.167 -0.263 -0.269 -0.283 (standard errors (0.032)*** etc.)
    - Other coefficients: -0.068 -0.049 -0.049 -0.045 -0.048 -0.047 -0.039 (standard errors (0.014)*** etc.)
    - ULC and other controls: 0.111 0.094 0.093 0.074 0.054 0.048 0.019 (with standard errors including (0.019)*** etc.)
    - R2 values: 0.03 ; 0.03 ; 0.03 ; 0.03 ; 0.06 ; 0.06 ; 0.06
    - Constant terms: 0.988 1.018 1.024 1.045 0.967 1.089 0.956 (standard errors (0.124)*** etc.)
  - Note: EL, NG, and OD correspond to the domestic prices of electricity (average of tariffs for medium and large firms), natural gas, and oil derivatives respectively, deflated by the CPI. The operator ∆ corresponds to the percent change. ULC denotes unit labor costs. WTI denotes West Texas Intermediate oil prices. U.S. r denotes ex-post real interest rates measured as the difference between the yield on BAA Moody’s-rated corporate bonds (all industries) and core inflation. * p<0.1; ** p<0.05; *** p<0.01.

- Table 3. Estimates When Energy Inputs Enter Separately (sample: 1996–2012, annual; N 4,352 or 3,584 when OD included)
  - Example coefficients (selected):
    - 0.035 0.033 0.009 0.034 (standard errors (0.037) (0.038) (0.041) (0.038))
    - -0.174 (0.028)***
    - -0.024 (0.008)***
    - 0.038 (0.022)*
    - -0.091 (0.022)***
    - R2: 0.08 ; 0.07 ; 0.03 ; 0.07
    - Constants: 1.543 1.292 1.108 1.390 (standard errors (0.114)*** (0.127)*** (0.090)*** (0.115)***)
  - Note: EL, NG, OD, and LNG correspond to domestic prices of electricity, natural gas, oil derivatives, and liquefied natural gas respectively. Sample period is 1996–2012 except when OD included (1998–2012).

- Table 4. Differential Effects Across Subsectors (sample period: 1996–2012; N 3,584)
  - Selected coefficients and statistics (preserve exact values):
    - Coefficients: 0.008 0.011 0.011 (std. errors (0.040) (0.040) (0.040))
    - -0.282 -0.529 -0.514 (std. (0.041)*** (0.075)*** (0.080)***)
    - -0.116 -0.036 -0.044 (std. (0.031)*** (0.014)** (0.031))
    - Other subsector-specific coefficients appear (e.g., *Food, beverages, and tobacco  0.124  0.029  (0.029)***  (0.031))
    - R2: 0.07 ; 0.07 ; 0.07
    - Constants: 0.956 0.957 0.958 (std. (0.130)*** (0.130)*** (0.130)***)

### Concluding remarks and policy implications
- Main conclusion:
  - The energy reform is likely to have important real effects through its impact on manufacturing activity via lower electricity prices, on top of direct growth effects from increased investment and production in the energy sector.
  - Technology spillovers and increased foreign direct investment in manufacturing as the sector becomes more competitive could amplify these estimated effects.
- Policy priorities and recommendations:
  - Increasing gas pipeline capacity to allow larger natural gas imports from the U.S. will yield the most immediate gains.
  - Existing fuel oil-based plants should be adapted to operate with natural gas to realize substitution gains.
  - As reform attracts private investment in low-cost electricity generation, gains from fuel substitution are more likely to become permanent.
  - Further gains will follow from increasing availability of gas throughout the country by an expanded network of pipelines.
  - A strong regulator is critical to ensuring competition in electricity generation and thus making any reduction in electricity prices long-lasting; synergy with the antitrust reform will play an important role.
  - For efficiency gains, operation of transmission and distribution lines must encompass the right incentives to improve infrastructure; distribution and transmission lines will remain property of the state, so efficiency gains depend on incentives for the new administrator to invest and lower high technical losses.
  - Caution: Reducing distribution losses will be challenging and thus the gains from increased efficiency, while in theory important, are not guaranteed.

*Source: Appendix and accompanying tables and figures from the provided IMF content unit.*

### 1.      Gross capital inflows (particularly portfolio inflows) to Mexico have increased

### _cr14320 - 1.      Gross capital inflows (particularly portfolio inflows) to Mexico have increased

### Overview: recent capital inflows and investor links
- Gross capital inflows (particularly portfolio inflows) to Mexico have increased significantly in recent years.
- Mexico has particularly strong links to the United States, its largest trading partner and the main source of portfolio and foreign direct investment.
- Mexico’s inclusion in Citigroup’s World Government Bond Index (WGBI) in 2010 attracted new groups of foreign investors.
- The share of non-resident holdings of domestic sovereign debt reached 36 percent in April 2014.
- Pension, insurance, and mutual funds now account for about half of the financial system (more than 40 percent of GDP).

### Volatility and risk of portfolio inflows
- Portfolio capital inflows have become more volatile, with episodes described:
  - Surge prior to the global financial crisis (GFC).
  - Sharp contraction during the GFC (2008–09).
  - Rebound to a higher level in 2010.
  - Fall during the European crisis in 2011.
  - Surge in 2012.
  - Sharp contraction after the U.S. Federal Reserve tapering announcement in May 2013.
- Excessive volatility in capital flows could:
  - Amplify economic cycles.
  - Increase financial system vulnerabilities.
  - Aggravate overall macroeconomic instability.
- The paper does not assess causation but notes Mexico remains vulnerable to risks from U.S. monetary policy normalization and geopolitical events.

### Domestic investor base developments
- Domestic investors’ flows have become more important as the domestic investor base expands and diversifies.
- Over the last 10 years pension funds’ assets increased by about 18 percent annually.
- Gradual regulatory changes have allowed pension funds to diversify and invest abroad.
- Domestic investors increased holdings of foreign assets more than domestic government debt holdings, and reversed positions during crisis (sold foreign assets and repatriated funds).

### Research questions and data
- Central questions addressed:
  - Have foreign investors behaved differently from domestic investors, particularly during market stress?
  - Have foreign investors been more destabilizing?
  - Do domestic investors have a mitigating role to play?
- Data and scope:
  - Three unique datasets on Mexico (including extensive mutual fund datasets).
  - Mutual fund datasets: about 400 international mutual funds and 540 Mexican mutual funds at monthly frequency, latest observation April 2014.
  - Sample periods:
    - Full fund-level sample: January 2007 to April 2014 (covers GFC and tapering announcement in May 2013).
    - Shorter sample for momentum trading comparisons (domestic funds available from January 2011 to April 2014).
  - For panel regression: 546 domestic mutual funds and 375 foreign mutual funds (foreign sample includes 112 bond funds and 263 equity funds).

### Methodology for identifying extreme gross flow episodes
- Focus on gross inflows and outflows to differentiate foreign-initiated vs. domestic-initiated movements.
- Four types of extreme portfolio capital flow movements (1995–2013, quarterly), following Forbes and Warnock (2012):
  - Surge: sharp increase in nonresidents’ gross capital inflows.
  - Stop: sharp decrease in nonresidents’ gross capital inflows.
  - Flight: sharp increase in residents’ gross capital outflows.
  - Retrenchment: sharp decrease in residents’ gross capital outflows.
- Forbes and Warnock (2012) criteria (as applied):
  - (1) Year-over-year changes in four-quarter gross flows more than two standard deviations above or below historic average for at least one quarter of the episode;
  - (2) Episode lasts for consecutive quarters with year-over-year changes more than one standard deviation above or below historical average;
  - (3) Episode length greater than one quarter.

### Empirical findings — mutual fund behavior: herding
- Two co-movement measures:
  - Proportion of all funds active in Mexico that are net sellers in a month.
  - Herding index based on Lakonishok, Shleifer, and Vishny (1992), adjusted using the share of funds that are net buyers across all emerging markets and allowing time-variation.
- Key empirical observations:
  - Foreign mutual funds (bond and equity) showed a strong tendency to sell Mexican assets during periods of heightened global uncertainty.
    - After Lehman Brothers collapsed in September 2008, around 75 and 95 percent of equity and bond funds active in Mexico were net sellers, respectively.
  - Foreign mutual funds are more inclined to exhibit herding than domestic mutual funds.
    - During the tapering announcement in May 2013:
      - Around 50 percent of domestic mutual funds were selling Mexican assets.
      - Net sellers among foreign mutual funds rose to above 70 percent.
  - Herding indices for foreign bond and equity funds increased significantly around the tapering announcement in May 2013 relative to prior periods.

### Empirical findings — mutual fund behavior: positive-feedback (momentum) trading
- Estimation setup:
  - Regress change in a fund’s current asset position on lagged returns (one-month lag), stress dummy, and fund fixed effects.
  - Returns used:
    - Peso-denominated return on Mexico’s 3-month government bonds (for bond funds) or stock market return (for equity funds).
  - Stress dummy equals one for Global Financial Crisis and tapering announcement periods, zero otherwise.
  - Robustness checks include using Mexico’s 10-year government bond returns and U.S. dollar–denominated returns.
- Main results:
  - Strong evidence of positive-feedback trading among foreign bond funds during market stress episodes.
  - Domestic bond funds do not show statistically significant positive-feedback trading.
  - Foreign bond funds’ positive-feedback trading tended to be stronger during market stress than during normal times.
  - Domestic equity funds exhibit negative-feedback trading (buy when returns fall, sell when returns rise).
  - Foreign equity funds do not show significant positive-feedback trading in normal times, but do exhibit positive-feedback trading during crisis times.
  - The stress dummy coefficient is negative, reflecting funds’ overall tendency to reduce exposures to Mexico during stress.

### Interpretation and implications
- Foreign mutual funds contributed to synchronized selling and momentum-driven outflows during stress episodes, increasing gross flow volatility.
- Domestic investors have at times offset foreign behavior (e.g., prior to Lehman domestic flights offset foreign surges), but during recent episodes (May 2013) domestic offsetting was only partial.
- Policy-relevant considerations include understanding investor-type behavior for flow volatility management and considering measures to enhance market transparency and domestic investor capacity to mitigate external shocks.

*Prepared by Jianping Zhou, Han Fei, and Jasmine Xiao (IMF).*

### 17.      These micro-level findings provide one plausible explanation for what we observe

### _cr14320 - 17.      These micro-level findings provide one plausible explanation for what we observe

### Micro-level evidence on portfolio flows and investor behavior
- Between Q2 and Q1 2013, capital inflows (by non-residents) fell by US$24.5 billion, of which US$14 billion was due to a sudden stop in portfolio inflows.
- Much of the decline reflected a sharp reduction in foreign mutual funds’ investment in Mexico, especially by small retail funds.
- Foreign mutual fund investors exhibited potentially destabilizing trading behaviors:
  - Evidence of herding behavior (sell when others sell).
  - Positive-feedback trading (buy high and sell low).
- No significant evidence found that domestic mutual funds behaved similarly.

### Trends in foreign participation in Mexican sovereign debt
- The share of non-resident holdings of domestic sovereign debt in Mexico rose rapidly, especially relative to residents’ holdings (sample period 2000–14).
- Mexico attracted large U.S. investor participation and growing holdings from Europe and Japan since 2007.
- Inclusion in Citibank’s World Government Bond Index (WGIB) in October 2010 increased foreign participation.
- Mexican peso noted as one of the most traded EM currencies globally, facilitating hedging of currency risk for foreign investors.

### Empirical strategy: OLS and multivariate GARCH analyses
- Two time-series models estimated: OLS and multivariate GARCH (MGARCH).
- High-frequency samples at weekly and daily frequencies covering 2000–14.
- Dependent variable: volatility of Mexican 10-year local-currency sovereign bond yields (two measures):
  - Within-week volatility: standard deviation of daily yields within each week normalized by the weekly mean.
  - 5-day rolling volatility: standard deviation of a 5-day rolling window of daily yields, normalized by the mean.
- Key explanatory variables:
  - Foreign and domestic participation: daily holdings of LT Mexican local-currency sovereign bonds by foreigners and five domestic investor types (pension funds, insurance companies, mutual funds, banks, other domestic investors) as shares of total outstanding LT sovereign bonds.
  - Global financial shocks: VIX and volatility in U.S. 10-year Treasury yields.
  - Domestic controls: official international reserves (weekly, excluding gold), peso/US dollar exchange rate (daily), interbank interest rate (daily). The 28 day TIIE interbank equilibrium rate used; first-order differences of reserves and interest rates and first-order differences of log exchange rates used in regressions.
- MGARCH specification:
  - Jointly model daily changes in Mexican 10-year sovereign bond yields and Mexican stock market returns.
  - Mean equations include asset returns and global shocks Z (VIX and change in U.S. 10-year yields) and X (first-order log difference of exchange rate).
  - Volatility equations assume diagonal conditional variance matrix with GARCH(1,1) processes and exogenous regressors; constant conditional correlation between returns assumed.

### Key empirical findings
- Foreign participation amplifies the impact of global financial shocks on Mexican sovereign bond yield volatility, notably during market stress periods.
  - Consistent across OLS and MGARCH specifications and robust to various model choices.
  - MGARCH results: higher foreign participation increases sensitivity of the volatility of the 10-year Mbono yield to a shock to the VIX, without necessarily affecting the level of the Mbono yield.
  - Increased foreign participation makes both the level and volatility of the Mbono yield more sensitive to shocks to the yield on 10-year US Treasuries.
- Domestic investors’ roles were mixed and depended on investor type and shock nature:
  - Banks tend to amplify the impact of global shocks on domestic yield volatility during normal times; their role during stress periods is unclear.
  - OLS results suggest pension funds and banks dampen the impact of VIX shocks during stress periods, but:
    - No significant and robust evidence that insurance sector or mutual funds mitigate or amplify during stress periods.
    - The mitigating role of pension funds disappears if the shock is to the U.S. 10-year Treasury yields.
  - Overall, domestic investors are not uniformly different from each other; behavior depends on the nature of global shocks.
- Robustness:
  - Replacing sovereign-yield volatility with exchange-rate-based volatility in OLS regressions does not change the qualitative result that foreign participation tends to amplify global financial shocks.
  - Replacing investor-share interaction terms with first-order differences of shares also leaves main results intact.

### Selected quantitative details from fund-flow regressions and samples
- Table 1a (Bond Funds): Sample size: 2011 January – 2014 March (stress episode: tapering announcement).
  - Domestic Funds (Fund Flow): Lagged Mex 3-month bond yield coefficients: 0.00726 and 0.00628; R-squared: 0.019; Number of Observations: 20202.
  - Foreign Funds (Fund Flow): Lagged Mex 3-month bond yield coefficients: 0.00725*** and 0.00660***; Lagged bond yield * Stress: 0.0171***; Stress coefficient: -0.0677***; Constant: -0.0288*** and -0.0260***; Number of Observations: 3422; R-squared: 0.172 and 0.178.
  - Standard errors (robust) reported; significance levels: *p<0.10 **p<0.50 ***p<0.01.
- Table 1b (Equity Funds): Lagged stock market return coefficients:
  - Domestic Funds: -0.0211* and -0.0269**.
  - Foreign Funds: 0.00228** and 0.000615.
- Time period for main econometric sample: 2000–14 (daily and weekly frequencies).
- International reserves noted: International reserves stand at about US$190bn as of June 2014.
- Example macro movement tied to tapering announcements:
  - Capital inflows by non-residents dropped by US$24.5 billion from Q1 to Q2 2013.
  - Exchange rate depreciated by 8 percent.
- FCL/Reserves metric example: Mexico: 38.1% plus 15.4% of FCL (presented in chart context).

### Policy implications and concluding remarks
- Mexico remains vulnerable to global financial shocks due to highly integrated financial markets and larger, more volatile gross capital inflows (particularly portfolio inflows).
- A resurgence of investor uncertainty from an earlier or sharper-than-expected rise in U.S. interest rates poses a near-term risk; such shocks can trigger significant portfolio outflows and high bond-price volatility.
- Policy credibility is key to maintaining market confidence during stress. Mexico’s strengths cited:
  - Sound public finances, low and stable inflation, a sustainable external sector position, a healthy banking system.
  - Strong policy framework: monetary policy guided by an inflation targeting framework under a flexible exchange rate; fiscal policy anchored by a fiscal responsibility law.
  - One of the highest credit ratings among emerging markets.
- Country insurance and buffers recommended:
  - Exchange rate flexibility acted as a shock absorber; high reserves provided a useful buffer.
  - FCL arrangement treated as precautionary and complemented reserve buffers, providing extra protection against tail risks.
- Importance of developing a deep and diverse domestic investor base:
  - Pension reforms and establishment of insurance and mutual fund industries have expanded local investor base.
  - Domestic investors can play a stabilizing role (e.g., improving liquidity in government securities secondary markets).

*Source: IMF chapter/section text (pages excerpted).*

### Box 2. Data on Foreign Mutual Funds

### Box 2. Data on Foreign Mutual Funds

### Data source and coverage
- Data source: EPFR Global.
- Coverage: about 11,000 equity funds and about 4,500 bond funds, all of which have $22 trillion in total assets as of the end of 2013.
- EPFR Global reports it tracks more than 95 percent of EM-focused bond and equity funds.
- EPFR data are recorded on a nationality basis (contrast: Balance of Payments data report on a residency basis).

### Advantages and limitations of EPFR Global
- Advantages:
  - High-frequency (weekly or monthly), detailed information at the fund-level.
  - Fund-level detail on domicile and geographic focus.
  - Records data on a nationality basis.
- Limitations:
  - EPFR generally tracks only mutual funds.
  - Country-level data are estimated by combining two fund-level datasets (see below); the estimation method multiplies the country allocation weight at the end of month by the aggregate flow into funds with specific characteristics and therefore:
    - Generally captures changes in flows from ultimate investors but not changes in allocation weights.
    - If asset managers shift allocations (e.g., from EM economies to advanced economies or cash) in response to deteriorating global financial conditions, estimated outflows from EM economies can be underestimated.
- Relevance note from source: IMF (2014) reports that mutual funds are more sensitive to global financial conditions and are more likely to engage in return chasing than other types of investors.

### EPFR data types and estimation method
- Two fund-level datasets used:
  - Fund flow data set: reports dollar-denominated flows, returns, assets under management (AUM), and fund characteristics (domicile, geographic focus). Flows, returns, and AUM are not disaggregated by destination economy.
  - Country allocation data set: reports country allocation weights over more than 130 developed and emerging economies on a monthly basis.
- Country-level data: estimated using the two fund-level datasets by multiplying month-end country allocation weights by aggregate flows into funds with specific characteristics; country-level data can be decomposed by fund characteristics.

### Empirical evidence (selected quantitative results from regressions and robustness checks)
- General sample and stress definitions:
  - Sample size for many fund-level regressions: 2011 January – 2014 March (stress episode: tapering announcement).
  - Longer-sample robustness checks: 2005 January – 2014 March (stress episodes: global financial crisis & tapering announcement).

- Table 1a. Bond Funds: Foreign funds (selected coefficients)
  - Lagged Mex 3-month bond yield: 0.00725*** (0.000765)
  - Lagged bond yield * Stress: 0.0171*** (0.00567)
  - Stress: -0.0677*** (0.0229)
  - Number of Observations: 3422
  - R-squared: 0.178

- Table 1b. Equity Funds: Foreign funds (selected coefficients)
  - Lagged stock market return: 0.00228** (0.00113)
  - Lagged stock market return * Stress: 0.0138** (0.00657)
  - Stress: -0.0000949 (0.000118)
  - Number of Observations: 18595
  - R-squared: 0.140

- Table 2a. Robustness Check (1) — Longer sample for foreign mutual funds (selected coefficients)
  - Foreign Bond Funds: Lagged return 0.000557*** (0.000204)
    - Number of Observations: 7287
    - R-squared: 0.220
  - Foreign Equity Funds: Lagged return 0.0172*** (0.00604)
    - Number of Observations: 38724
    - R-squared: 0.075
  - Sample size: 2005 January – 2014 March

- Table 2b. Robustness Check (2) — Using dollar-denominated return on the 3-month Government Bonds (selected coefficients)
  - Foreign Funds: Lagged dollar return on Mex 3-month govt bonds 0.0705*** (0.00803)
  - Lagged bond yield * Stress (foreign funds): 0.261*** (0.0701)
  - Stress (foreign funds): -0.0852*** (0.0228)
  - Number of Observations: 3422
  - R-squared: 0.176
  - Sample size: 2011 January – 2014 March (stress episode: tapering announcement)

- Selected multivariate results (VIX and U.S. tapering shocks; t-statistics in parentheses; significance indicators preserved in source)
  - Table 3a (VIX shock, OLS):
    - VIX*Foreign share*Stress: 0.0059*** (6.56) and 0.0053*** (7.13) in alternative specifications.
    - VIX*Pension share*Stress: -0.0073* (1.89) and -0.0071* (1.83).
    - Adjusted R2: 0.33 and 0.32 in two specifications.
  - Table 3b (U.S. tapering shock, OLS):
    - Volatility of U.S. 10-year bond yields*Foreign share*Stress: 0.043*** (4.78) and 0.040* (5.57).
    - Adjusted R2: 0.25 and 0.22.
  - Table 4a and 4b (Multivariate GARCH results):
    - VIX*Foreign share*Stress: column IV-A 0.0079*** (9.13).
    - U.S. 10-year bond yields*Foreign share*Stress: column IV-B 3.73*** (5.58).
    - Exchange rate depreciation appears with large positive t-statistics in several GARCH specifications (examples: 2.87*** (16.84); 3.44*** (18.95)).
    - ARCH and GARCH terms reported (examples: ARCH term 0.25*** (8.34); GARCH term 0.30*** (6.05); other specifications show ARCH term 0.26*** (7.83) and GARCH term 0.56*** (10.95)).

### Implications highlighted in the source
- EPFR’s fund-level, high-frequency data facilitate analysis of mutual fund behavior and its sensitivity to global financial conditions.
- The EPFR estimation approach for country-level flows may understate outflows from EM economies when asset managers change allocation weights (e.g., shift from EM to advanced economies or cash) during stress episodes.

*Source: Box 2. Data on Foreign Mutual Funds (extracted from the provided IMF content)*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2014/_cr14320.pdf_