Petroleum fiscal systems and taxation

The state as owner sells not the oil but the right to produce it, and it takes its part of the resource rent (the value of the oil above all costs and the return the investor needs) through payments. Which payments are possible depends on the type of agreement (Licences and contracts for subsoil use). The lesson on the cash flow model shows where each one sits in that model. Here another question matters: what each payment is calculated from, and how it behaves when the price, the costs and the size of the field change.

Fixed payments. The signature bonus is paid when the company gets the right. A commercial discovery bonus is paid when a discovery is declared commercial (worth developing), and production bonuses when production reaches agreed levels. The area fee grows with time, so that a company does not hold land without working it. For the budget this is the most reliable and earliest money. For the investor it is the heaviest, because it is paid before anything is known about the result. At an auction the bonus is bid up, and competition brings it closer to the value of the area (the auction in Kazakhstan).

Payments on production. Royalty is paid in money or in oil (in kind), at a fixed rate or at a sliding rate that depends on daily production, on the price or on the depth of the sea. A production tax (a tax on the oil produced) works in the same way but is called a tax. In the USA, such state taxes are called severance taxes. Both need a price. In deals between independent parties (at arm's length), the price of the deal is used. When the oil is sold to the company's own trading company, a calculated price from market quotations is used, because otherwise revenue can be moved away by a low transfer price (a price set between companies of the same group). For marketable oil, including export oil, the Kazakhstan MET is calculated at the world price (the quotations of Urals, KEBCO or Brent), and for sales to Kazakhstan refineries at the plant's actual purchase price (Article 776 of the Tax Code).

Export payments and mandatory deliveries. An export duty and an export tax are charged on the volume or the value of the oil exported. An obligation to sell part of the oil or gas on the domestic market at a price below the export price is a hidden tax. The difference between the export revenue and the domestic revenue goes to consumers, and it too is counted in the government take. In Kazakhstan in 2026, a coefficient of 0.5 is applied to the MET rate for oil sold on the domestic market (Article 778).

Income taxes. The base is revenue minus the allowed expenses. Operating costs are deducted at once, capital costs through depreciation (spreading the cost of the equipment over several years), and losses are carried forward to the following years. When the income and costs of a project are taxed separately from the rest of the company's business, this is called ring-fencing: the losses of a new project do not reduce the tax on the old ones. It also matters what income tax does not deduct: the cost of capital. So it taxes not only the rent but also the normal return on the investment.

Taxes on rent and on excess profit try to tax only what is above the normal return. The classic resource rent tax is charged on the positive balance of the project's cash flow. Past negative balances are carried forward with an uplift (they grow by an extra percentage each year), which sets the threshold return. Other systems raise the rate by the R-factor, the ratio of the investor's accumulated receipts to its accumulated costs, or by the yearly return, like the Kazakhstan excess profit tax. A windfall profits tax is introduced once or for a time when prices have risen sharply. It takes part of the unexpected income, but it shows investors that the rules can change after the fact.

Production sharing and state participation. In a production sharing agreement, the contractor first takes back its costs as cost oil, and the state gets a share of the profit oil, which is what is left of the production (the three families of agreements). The tax character of the agreement is set by the cost recovery limit and the sharing scale. In some agreements, the contractor also pays income tax on its own share; in others, the tax is counted as included in the state's share. A state company can take part in a project as an ordinary partner, putting in its own share of costs. Or it can hold a share whose costs the investor pays at some stage, and then the participation works as a tax (see the section below).

The state and the investor ask a system the same questions. How much of the project's money does it take? When? And what will happen to that share if the price, the costs or the size of the field turn out to be different? Each question has its own measure.

The average take says how much the state takes now, but it does not say what will happen at another price. For that we need the marginal government take $m$: the part of one extra dollar of cash flow (from a higher price, lower costs or a new accumulation) that the payments will take. If the taxes are calculated in layers, each on what is left after the previous ones, it is built up like this:

The comparison of the marginal and the average take defines the progressivity of a system. Let the government take be $\mathrm{GT} = G/\Pi$, where $\Pi$ is the flow of the project before payments over its whole life, and $G$ is the payments. Then, when $\Pi$ changes (for example, with the price):

The third measure is the effective royalty rate: the smallest share of revenue that the state gets in any year, even a loss-making one. In a concession (a licence under which the company owns the oil), it is the sum of the payments on revenue. In production sharing, it also includes the part of the profit oil that goes to the state while costs are not yet recovered and the cost recovery limit applies:

Finally, front-end loading: how much of the payments to the state falls in the years before the investor has got its investment back. Bonuses and royalty make it large. The state is protected from the failure of the project, and the investor carries more risk and waits longer for payback (Cumulative cash flow and payback period). From the investor's side, the result of a system shows in the NPV and IRR (Net present value (NPV), Internal rate of return (IRR)), in the oil price at which the NPV is zero (the break-even price), and in the economic limit of production.

Let us compare three assumed regimes on one and the same assumed field, with the medium size and costs of this lesson, at three oil prices. The regimes are different on purpose. The first rests on a payment on revenue, the second on production sharing with a sliding scale, and the third on income and rent taxes.

The condition of progressivity explains why the lines part. At 60 dollars a barrel, the marginal take of regime A is 40.5%, as the formula with layers also gives, and its average take is 49.8%. The marginal take is smaller than the average, so the take falls as the price rises. For regime B, the marginal take over the whole life of the project is 62.8% against an average of 49.3%. It is even higher than the state's 30–60% share of profit oil in single years, because as the price rises, $R$ grows too, and with it the state's share in every following year. For regime C it is 60.6% against an average of 55.1%. In shares of the rent (in brackets in the table of takes above) the picture is different. At 40 dollars a barrel, the rent is small and all three regimes take most of it. As the price rises, the share of regime B in the rent first falls and then grows, while the share of regime C falls, from 70.7% at 40 dollars to 61.0% at 80. The income tax also taxes the normal return on investment, and its part of the rent falls from 55.9 to 22.5%, while the part of the rent tax grows. So when we call a system progressive, we say by which measure.

Price is not the only thing that makes a project poor or rich. Here are the same three regimes at US$60/bbl on three fields: a small one, the medium one from the previous section, and a large one. The small field has 8 million bbl over 8 years of production (years 3–10), capital costs of US$180 million (80 and 100 in years 1–2), operating costs of US$8 million a year plus US$6/bbl, and abandonment of US$15 million in year 11. The large field has the production profile of the medium one 2.5 times larger (75 million bbl), with capital costs 1.8 times larger (US$720 million), fixed operating costs of US$24 million a year, the same variable US$5/bbl, and abandonment of US$60 million. The fields are assumed.

Regime A takes 107.3% of the rent of the small field and 51.9% of the large one. The same royalty rate is heaviest for the one that has the least rent. The takes of regimes B and C without discounting, on the contrary, grow with size. In shares of the rent, regime C shows the price of the income tax: on the small field it alone takes 46.9% of its modest rent, because depreciation does not count the cost of capital.

A payment on revenue hurts small and old fields also by stopping production earlier (the economic limit). The small field has fixed costs of US$8 million a year and variable costs of US$6/bbl. Without a royalty, production at US$60/bbl pays for itself as long as it is above $8/(60 - 6) = 0.148$ million bbl a year. With a royalty of 15%, it pays for itself only while it is above $8/(51 - 6) = 0.178$ million bbl. The wells will be closed earlier, and part of the oil will stay in the reservoir. For the medium field, the same three regimes give the break-even prices of 42.9, 37.8 and 36.4 dollars a barrel against 33.6 dollars before payments. The better the field, the smaller the difference between the regimes in absolute numbers, but it still cuts off some of the weaker projects.

How a production sharing agreement is built is covered in the lesson Licences and contracts for subsoil use. On the tax side it has three levers. The cost recovery limit sets the effective royalty rate. The lower it is, the more profit oil the state gets from the first barrel, and the closer the agreement is to a regime with a royalty. The sharing scale sets the progressivity. The state's share can depend on daily production (then it grows with the size of the field but not with the price), or on the R-factor or the contractor's accumulated return (then it grows with the price and with the quality of the field). The uplift is the right to recover, for example, 20% more capital costs than were spent. It encourages investment, but also the padding of costs: each extra dollar comes back with a profit, so the state needs a strict audit of costs.

Production sharing differs in one more way: the contractor's barrels depend on the price. The more expensive the oil, the fewer barrels are needed to recover costs (the contractor's barrels). This also affects the reserves that the contractor shows in its reporting (Classification of reserves and resources).

Participation of a state company. If the company puts in its own share of the costs, it is an ordinary partner. It gets a share of the income for its own money, and for the assessment of the fiscal system this is an investment, not a tax. But if the investor pays the company's share of the costs at some stage, and the company gets its share of production, this is a carried interest, and it works as a tax on the size of the costs paid. In partnership deals between companies, the same arrangement is used to share risk (partnership).

In Kazakhstan, the right to hydrocarbons is given by a contract, and no new production sharing agreements have been signed since 2009 (the history of the subsoil regime). Taxes are set by the Tax Code of 18 July 2025 No. 214-VIII, in force from 1 January 2026. Its special payments and taxes of subsoil users are: the signature bonus; the payment for reimbursement of historical costs (what the state spent earlier on studying the area); the alternative subsoil use tax; the royalty; the Republic of Kazakhstan's share under production sharing; the mineral extraction tax (MET); and the excess profit tax (paragraph 3 of Article 753). In the Code, the royalty is a payment on solid minerals only (it has a chapter of its own), and the Republic's share applies to production sharing agreements. For oil, MET plays the role of the payment on production. Besides the special payments, a subsoil user pays the general taxes: corporate income tax (CIT), value added tax (VAT; 16%, and 0% on exports), property tax and social tax. An oil exporter also pays rent tax on export and an export customs duty.

• MET on oil. The base is the value of the oil produced in the quarter (Articles 775, 776). In 2026 the rate depends on the planned annual production under the contract: 5% up to 250 thousand t, 7% up to 500 thousand, 8% up to 1 million, 9% up to 2 million, 10% up to 3 million, 11% up to 4 million, 12% up to 5 million, 13% up to 7 million, 15% up to 10 million t, and 18% above that (Article 778). One rate is applied to the whole production of the year. If the actual production turns out to be in another tier, the tax for the year is recalculated (Article 777).
• Rent tax on export. One rate applies to the whole value of the oil exported: 0% at a world price up to US$40/bbl, 7% up to 50, 11% up to 60, 14% up to 70, 16% up to 80, and then up to 32% at a price above 180 (Article 749). In 2026 MET payers also pay it: the exclusion of MET payers from the payers of this tax is suspended until 1 January 2027 (Article 835).
• Export customs duty on crude oil. Below US$25/bbl there is none. At an average market price from 25 to 105 dollars a barrel, the duty in dollars per tonne equals the price in dollars per barrel. Above 105, it is from 115 to 236 dollars per tonne by a table. It is not paid, among other cases, on exports to countries with which there are treaties on exemption, under production sharing agreements signed before 2009 that have such an exemption, and for complex projects, for 10 or 20 years from the start of export (Order No. 169-NQ).
• CIT is 20% (Article 357), and a loss is carried forward for 10 years (Article 339). Accounts are kept separately for each contract, and the losses of a contract are covered only by the future income of the same contract (paragraph 6 of Article 757). This is the Kazakhstan form of ring-fencing. MET, the rent tax and other taxes reduce the CIT base. CIT itself, the excess profit tax and the alternative tax are not deductible (Article 272).
• Excess profit tax is calculated for a year, for each contract (Article 798). The part of the net income above 25% of the deductions is taxed (Articles 800, 801). The net income here is the taxable income for this tax minus the CIT on the contract. The deductions are the expenses deductible for CIT, and the actual capital costs of the year instead of depreciation (Article 804). An excess of deductions over income is carried forward to the following years (paragraph 2 of Article 802). The rates are by tiers (Articles 807, 808), with the formula below.
• The alternative subsoil use tax replaces the payment for historical costs, MET and the excess profit tax. It can be chosen by contracts on areas in the Kazakhstan sector of the Caspian Sea, on accumulations whose top is at a depth of 4,500 m or more and whose bottom is at 5,000 m or more, and on depleting fields. The choice is made by a notification within 30 calendar days of registration of the contract or of an addendum to it, with no right to change one's mind until the contract ends. For complex projects it is mandatory (Article 812). The base is income minus deductions under the CIT rules, without interest payments and CIT. The rate depends on the world price: 0% up to US$50/bbl, 6% up to 60, 12% up to 70, and so on up to 42% above 150 (Articles 813, 814).
• Bonus and historical costs. The starting signature bonus for exploration and production without approved reserves is 3,000 MCI plus 10 MCI for each block above 300. With approved reserves it is 0.04% of the value of reserves of categories A, B and C1 and 0.01% of the value of category C2 and of resources C3, but not less than 10,000 MCI (A, B, C1 and C2 are categories of the national reserves classification of Kazakhstan and Russia). It is paid within 20 working days (Articles 761, 764; the auction). The payment for reimbursement of historical costs gives the state back its costs of studying the area. Up to 10,000 MCI it is paid at once, and above that in equal parts for up to 10 years (Article 769).

From 2027 the MET rate on oil is taken from a matrix. The rows are annual production (up to 250 thousand t, …, above 10 million t) and the columns are the world price from US$20 to US$200/bbl (subparagraph 2 of paragraph 1 of Article 778). The rate grows with the volume and, almost everywhere, with the price. In the row “up to 1,000,000 tonnes” it is 22% in the columns of 20, 30 and 40 dollars, 29% in the column of 50, 33% in 60 and 38% in 80. At the same time, oil produced by MET payers stops being charged the rent tax on export (Article 746). One exported barrel shows how this changes the payments on revenue.

In the terms of this lesson, the Kazakhstan system looks like this. The main burden is payments on revenue (MET, the rent tax, the duty). They are regressive with respect to costs: a costly field pays the same per barrel as a cheap one. The MET scale by annual production makes them heavier for large contracts, and the rent tax and the 2027 matrix make them heavier at high prices. CIT, the excess profit tax (a progressive layer) and the alternative tax are linked to profit. The state offers the alternative tax to offshore, deep and depleting fields, where taxes on revenue would stop projects. For weak fields there is a MET relief. The MunaiHub economic model, in the “Investors” section, calculates the NPV and IRR of a project with the 2026 taxes (it has no signature bonus), and it is convenient for repeating the conclusions of this lesson for your own field.

There is no perfect fiscal system, because its aims argue with each other. The state wants to take most of the rent, to get the money early and reliably, to collect it simply, and not to stop projects. The investor wants to keep its return and the predictability of the terms.

• Neutrality. A tax should not change decisions: whether to develop, how much to recover, when to stop production. The closest to this is a rent tax with a threshold equal to the investor's rate (see the advanced section). The furthest are bonuses and payments on revenue.
• Progressivity. A system that adjusts itself to the price and the quality of the field needs revision less often. At high prices the state gets more without new laws, and at low prices projects keep working.
• Reliability and timing of the budget's income. Royalty and bonuses come early and at any price. Rent taxes come late and only from successful projects. So almost all systems mix both kinds of payments.
• Simplicity. A royalty is calculated from the volume and the price. Income taxes and rent taxes require checking of costs and transfer prices, and production sharing also requires an audit of the recoverable costs.
• Stability. Each change of rules after the fact adds a risk premium to the investor's rate and reduces the rent that can be divided (the risk of terms, Market, financial and regulatory uncertainty).

Stepped scales. If one rate is applied to the whole base, and the rate itself depends on the price or the volume, then at the border of a step the income drops with a jump. The rent tax on export at a price of exactly US$50/bbl is 7%, and the exporter keeps $50\cdot0.93 = 46.5$ dollars a barrel. At 51 dollars it is already 11%, and the exporter keeps $51\cdot0.89 = 45.39$ dollars. The exporter keeps 46.5 dollars again only at a price of $46.5/0.89 = 52.25$ dollars a barrel. The 2026 MET scale by annual production works in the same way. With a plan of 1.00 million t the rate is 8%, and with 1.01 million t it is 9% on all the oil. MET grows by $0.09\cdot1010 - 0.08\cdot1000$ $= 10.9$ thousand t of oil in terms of its value, which is more than the extra 10 thousand t are worth. Scales in which each rate applies only to its own part of the base (like the excess profit tax) do not give such jumps.

• Transfer prices. A tax on revenue and profit can be reduced by selling the oil cheaply to the company's own trading company. So the base is calculated from quotations or by the rules of transfer pricing.
• Padding of costs. If costs are recovered with an uplift or reduce the rent tax, the investor has a weaker reason to save them. The state answers with audits and maximum norms.
• The borders of an object. Scales by the production of a field or a contract and the rules of ring-fencing make it important what counts as one object. So the law itself defines what a scale is applied to.
• Timing. The R-factor and yearly thresholds can be moved by shifting costs from year to year. A rent tax with a carried-forward balance is not very sensitive to this.