Out of Gas: An Empirical Analysis of the Fiscal Regime for Exploration in India

Out of Gas: An Empirical Analysis of the Fiscal Regime for Exploration in India Anupama Sen Senior Research Fellow, Oxford Institute for Energy Studies, United Kingdom JUNE 2014

Outline 1. Natural Gas in India 2. Conceptual Literature on Resource Taxation 3. Empirical Literature 4. Method 5. Analysis 6. Results 7. Observations & Conclusions

1. Natural Gas in India Gas forms 11% of India s primary energy consumption Relatively young but growing sector India s gas reserves 0.7% of world reserves Range of estimates Mostly offshore Consumption 60 Bcma (20-30% LNG) 70% used in power generation, fertiliser industry, household use, transportation Policy on gas Substitute for coal, naphtha, fuel oil Security of supply / mitigating energy shortages Environment

Potential for gas in electricity and transportation Per Capita Emissions by Sector, 2010 (kg of CO 2 per capita) Electricity Other Energy Industry Own Use Manufacturing Industries and Construction Transport Residential Other Total CO 2 Emissions from Fuel Combustion OECD 4,007 558 1,423 2,699 797 1,408 10,096 UK 2,873 519 822 1,919 1,325 1,643 7,776 USA 7,448 845 1,893 5,229 1,038 1,897 17,312 India 748 52 342 138 64 108 1,388 China 2,659 205 1,734 382 225 416 5,395 Brazil 230 129 585 852 87 194 1,989 Africa 414 39 138 215 56 104 910 Middle East 2,715 786 1,577 1,651 623 829 7,559 Source: IEA (2008)

Pricing Prices for domestically produced gas are controlled To keep prices to consuming sectors low Prices are set according to the fiscal regime governing a producing field The fiscal regime has evolved over time; thus multiple pricing mechanisms India s 1999 liberalised upstream fiscal regime Prices based on an S curve formula with a cap and ceiling SP = $2.5 + (CP 25) 0.15 + C Five year reviews Formula rendered redundant by high oil price Ceiling of $60/bbl breached very early on Review of pricing currently underway; proposals to link it to Henry Hub, NBP and Japan LNG

India Gas Price vs International Benchmarks India LNG Imports Henry Hub Source: BP Statistical Review; Govt of India

Fiscal Regime for Gas Liberalised regime launched in 1999 New Exploration Licensing Policy Based on a Production Sharing Contract (PSC) Hybrid system (royalty + taxes on rent ) Main features 1. Royalty 10%, 5% for first 7 years 2. Cost recovery Opex, capex & royalty up to 100% pa 3. Profit Sharing Based on a R-factor, or the ratio of cumulative income to cumulative capital expenditure. Two tranches, 1.5 and 3.5. Shares in-between worked out using an extrapolation formula Z = a + [(b-a)*(x-1.5)/2] 4. Corporate Income Tax 30-40% 5. Minimum Alternate Tax 18.5% of book profits if IT payable is < 18.5% of book profits

Graphical Illustration of PSC Regime Royalty Gross Production (-) (-) Cost Recovery Total Profit Oil Government Share Contractor Share (+) Taxable Income (+) Government Take (+) Tax (-) (+) Contractor Take

Performance of India s Fiscal Regime Production Large initial increase (2009) followed by dramatic decline in production; currently a third of original targets Driven by private sector NOC production has remained stagnant No significant new discoveries since 2004 Investments Down from $4.7Bn in 2007 to $1.6 Bn in 2013 Arbitration & delays Companies accused of gold plating in the R Factor Alleged loss of revenues to exchequer Proposed Policy Solution Replace the PSC regime with a simple Revenue Sharing Contract (RSC) Regime Eliminate royalty, cost recovery and profit sharing (R Factor) Companies simply share a percentage of their revenues from day 1 of production Production slabs pre-specified by the government. Percentage revenue shares biddable by companies at the time of the auction rounds

Performance (contd) Drop in Production Increase in Consumption & Imports Source: BP Statistical Review, PPAC India (2013/14)

2. Conceptual Literature on Resource Taxation What to Tax - The Base Net income / rent / liberal regimes Pure rent = surplus or financial return not required to motivate economic behaviour Supply price of investment = costs of capital and operation plus risk premium. Lower supply price, higher rent. Risk premium affected by policy Zero marginal fiscal take Requires information on production volumes, costs, prices, investments Gross income / proprietorial regimes Positive marginal rent, reservation ground rent, and excess profits Information on prices and volumes only Mommer (1999), Baunsgaard (2001), Hotelling (1931) How to Tax - The Instruments Net Income /rent Resource Rent Taxes (R Factor, Rateof-Return), Progressive Profits Taxes, Corporate Income Tax, VAT Gross Income Royalties, Bonuses, Fixed Fees, Minimum Work Programs Other Government equal partner Government/NOC equity participation, Brown Tax ( Most systems use a hybrid or combinations of both elements

Equivalencies in Fiscal System Design Production Sharing Royalties Corporate Income Tax Royalty rate = Tax rate = cost oil cap government profit share Resource Rent Tax Brown Tax Paid Equity Tax rate = equity share Carried Interest Tax rate = equity share Target real rate of return = interest rate Concessional Equity Source: Baunsgaard (2001) Tax rate = share of initial concessional investment Tax rate = equity share Target rate of return = interest rate

Desirable Features & Tradeoffs Desirable features of a fiscal system (Nakhle, 2008) Efficiency Neutrality Equity Sharing of fiscal risk Simplicity Stability of fiscal terms and of revenues Trade-offs between features Neutrality trades off with efficiency and with revenue generation Neutrality trades off with simplicity Equity trades off with simplicity and efficiency Stability trades off with fiscal risk sharing Selection of instruments is contingent upon the government s objectives in fiscal system design

Analogies with Price Cap versus RoR Regulation PSC regime analogous to rate-of-return system; RSC regime to price cap system PSC (RoR) necessitates information on volumes, costs, prices and investment RSC (Price Cap) necessitates information only on volumes and prices Under a price cap type system, the government will most likely need to accept a lower overall expected level of taxation and lower take but stable (yet lower) revenues Conversely, investors face greater risks in sinking investments into difficult (offshore) areas In utilities and resource taxation price cap systems almost always revert towards rate-of-return (Tapia, 2012; Helm, 2010; Liston, 1993) The setting of appropriate caps ultimately necessitates information on cost functions Governments may therefore be compelled to intervene ex post There is a case for incorporating RoR elements into fiscal design ex ante

3. Empirical Literature Static Models: Discounted cash flow models using data on production, costs & prices under a high, medium and low case scenario. Fiscal terms used to compute economic measures. One or more of these are held constant to observe the effect of changes in the others. Results anchored to initial conditions employed. Restrictions on graphical representation of multidimensional data. Key Studies: Goldworthy and Zakharova (2010) Johnston (2003) Dongkun and Yan (2010) Daniel and McPherson (2010) Nakhle (2008) Tordo (2007) Al-Attair and Alomair (2005) Dharmadji and Parlindungan (2002) Iledare (2001) System Measures: AETR, IRR, sensitivity analysis AGR/ERR, savings index, take stat Government take, front loading index METR, AETR, NPV NPV, IRR, take NPV, IRR, take E&P costs Contractor take, IRR, NPV Prices, production, NPV, take Meta Models: A cash flow model of the system constructed and parameters of the system are defined and bound through design intervals based on historical data or reasonable assumptions combined with the given parameters in the fiscal system. Parameters of the system sampled from the design space and evaluated in the cash flow model. The results of the model and system parameters are then analysed and a linear model is constructed from the generated data. Kemp and Stephen (2011) Hong and Kaiser(2010) Adenikinju and Oderinde (2009) Iledare and Kaiser (2006) Kaiser and Pulsipher (2004) Prod, costs, invest., tax revenues NPV, IRR, GT, CT NPC, IRR GT CT NPV, IRR, GT CT NPV, IRR, GT, CT

4. Method Static Analysis

Meta Modelling

5. Analysis Static DCF model constructed for India s PSC regime Inputs Price Fiscal terms: royalty, cost recovery ceiling, corporate income tax, minimum alternate tax, share of profits from production to the government at the lower and upper tranches of the R-Factor investment multiple, and the contractor and government discount rates Inputs are applied to a real (representative) offshore gas field in India We obtain a set of economic measures Present value of cash flows to the contractor and government (NPGc, NPVg), the internal rate of return (IRR) and the discounted contractor and government take (NPVct, NPVgt)

DCF Model Inputs and Project Parameters Inputs (Fiscal Terms) Inputs Units Varnames Numbers Price $/MMBtu P 4.20 Royalty Rate % Roy 5, 10 Cost Recovery Ceiling % CR 80 Corporate Income Tax % CIT 34 Minimum Alternate Tax % MAT 18.5 Government Profit Share % IM1 0.3 at Lower R Factor (<1.5) Government Profit Share % IM2 0.8 at Higher R Factor (>3.5) Contractor Discount Rate % Dc 15 Government Discount Rate % Dg 12 Key Project Parameters Parameter Name Recoverable Reserves Field Life Total Capex Full Cycle Opex Measures 11 tcf 33 years $14.78 bn $0.76 /Bcf

DCF Model Results Set of Economic Measures Economic Measures Results NPVc $ 1,213 Mn NPVg $ 3,298 Mn IRR 22% NPVct 27% NPVgt 73%

Assumption variables The Meta Model (1) P, Roy, CR, CIT, IM1, IM2, Dc, Dg. Forecast variables NPVc, NPVg, IRR, NPVct, NPVgt. A design space is constructed for each assumption variable Based on reasonable assumptions or historical data (fit to distribution) 100, 500 and 1000 trial simulations of the assumption variables Obtain corresponding forecast variables

Design Space for Meta Model Parameter Price LN (3.05, 16.75) Royalty Rate U (0.05, 0.20) Cost Recovery Ceiling U (0.10, 1.00) Corporate Income Tax U (0.30, 0.40) Government Profit Share at Lower R Factor (<1.5) U (0.10, 0.90) Design Space Government Profit Share at Higher R Factor (>3.5) U (0.10, 0.90) Contractor Discount Rate U (0.12, 0.40) Government Discount Rate U (0.07, 0.12)

Dependent variables: NPVc, NPVg, IRR, NPVct, NPVgt Independent variables: P, Roy, CR, CIT, IM1, IM2, Dc, Dg The Meta Model (2) The dataset generated mimics time-series data. We test for stationarity We are interested in the relative influence (elasticity relationships) of indepvars upon the depvar. We estimate the following equation for depvars NPVc, NPVg, IRR, NPVct, NPVgt: L.φ (f, F) = k+ βl.p+ βl.roy+ βl.cr+ βl.cit+ βl.im1+ βl.im2+βl.dc+ βl.dg +ε We test for serial correlation and use robust standard errors. The coefficients stabilise at around 500 trials

Same procedure for Revenue Sharing Contract (RSC) Inputs (Fiscal Terms) Inputs Units Varnames Numbers Price $/MMBtu P 4.20 Corporate Income Tax % CIT 34 Minimum Alternate Tax % MAT 18.5 Revenue Share at Tranche A % R A 9 Revenue Share at Tranche B % R B 13 Revenue Share at Tranche C % R C 16 Revenue Share at Tranche D % R D 25 Contractor Discount Rate % Dc 15 Government Discount Rate % Dg 12 Key Project Parameters Production Slabs Parameter Name Recoverable Reserves Field Life Total Capex Total Opex Measures 11 tcf 33 years $14.78 bn $8.8 bn Production Govt Revenue Slabs (Bcf) Share A 0-73 9% B 74-1,825 13% C 1,826-2,920 6% D > 2,921 25% Source: Kelkar Committee Report, 2013

Set of Economic Measures Economic Measures Results NPVc $ 5,047 Mn NPVg $ 6,053 Mn IRR 91% dct 54.53% dgt 45.47% Design Space for Meta Model Parameter Price LN (3.05, 16.75) Corporate Income Tax U (0.30, 0.40) Revenue Share at Tranche A U (0.01, 0.50) Revenue Share at Tranche B U (0.01, 0.50) Revenue Share at Tranche C U (0.01, 0.50) Revenue Share at Tranche D U (0.01, 0.50) Contractor Discount Rate U (0.12, 0.40) Government Discount Rate U (0.07, 0.12) Design Space

6. Results Results from all simulations-psc Regime L.NPV C 100 500 1000 L.NPV G 100 500 1000 L.IRR 100 500 1000 L.NPV GT 100 500 1000 L.NPV CT 100 500 1000 φ (f, F) = k+ βl.p+ βl.roy+ βl.cr+ βl.cit+ βl.im1+ βl.im2+ βl.dc+ βl.dg +ε Trials k βl.p βl.roy βl.cr βl.cit βl.im1 βl.im2 βl.dc βl.dg R 2 Adj R 2 F Stat DW 2.21(4.73) 0.15(0.03) + -0.03(0.52) 0.25(0.37) -2.51(2.22) -0.67(0.35) 0.51(0.37) -2.43(0.62) + 0.77(1.46) # 0.53 0.43 6.15 + 2.5 0.52(1.42) 0.14 + -0.20(0.18) 0.01(0.13) 0.34(0.87) -0.68(0.13) + 0.04(0.13) -2.88(0.20) + -0.63(0.44) 0.58 0.57 45.17 + 1.9 2.31(1.0) 0.10 + -0.27(0.11) * 0.33(0.09) + -0.35(0.57) -0.48(0.08) + -0.11(0.08) -2.37(0.14) + -0.05(0.29) 0.53 0.53 75.86 + 2.0 6.51(0.96) 6.36(0.38) 7.24(0.34) -3.51(5.24) -4.50(1.61) -2.50(1.19) 0.05(0.14) 0.20(0.06) 0.12(0.04) -6.55(3.85) -6.90(1.08) -5.58(0.71) 0.13 + 0.12(0.003) + 0.10(0.002) + 0.20(0.10) + 0.19 + 0.13 + -0.001(0.001) + -0.004(0.0005) + -0.002(0.002) + 0.05(0.02)* 0.05 + 0.03(0.004) + 0.08(0.11) 0.15(0.05)* 0.07(0.04) # -0.12(0.57) -0.10(0.21) -0.32(0.14) * 0.03(0.02) 0.01 0.02 + -0.14(0.42) -0.32(0.14)* -0.34(0.08) + -0.13(0.07) * -0.15(0.03) + -0.05(0.03) # 0.54(0.41) 0.20(0.14) 0.63(0.11) + -0.01-0.02 + -0.02(0.003) + 0.32(0.30) 0.15(0.10) 0.39(0.06) + 0.27(0.49) 0.61(0.22) + 0.59(0.18) + -2.53(2.46) 0.84(0.94) -0.11(0.68) -0.03(0.07) 0.01(0.03) 0.03(0.02) -2.89(1.81) -0.45(0.66) -0.73(0.40) # 0.38(0.08) + 0.27(0.03) + 0.31 + -0.22(0.38) -0.32(0.15)* -0.07(0.10) 0.03 # 0.05 + 0.04(0.003) + -1.15(0.28) + -0.99(0.10) + -0.90(0.06) + -0.01(0.07) -0.001(0.03) 0.0001(0.03) 0.45(0.41) 0.04(0.15) -0.14(0.10) -0.001-0.003-0.002(0.003) 0.48(0.30) 0.05(0.10) -0.002(0.003) -0.15(0.14) -0.03(0.05) 0.004(0.05) 0.15(0.69) -0.28(0.23) 0.26(0.17) 0.09(0.02) + 0.08 + 0.07 + -2.59(0.51) + -2.86(0.15) + 0.07 + -1.85(0.29) + -2.17(0.12) + -1.85(0.10) + 0.59(1.62) -0.83(0.51) -0.08(0.35) -0.05(0.04) 0.01(0.02) -0.02* 2.18(1.19) # 1.43(0.34) + -0.02 # 0.83 0.78 0.73 0.59 0.52 0.50 0.33 0.38 0.31 0.51 0.65 0.31 0.82 0.77 0.73 0.52 0.51 0.49 0.28 0.37 0.30 0.42 0.64 0.30 55.44 + 214.02 + 333.58+ 8.21 + 35.41 + 66.39 + 5.69 + 37.02 + 55.65 + 5.94 + 61.08 + 55.65 + 2.0 1.9 2.0 2.4 1.8 2.0 2.4 2.0 1.9 2.5 2.0 1.9 + Significant at 1%, *Significant at 5%, # Significant at 10%, Standard errors in parentheses

L. φ (f, F) = k+ βl.p+ βl.roy+ βl.cr+ βl.cit+ βl.im1+ βl.im2+ βl.dc+ βl.dg+ε L.NPV C L.NPV G L.IRR L.NPV GT L.NPV CT k 0.52 (1.42) 6.36 (0.38) -4.50 (1.61) 0.20 (0.06) -6.90 (1.08) βl.p 0.14 ** 0.12 ** (0.003) 0.19 ** -0.004 ** (0.0005) 0.05 ** βl.roy -0.20 (0.18) 0.15* (0.05) -0.10 (0.21) 0.01-0.32* (0.14) βl.cr 0.01 (0.13) -0.15 ** (0.03) 0.20 (0.14) -0.02 ** 0.15 (0.10) βl.cit 0.34 (0.87) 0.61 ** (0.22) 0.84 (0.94) 0.01 (0.03) -0.45 (0.66) βl.im1-0.68 ** (0.13) 0.27 ** (0.03) -0.32* (0.15) 0.05 ** -0.99 ** (0.10) βl.im2 0.04 (0.13) -0.001 (0.03) 0.04 (0.15) -0.003 0.05 (0.10) βl.dc -2.88 ** (0.20) -0.03 (0.05) -0.28 (0.23) 0.08 ** -2.86 ** (0.15) βl.dg -0.63 (0.44) -2.17 ** (0.12) -0.83 (0.51) 0.01 (0.02) 1.43 ** (0.34) R 2 0.58 0.78 0.52 0.38 0.65 Adj R 2 0.57 0.77 0.51 0.37 0.64 F Stat 45.17 ** 214.02** 35.41 ** 37.02 ** 61.08 ** Durbin Watson 1.9 1.9 1.8 2.0 2.0 Trials 500 500 500 500 500 ** Significant at 1%, *Significant at 5%, Standard errors in parentheses Results PSC Regime

Results RSC Regime φ (f, F) = k+ βl.p+ βl.r A + βl.r B + βl.r C + βl.r D + βl.cit+ βl.dc+ βl.dg +ε L.NPV C L.NPV G L.NPV GT L.NPV CT k 3.55 (0.14) 7.24 (0.05) 0.23 (0.06) -3.46 (0.19) βl.p -0.014 0.01 4.42E-05-0.02 (0.02) βl.r A 0.011-0.003 (0.004) -0.003 (0.004) 0.01 βl.r B -0.25** 0.27** (0.004) 0.10** (0.0040-0.42** βl.r C -0.01 0.003 (0.001) -0.0004 (0.006) -0.01 βl.r D -0.001-0.004 (0.004) -0.003 (0.004) 0.001 βl.cit -0.30** (0.08) 0.45** (0.04) 0.16** (0.04) -0.60** (0.12) βl.dc -2.32** (0.02) 0.01 0.44** -1.88** (0.03) βl.dg -2.32** (0.02) -1.15** (0.02) -0.20** (0.02) 0.97** (0.06) R 2 0.97 0.94 0.87 0.93 Adj R 2 0.97 0.94 0.97 0.93 F Stat 1816.7** 1015.650** 414.46** 776.67** Durbin Watson 2.0 2.0 1.93 2.0 Trials 500 500 500 500 ** Significant at 1%, *Significant at 5%, Standard errors in parentheses

7. Observation & Conclusions Variables that have significant effects across all economic indicators: PSC Regime Price Share of profits to the government at the lower tranche of the R factor ( trigger tranche ) Coefficients larger than price Variables that have significant effects across all economic indicators: RSC Regime Share of revenues to the government at the first production slab ( trigger slab ) Corporate income tax Government discount rate Trigger tranche / trigger slab reflect rate-of-return characteristics

Observations & Conclusions (2) The results for trigger tranche/trigger slabs in the PSC/RSC system are indicative of rate-of-return influences on the fiscal system With a RSC, it could deter investment in difficult (offshore) areas as there is less traction in the system to offset investor risk The government s objective in fiscal design appears contrary in the RSC: it desires higher production and higher (and earlier) revenues, but there is not enough traction to offset the higher risks that come with it Alternatively, it could explicitly incorporate the rate-of-return element into fiscal system design This could be done through retaining the PSC, but adding an element of cost uplift, for instance It could also be done through designing a rate-of-return based tax combined with corporate income tax and/or royalties, instead of a RSC Areas for further work; RoR systems; testing functional relationships further

End anupama.sen@oxfordenergy.org

Significant Variables Compared NPVc NPVg IRR NPVct NPVgt PSC Regime Price Price Price Price (negligible) Price Trigger tranche (large) Discount rate (c) Trigger tranche Royalty Cost recovery Corporate income tax Trigger tranche Trigger tranche (large) Royalty Discount rates Trigger tranche Cost recovery Discount rate (g) (all negligible) Discount rate (g) RSC Regime Trigger slab Corporate income tax Trigger slab Corporate income tax No significant results Trigger slab Corporate income tax Trigger slab Corporate income tax Discount rates Discount rate (g) Discount rates Discount rates

Stationarity Tests PSC Regime Variable ADF Stats (at levels) Order of Integration NPV C -22.50 + I(0) NPV G -23.36 + I(0) NPV GT -22.47 + I(0) NPV CT -22.47 + I(0) P -23.25 + I(0) Roy -20.92 + I(0) CR -22.02 + I(0) CIT -23.28 + I(0) IM1-20.82 + I(0) IM2-22.76 + I(0) Dc -22.64 + I(0) Dg -21.97 + I(0) Critical Value at 1% -3.443228 Critical Value at 5% -2.867112 Critical Value at 10% -2.569800 +Significant at 1%

Variable ADF Stats (at levels) Order of Integration NPV C -24.37 + I(0) NPV G -22.13 + I(0) NPV GT -23.45 + I(0) NPV CT -23.45 + I(0) P -23.25 + I(0) R A -19.50 + I(0) R B -22.29 + I(0) R C -21.20 + I(0) R D -23.56 + I(0) CIT -21.65 + I(0) Dc -24.63 + I(0) Dg -22.85 + I(0) Critical Value at 1% -3.443228 Critical Value at 5% -2.867112 Critical Value at 10% -2.569800 +Significant at 1% RSC Regime