Drilling for Value, Pt. 4: The Economics of Petroleum Exploration and Production

Note: this post has been heavily redacted since its original data of publication in order to expand on the fundamentals of petroleum geology and the upstream business elsewhere. 


  • Economic models use assumptions which simplify the effects of accounting, taxes, regulations, and other minutiae in order to glean insights into the drivers of market behavior and value.
  • The effects depletion and commoditization, relatively low cash costs, and often prohibitive resource replacement costs drive the endemically cyclical petroleum investment cycle
  • Petroleum economics are strongly levered to petroleum prices and other extrinsic factors.
  • Maintaining a sufficiently low cost of supply is the primary operational lever capable of creating long-term investment value in the upstream business.
  • Timings of costs are a key consideration for evaluating investment decisions — known discount rates simplify decisions regarding timing preferences.

Figure 1: Pecos, Texas Oilfield
Source: Alexander Hogue. Pecos, Texas Oilfield. 1937

The Economics of the Upstream Petroleum Industry
The economics of the petroleum extraction is overwhelmingly colored by the economic factors of depletion and commoditization. Due to the fact that production depletes limited natural resources, the upstream industry must constantly explore for and develop additional resources. Given that the capital investments required to replace depleted resources are usually quite significant in relation to operating costs, resource replacement is a primary driver of costs. Commoditization describes the lack of differentiation in upstream business models and their end products. As a direct result of commoditization, the value propositions of upstream businesses are strongly levered to external market conditions (i.e., namely prices). Taken together, high replacement costs and supplier susceptibility to external market conditions have resulted in endemically cyclical petroleum supplies and prices.

Concepts of Economic Value
The goal of economic valuation is to derive estimates of real worth or value. The accumulation of future money flows over time forms the basis for economic value. Operating profits, revenues less expenses, are usually the main driver of earnings. Earnings are the portion of profits which accumulate (i.e.,  what you keep).

In a perfect world, future money flows are objectively represented with regard to their economic (vice accounting) significance and, depending on investor preference, are discounted to their present values. However, because regulatory and reporting frameworks allow accountants and managers to exercise a significant degree of discretion in how upstream assets are recognized and costs are expensed, accounting profits (i.e., profits prepared according to generally accepted accounting principles (GAAP)) usually provide limited information regarding economic earnings. In fact, it is not unusual for an oil and gas E&P company to run at a cash flow deficit almost indefinitely while still declaring an GAAP profitability as long as that company continues to inflate its balance sheet by liberally capitalizing costs.

Cash flows are therefore a better proxy for economic profits since they are less easily manipulated than GAAP earnings and are recognized in the economically correct time frame. However, since multiple types of business activities can throw off cash (e.g., in the form of earnings, financing, investments, or asset sales), reported cash flows for any finite time frame may not be indicative of long-term, sustainable earning power.

Figure 2: Upstream Oil and Gas: GAAP Profits vs. Free Cash Flow Margins
GAAP Income vs Free Cash FlowSource: Author; Portfolio123.com

Note to Figure 2: Margins are industry aggregates of all U.S. publically traded companies assigned to primary GICS Codes 10102010 (Integrated Oil & Gas) and 10102020 (Oil & Gas Exploration & Production). Net Income is defined as Net Income excluding Extraordinary Items. Free Cash Flow is defined as Cash Flows from Operations less Cash from Investing. Both earnings metrics are divided by quarterly Sales.

Other notions of value, such as accounting book value and market value, can inform economic estimates, but are also not necessarily indicative of them.

Accounting book value represents historical cost bases plus depreciation plus accumulated earnings. However, book value is subject to significant regulatory adjustments and managerial discretions which, when compounded over numerous decision points, complicate any reconciliation between assets’ historical cost bases and their holding values.

Markets tend to value assets very closely to their actual economic values, except when they are very wrong. Markets ideally function as price discovery mechanisms in which an auction matches competitive bids with offers in an orderly and rational manner. According to the efficient market hypothesis, a market which is sufficiently open and liquid should reflect the actual value of assets. However, markets are also susceptible to bouts of irrationality; the wisdom of crowds often gives way to, and may be indistinguishable from, the madness of the mob. Even when there is no market for a given asset, investors can infer estimates of fair value from market or transaction values of comparable assets and companies.

Price Equilibrium and Volatility
Petroleum prices are volatile and subject to persistent and unexpected price fluctuations (i.e., shocks) due to its integral role as the driver of modern economies. Despite all the hoopla about how manipulation of commodities futures contracts has caused massive shocks and assets bubbles, the total open interest on oil futures contracts on the NYMEX and ICE represent only 5% of the physical market for oil. Securitization cannot possibly explain endemic market shocks. Not shockingly, a classical interpretation of supply and demand best explains market action. According to Baumeister and Kilian (2016), “Most major oil price fluctuations dating back to 1973 are largely explained by shifts in the demand for crude oil“.

In classical economics, price and supply are positively correlated; when prices increase, there is more incentive for suppliers to bring additional supply to the market. Likewise, price and demand are negatively correlated; when prices decrease, demanded goods will appear more attractively priced compared to substitute goods.

Economic factors which affect petroleum supply include resource depletion, production declines, technological advances, and capacity additions. Factors which affect demand include weather seasonalities and consumer behavior. Resource depletion, secular shifts in consumer preferences, and the emergence of substitute forms of energy and hydrocarbons have the most meaningful, long-term impacts on supply and demand.

Elasticity describes how sensitive a dependent variable is to a change in an independent variable (i.e., the price elasticity of supply is the expected change in price for a change in quantities supplied). Because petroleum is a staple good for there are no readily available commodity substitutes, demand is relatively inelastic to changes in prices. The elasticity of petroleum supply can even become inverted when lower petroleum prices drive producers to make up lost revenues through increased production.

When the quantity of product supplied equals the quantity demanded, prices are said to be in equilibrium. The actual price which someone pays for a barrel of oil (or cubic foot or natural gas) is determined by the marginal barrel. The cost of supply required to sate the very last barrel of demand sets the price of the marginal barrel. Price differences between various quantities of petroleum are mainly due to marketing and regulatory costs, yet global petroleum prices are kept in lock step by arbitrageurs (i.e., vile speculators). The existences of significant violations to the no arbitrage principle are extremely difficult to prove and probably rare.

Due to cyclically varying of supply and demand elasticities, petroleum prices can simultaneously have multiple equilibria. Furthermore, prices tend to be sticky around a given level until a force majeure displaces them to the next feasible equilibrium. The phenomenon whereupon prices “jump” to next equilibrium explains many characteristics of petroleum market volatility. A large body of academic research indicates that jump stochastic models — vis-a-vis a Poisson process — closely resemble empirical oil price volatility.

In more eloquent terms:

The volatility of oil prices is inherently tied to the low responsiveness or inelasticity of supply and demand to price changes in the short term. Crude oil production capacity and equipment that uses petroleum products as its main source of energy are relatively fixed in the near term. It takes years to develop new supply sources or vary production, and it is hard for consumers to switch to other fuels or to increase fuel efficiency in the near term when prices rise. Under such conditions, a large price change can be necessary to rebalance physical supply and demand.
EIA. Energy Explained: Oil Prices

Figure 3: Petroleum Equilibrium Curve
Upstream Supply and Demand Equilibria
Source: Author

The Petroleum Investment and Commodities Cycles
The petroleum industry, like any other, is subject to the laws of supply and demand, except when it isn’t. Capital investments for the development of oil and gas production are highly correlated to petroleum prices. This correlation leads to production booms and busts which further lead to over and under-investment cycles. These cycles sometimes result in counter-doctrinal market reactions which require disproportionately large price swings to rebalance the market.

The upstream industry requires frequent influxes of investment capital to replace depleting resources and declining production. Since production tends to be proportional to the quantities of remaining developed resource, production will decline unless producers invest in future production. In the case that a company invests nothing into replacing resources, future production and economic value will be limited to existing wells. When the amounts of excess cash which upstream projects throw off are insufficient to satisfy investment and asset replacements costs, producers must tap external sources. With oil prices at or below $60 per barrel, few resource plays throw off enough cash to internally sustain (let alone grow) production.

The fact that high prices attract capital investment, while low prices discourage investment, inevitably leads to cyclically recurring supply surpluses and deficits. The economically counterproductive nature of the self-inflicted cycle seems obvious enough, yet many operators, acting on real or perceived pressures from financial markets (e.g., many companies’ executive salaries and bonuses are based on production and/or reserve-growth), clamor for production and asset growth often in spite of the costs. The rush for production growth usually reaches its apex during periods of maximum market euphoria (or about the time when commodity prices reach their highs) which is the worst time to over-invest. The below figure, taken from the EIA, demonstrates the stark correlation.

Figure 4: Correlation between Petroleum Prices and Investment Capital Influx
EIA - Sustained low prices could reduce exploration and production investment
Source: EIA. Sustained low oil prices could reduce exploration and production investment. Today In Energy. 24 Sep 2015

Similarly, Baker Hughes’ United States drilling rig count data correlates strongly to and lags slightly behind oil prices.

Typified Petroleum Investment Cycle
As a result of the lag between petroleum price innovations and supply responses, petroleum production tends to cyclically over- and under-shoot demand, thereby exacerbating perennially cyclical market conditions.

During normal market conditions, prices may not be high enough to sanction sufficient investment into future production. Over time, production declines and demand growth lead to the eventual supply crunch (whether actual or perceived). Prices then increase and, according to conventional economic theory, supply also increases.

However, based on the history of the lag between prices and supply, it takes about five years from peak oil prices to generate peak global oil production — finding and development (F&D) costs are front-loaded and take several years to result in significant new production. If incremental increases in supply from swing producers (e.g., producers with excess capacity; marginal producers with shut-in capacity; or, producers with an inventory of short-cycle investments) are not able to sate demand, prices must increase beyond the cost of adding new supply. Often this increase is large and rapid.

As figure 4 shows, however, there is virtually no lag between prices increases and spending. As a result of lag between capital and production, many projects tend to come online at once and production overshoots. The result of excess production is growing inventories and the ensuing supply glut.

As a result of growing inventories, prices drop, but production typically does not fall commensurately since production costs tend to be very modest compared to F&D costs. In addition, efficiencies tend to increase during the down-cycle. Even though many projects, over their full life cycle, will prove to be uneconomic, the majority of the capital costs have already been sunk – thus, it is usually less economic to shut off uneconomic production once it has commenced. As long as operating netbacks remain positive, producers will continue to produce from existing, developed reservoirs.

In many instances, a drop in petroleum prices can actually lead to increased production as companies with significant social and fiscal liabilities increase production to offset lost revenues. This phenomenon violates linear supply-demand equilibrium models (i.e., the kind which are taught in schools). Producers which uneconomically maintain or grow production during downturns will actually exacerbate any over-supply concerns and thereby prolong the amount of price change and length of time needed to purge markets of malinvestment.

In order to for excess supply to be wiped out, prices need to fall lower than marginal cost of existing supply and stay there as long as it takes for markets to rebalance. While simple in premise, an unbalanced market often takes much longer to balance than many experts initially expect. For example, during commodity price downturns, lower prices exert pressure on the entire supply chain which causes the marginal cost of supply to also fall. Lower costs exert a positive feedback loop on the commodities cycle, allowing prices to drop even further.

Underinvestment is common market response to imbalances. After prolonged periods in which per unit resource replacement costs exceed operating netbacks, broadly based producer spending cuts can further exacerbate market imbalances, albeit in the other direction.

Thus the investment and commodities cycles are mutually and self perpetuating. Or at least, this is how things have worked out since the industrial revolution. Absent of the any major changes to the world economic system or human nature, I foresee no substantive change to these basic cycles.

Figure 5: Typified Petroleum Investment Cycle
Typified Petroleum Investment Cycle3
Source: Author

Geopolitical Risks and Other Extrinsic Factors
Petroleum is a strategic commodity — in his Pulitzer Prize winning novel, “The Quest“, Daniel Yergin thoroughly documents petroleum’s transformative role in man’s and the state’s quest for power and wealth.

Although geopolitical factors are extraneous to the natural economic order, they nonetheless exert significant influence over market supply. National and corporate interests share a mutual interest in counteracting cyclical markets, thereby maximizing prices but not to the point of demand destruction. In order to stabilize markets while keeping prices high, nations have formed non-governmental entities (i.e., cartels?) out of their oil companies which seek to better align supply with demand (i.e. set prices?). Most notable is the Petroleum Exporting Countries (OPEC), but other often more loosely organized entities exist, such as the Organisation for Economic Co-operation and Development (OECD), which seek to promote beneficial national and economic policies (i.e., a milder form of market manipulation). These cartels ostensibly serve political agendas, but the line between corporate and state interests are blurred when states directly own or control corporations — note the preponderance of National Oil Companies (NOCs).

The cartels have met with mixed success over the years. Sometimes they are able to exert themselves. Other times, their power comes from a captive audience (i.e., perception). But every once in a while, market forces overwhelm these cartels, which had been the case as of late. Since 2014, OPEC has lost much influence due to innovations in the private sector which have unlocked vast quantities of previously uneconomic resources. In November 2016, OPEC reasserted itself, agreeing to its first production cut in 8 years. Only time will tell how effective it will be — but if Saudi Arabia’s former Minister of Petroleum and Mineral Resources is correct, its effect will be mitigated by the reality that “we tend to cheat“.

Politically motivated cartels sometimes counteract commodities cycles, but also sometimes exacerbate them, yet always convolute the driving forces of supply.

Geopolitical forces can also result in unforeseen markets shocks due to including wars and coups.

Shocks due to natural causes (e.g., weather, natural disasters, etc…) are yet more extrinsic forces which can affect investment and commodities cycles. Since weather and natural disasters tend to be regional in nature, they are more important in setting natural gas prices. However, oil, due its ability to be easily transported, is more resilient to extrinsic shocks. Technologies and facilities which facilitate the transportation of natural gas — including pipelines, liquefied natural gas plants (LNG) and tankers, and gas-to-liquids (GTL) conversion plants — may counteract seasonal effects, but their high capital costs require very favorable market conditions (i.e., large and sustained price differentials between regional gas prices and/or gas-liquids spreads) for economic sanction.

Figure 6: Crude oil prices react to a variety of geopolitical and economic events
EIA_Crude oil prices and geopolitical and economic events
Source: EIA. What Drives Crude Oil Prices? An analysis of 7 factors that influence oil markets, with chart data updated monthly and quarterly. Energy & Financial Markets

Prices, Realizations, and Revenues
Upstream petroleum companies are price-takers — i.e., price realizations are largely dictated to them by commodity market conditions. Outside of projects which seek to uplift price realizations or hedging activities, upstream revenues are simply the product of production and prices. The upstream business model is characterized by a general lack of economic self determination, implying that most upstream businesses are almost entirely dependent on ever increasing petroleum prices to justify their value propositions.

Overwhelmingly, the evidence supports this notion: investments into upstream companies are primarily leveraged commodity price plays. This observation has significant implications for prospective investors. As a result, prognosticators have devised countless ways to combine their knowledge of economic, market, and geopolitical forces to forecast oil prices in the hopes that this will lead to advantageous market timing and/or investments into the right types of assets during the right phase of the cycle.

Figures 7 and 8: Upstream Total Return Index versus Crude Oil Prices
Upstream Total Return Index vs. Crude Oil Prices - time series2

Upstream Total Return Index vs. Crude Oil Prices - scatterplot
Source: Author; Portfolio123.com

Note to Figures 7 and 8: Custom indices are constructed according to a modified capitalization weighted indexing methodology. Upstream Total Return Index includes all U.S. publicly traded companies in the Portfolio123 database which are assigned primary GICS Codes 10102010 (Integrated Oil & Gas) and 10102020 (Oil & Gas Exploration & Production). Dividends explain nearly half of the returns of holding a capitalization weighted index of oil and gas producers. A price return index which excludes dividends would have increased about 58% of the rate of the total return index.

Prognostications usually fall flat, however. More overwhelming empirical evidence shows that experts are more likely to get things wrong, especially the big moves. Moreover, even good economic forecasts may not be very useful to investors of upstream companies or their managements. From a capital allocation perspective, economic forecasts may be meaningful to select the most advantageous sectors or products to invest in. However, if constrained to investments in assets or securities within the upstream petroleum sector, macroeconomic factors are mostly extraneous since neither managements nor investors exert meaningful influence over them.

Reference Appendix B for a more detailed discussion of asset price forecasts.

In response to their lack of economic self-determinism and inabilities to estimate long-run prices, many upstream businesses seek to enhance their market positions by improving price realizations (i.e., achieve price uplift). Most examples of uplift projects are required to justify commercial and/or economic viability — without which, many projects would be uneconomic. In fact, such attempts can easily erode value propositions due to the significant costs incurred and prolonged project life cycles.

Additionally, many upstream companies attempt to hedge revenues against adverse market conditions through selling forward and engaging in derivatives contracts. While hedging can “lock in” advantageous prices and limit the effects of adverse market conditions, it is also likely to lock in disadvantageous prices and limit upside. Academic research indicates that oil producers historically paid a risk premium to speculators, but that relationship has flipped in recent history, likely due to increasing retail investor interest in owning oil contracts as a form of portfolio diversification. In any case, negative carry when futures terms structures (i.e., strip prices) are downward sloping (i.e., backwardated) results in an expected cost to producers who sell forward. Even when term structures are steeply upward sloping (i.e., in contango) the expected gain from selling forward is small compared to the price risks. Moreover, if a business justifies its activities under the assumption that commodity spot prices will rise over the long-run, then hedging is likely to either destroy value or the business is likely to fail. Academic research corroborates this intuition, finding that although hedging dampens stock price sensitivity to fluctuations in oil and gas prices, it is not a panacea for market value creation1.

The measured costs of petroleum extraction depend on the metric and accounting treatment. The tendency is for companies to understate actual costs in order to appear more profitable. In addition lack of pricing control, petroleum production cost structures are also strongly directly and indirectly linked to energy prices which further erodes the prospects for self-determination. Additionally, the upstream business model is subject to perennially diminishing returns of invested capital due to resource depletion. Despite the challenges involved, select producers have been able to maintain sufficiently low costs of supply and/or capital flexibility such that they are able to achieve acceptable returns in nearly any conceivable commodity market. The ability to control costs is therefore the main operating lever available to upstream managers which can reliably result in long-term investment value (the major exception to this rule, rational integration, is discussed at a later point).

There are basically two types of costs incurred in any business: expensed costs and capitalized costs. Expensed costs are those realized on the income statements as costs. Capitalized costs are those which, like expensed costs, result in a use of funds (i.e., cash), but are added to the balance sheet as an asset and expensed over time through depreciation, depletion, and amortization (DD&A), but can also be written off or reduced in other, sometimes borderline surreptitious, ways (e.g., through impairments, disposals, write-downs, write-offs, charge-offs, and even extraordinary items).

Successful efforts (SE) and full cost (FC) accounting treatments attempt to standardize how costs of extractive industries (e.g., petroleum E&P) are capitalized. US GAAP allows companies to choose either method or a hybrid, while IFRS mandates full cost only. Still, there exists significant managerial discretion as to cost treatment under either method, and the upfront cost-intensive nature of the business dictates that asset inflation is the norm. Moreover, the lack of economic self determination of the business also means that impairments and write-offs of these capitalized costs are frequent and sometimes massive, especially for companies which utilize the FC method. In any case, the tendency for many companies is to aggressively capitalize costs in order to increase income in the near-term naturally increases depletion costs. In order to manage the growing depletion costs going forward, these companies then realize massive impairments and write-downs which are then dismissed as “one bad quarter”. After such an impairment, a company can often reassure shareholders by returning to GAAP profitability, even if it still continues in its profligate ways.

In petroleum economics, cost of supply refers to the all-in cost of bringing supply to the market. Half-cycle costs refer to those which are realized during D&C and production — it excludes basically all exploration and land acquisition costs. Full-cycle costs — equal to the sum of all expensed and capitalized costs — include all half-cycle costs in addition to “non-core” operating costs (e.g., F&D costs such as land/leasing, seismic, processing infrastructure, et cetera), costs incurred in related capital projects (such as processing and treatment facilities), asset retirement costs (i.e., decommissioning and environmental costs), and optionally general corporate and financing costs. Full-cycle economics provide a better measure of capital efficiency since all non-discretionary expenses of exploring for, developing, and producing petroleum assets are captured. Well level economics are almost always stated using half-cycle cost measures, which often obscures and over-states their actual economic values within a larger corporate entity. Corporate presentations usually depict field and well-level metrics (e.g., NPVs, IRRs, Pay-back periods, etc…) using half-cycle costs only. This practice is at least somewhat disingenuous as it tends to vastly understate actual costs and overstate actual economic value-add. This is only one of a several games that petroleum producers play on investors. Furthermore, both buy and sell-side equity analysts often mirror the corporate practice of aggregating wells using half-cycle economics but — never fear — “compensate” by using higher discount rates as proxies for corporate overhead and “risk” — these are very weak assumptions.

Figure 10: Upstream Business Phases and Associated Costs

Source: Author

In order to control costs, management can exercise fiscal and capital discipline, and harness internal efficiencies. However, many components of costs are linked to factors over which management has no control. Over the short term, costs are linked to energy prices due to the fact that energy production expends energy. This is especially true in wells which utilize secondary and tertiary recovery techniques such as artificial lift and liquid/gas/polymer injection.

Costs are also correlated with service firm margins which are in turn correlated with energy prices.

Figure 11: Oil Service Firm Operating Margins and Oil Prices
BTU Analytics - HAL SLB Operating Margins and WTI Pricing
Source: BTU Analytics. Oil Service Costs To Rise, But How Much And When? Seeking Alpha. 28 Jun 2016

A recent joint EIA-IHS report, Trends in U.S. Oil and Natural Gas Upstream Costs, corroborates the price-cost relationship, citing that during the challenging price environment of 2015, average well drilling and completion costs in five onshore areas were down between 25% and 30% from their 2012 levels. Some of these cost reductions were due to improved efficiencies, others were due to decreased energy costs, while the rest can be attributed to pressures on the overall petroleum supply chain. When prices inevitably rise again, so will costs.

At the well level, per unit of production (i.e., per barrel equivalent) production costs tend to be low for newly completed wells, but rise over time due to increased proportions of water production and the necessity for artificial lift and recovery methods to sustain production – older wells also require more maintenance.

Figure 12: Typified Well-level Marginal Cost of Supply

Sources: Höök, Davidsson, et al. Decline and depletion rates of oil production: a comprehensive investigation. 2 December 2013; Author

At the field or reservoir level, costs tend to decrease for a while as geological uncertainty decreases, and as processes and technology improve, but eventually rise for the same reasons.

Figure 13: J-Curve Marginal Cost of Supply and Diminishing Returns on Invested Energy (Field and/or Reservoir Level)

Source: Author

The law of diminishing returns drives rising per unit (real) production costs. Economic conditions aside, resources deplete as resources are extracted and grow as technology improves. Likewise, as the technical difficulty of finding, developing, and extracting resources increases, so does the the energy intensity. At some point, however, natural laws dictate that more energy is expended than recovered. At this point, any further extraction is unsustainable.

For an explanation as to how the cost-curve in Figure (12) manifests itself as the shape observed in Figure (13) when aggregated across multiple wells, please refer to Dennis Coyne’s discussion on convolution functions.

Declines, Depletion, Energy Intensity, and Diminishing Returns
Petroleum is a non-renewable resource, and as such, the relative ease by which it can be extracted is proportional to the amount remaining in the ground. As remaining resources are increasingly depleted, natural declines set in. The characteristics of these declines are overwhelming idiosyncratic — they vary at the well, reservoir (i.e., pay zone), field, play, basin, and global levels. One quality, though, is consistent: natural declines (i.e., the rate of production decay) are proportional to the rate of depletion (i.e., the production rate with respect to remaining resource).

The onset of a natural decline always implies that increasing amounts energy are be required to maintain constant production. Declining Energy Returns on Energy Invested (EROEI) for petroleum resources have been meticulously and consistently documented. In many regions and petroleum resource categories, petroleum is already in danger of becoming sub-economic.

A more detailed discussion of energetic factors of resource economics is contained in Appendix A.

In the end, no amount of innovation can buy enough time to overcome the law of diminishing returns. However, efficiency gains to refining and power generation processes offset, and at times reverse, declining returns — converging evidence supporting the economic advantages of integration. Moreover, recent well and field-level advances (e.g., hydraulic fracturing; advanced seismic imagery analysis; horizontal drilling; etc…) have shifted the balance of power back towards producers and short-term projects. How much of that shift is sustainable remains to be seen, but it does seem apparent that it is only a matter of time before advances which took place in North American shale find roots on foreign shores. EIA’s assessment for “Technically Recoverable Shale Oil and Shale Gas Resources” investigates the resource potentials of 137 Shale Formations in 41 countries outside the United States2. Not surprisingly, the study’s key finding is that geology is a global phenomenon. The online publication Drill or Drop? provides real-time updates on the global fracking phenomenon.

Value Creation
To claim that profit is the key to value creation is an extreme tautology. Growing the scale of profitable operations, all things equal, by definition increases value. However, since upstream companies are extremely limited in their abilities to control selling prices, growth also magnifies the potential for losses in the event of a downturn. Maintaining a sufficiently low cost resource base, which balances the needs for sustainability and scale, is therefore the only reliable operating lever available to managements to maximize profits, ensure company survivability throughout cyclical market conditions, and enhance the potential for long-term value creation.

Exceptions to the “low cost of supply” rule apply when commodity goods are advantageously integrated into a value chain. For example, a relatively high-cost natural gas producer which owns and operates its own gas treatment facility and is advantageously located near an under-supplied utility market may offset higher costs through much higher revenues. In this example, better price realizations, higher commodity prices, and low transportation costs improve the economics of a producer that under similar circumstances (i.e., same geology, same assets, but different location) may prove uneconomic.

Figure 16, below, speaks to the value drivers of the petroleum value chain. Note that from this perspective, the upstream, midstream, downstream, and chemicals divisions of fully integrated firms, such as Exxon Mobil, can be interpreted simply as sources of inexpensive supply for consumer retail sales of fuel and motor oil.

Figure 14: The Petroleum Value Chain

Source: Author

Economic Valuation Case Study
Due to the time value of money, timings of costs overwhelming influence economic value.

Ceteris paribus, projects with higher upfront costs will typically break even at lower petroleum prices, but also experience lower gearing to upside returns on higher prices. Furthermore, projects with long-lead times before production commences will typically break-even at even lower petroleum prices but will require longer periods to recoup the initial investment (i.e., payback period) and incur risks that economic assumptions under projects were originally sanctioned will prove to be overly optimistic. For this reason, projects with short payback periods tend to be considerably less risky.

To demonstrate this tautology, consider two petroleum development projects which have the following cash flows:

Figure 15: Cash Flows of Hypothetical Petroleum Development Projects
Upstream Oil and Gas - Comparable Project Valuation Case Study
Source: Author

If we assume that both projects have identical production rates and revenues streams, and have 10 year project lives, then the only difference between these projects are the costs. Project A has high front-loaded capital costs, but relatively low operating costs. Project B has modest upfront capital requirements but relatively higher operating costs.

The economics of these projects are summarized as such:

Table 2: Economic Summary of Hypothetical Petroleum Development Projects

Project A Project B
Capital Costs $5,000 $1,000
Net Cash Flow
Undiscounted $5,000 $4,000
Discounted @ 5% $3,458 $3,281
Discounted @ 10% $2,285 $2,724
Discounted @ 15% $1,371 $2,282
Break-even Price
Undiscounted $42.32 $43.86
IRR 26% 102%
Pay-back Period (years) 2.31 0.76
Recycle Ratio 2 5
DPIR (Discounted at 10%) 45.7% 172.4%

Source: Author

While Project A incurs higher upfront costs, it yields more undiscounted cash flows and breaks even at lower prices than Project B. On the other hand, Project B’s internal rate of return (IRR) is significantly higher and its pay-back period is significantly shorter. Project B’s recycle ratio (the quotient of per unit of production gross profit (i.e., netback) divided by the per unit of resource capital (i.e., F&D) cost) is also more favorable.

Which project is more economic? Academically, the right answer is the one with highest Net Present Value (NPV), but that metric often ignores price paid, capital capacity, and marginal return on additional capital employed. Moreover, all discounted value metrics presume that we already know the “right” discount rate(s).

Discounted profit to investment ratio (DPIR) is a more robust metric since it considers net present value received for every dollar of present value invested — it is basically the recycle ratio adjusted to discount the values of future cash flows.3 Assuming the finance and reinvestment rates are equivalent, the more commonly used but also more complicated modified internal rate of return (mIRR) yields the same relative rankings as DPIR.

An important but often overlooked consideration for calculating valuation ratios is pairing the correct denominator (i.e., type of invested capital) with a given numerator (i.e., type of cash flow) — what qualifies as invested capital depends on the category of money which flows into it, which in turns depends on the type of economic interest in question. When management estimates incremental returns on new projects, invested capital includes future capital expenditures. For investors, full-cycle invested capital might also include purchase prices of any financial instruments (e.g., shares of common stock) required in order to participate in economic interests.

Traditional discounted valuation analysis also disregards capital capacity, defined as the amount of capital which can be deployed at a given rate of return without significantly altering its economics. In the example, if constrained to a single instance of Project B, the answer depends on individual preferences regarding the timing of cash of flows. But since Project B results in higher returns on invested capital (DPIR), if we could invest in five Project Bs to equal the upfront capital investment of Project A, Project B would overwhelmingly produce greater value under any feasible discount rate (i.e., any discount rate greater than 0%).

Capacity constraints may not seem overly important for most individual investors, but they are acutely relevant to mineral resource lessors and owners. Most investors are capital constrained vice capacity constrained (i.e., most investors cannot deploy enough dollars to significantly impact the underlying economics). When capacity is not a limiting factor, valuation metrics scale linearly with capital employed. However, in reality, limited amounts of capital can be deployed at a given rate of return — this is especially true of limited and depleting resources. When potential low cost of supply areas have not yet been de-risked (i.e., a reduction of geological uncertainty related to recovery as a result of drill-bit operations) or when they are expected to dry up relatively quickly, lower return but less risky and/or larger scale/capacity capital investments will often be perceived as having greater value.

Even under the unrealistic assumption that perfect certainty can be had about future cash flows, the right metric for economic utility depends on individual preferences on timing, the types of economic interest involved, and constraints on capacity.

And all that presumes we already know the correct rate at which to discount valuations. Judging from industry norms, it seems likely that many valuation professionals do not fully grasp the underlying significance of the time of value of money principle.

Discount Rates
To simplify preferences on timing, investors select a rate by which to discount future money flows. In the conventional stream of thought, discount rates consist of a risk-free rate of return (i.e., inflation expectations) and expectations for future growth and return. It is also common practice, in this industry and in others to adjust the discount rate for risk. However, as Warren Buffett points out, “You can’t compensate for risk by using a high discount rate.”

The industry standard discount rate of 10% mirrors guidance from the Securities Exchange Commission (SEC) and Federal Accounting Standards Board (FASB) that requires public companies to disclose fair value estimates of petroleum resources within annual 10-K or 20-F reports. Present value calculations prepared using a 10% discount rate are informally known as “PV-10”. Banks typically discount at 9%. The June 2014 SPEE survey indicates that applied discount rates range from 4.96% to 29.24%.

However, empirical research on long-term asset returns and rational investor expectations prescribes the use of lower and more consistent rates. Compellingly, in 2016, the industry WACC fell between 3 to 4%. Moreover, nowadays one can rarely identity equity of any oil and gas producer which trades below its economic net asset value when a 10% discount rate and full-cycle costs are used. For investors who subscribe to “full cost disclosure” thesis, necessity dictates they lower their expectations in order to justify any capital allocation.

Figure 16: Weighted Average Cost of Capital (WACC) for Publicly Traded Oil and Gas Companies, 1999-2016

Source: Portfolio123; author’s calculations

Note to Figure 17: Industry aggregates consider all U.S. publicly traded companies assigned to primary GICS Codes 10102010 (Integrated Oil & Gas) and 10102020 (Oil & Gas Exploration & Production). The weighted average cost of capital is the sum of all quarterly cash dividends and interest payments divided by the sum of all invested capital (i.e., book values of equity plus debt and non-controlling interests).  

If all this focus on the discount rate seems superfluous, it is because it is. But that’s the point: spending more energy on reflecting “all the costs” and decreasing uncertainty in the numerator (i.e., the amounts, timings, and likelihoods of discrete cash flows) greatly decreases the need to adjust for risk in the denominator (i.e., discount rate). The differentiation between numerator and denominator is the fundamental precept of applying Buffett-esque margin of safety.

For a more detailed discussion on discount rates, reference Appendix C.

Advanced Economic Modeling
Straight discounted cash flow analysis are limited in their ability to distill one, correct answer regarding economic value since they assume future market conditions are known, do not account for optionality, may not explicitly define the estimated likelihood of a given set of cash flows, and do not account for the benefits of diversification. Different modeling techniques can help deal with these shortcomings.

Models of discounted future cash flows typically assume that future commodity prices are known (or constant). This is clearly a weak assumption, however, since all the evidence points to a contrary fact: that commodity price evolutions are virtually indistinguishable from a stochastic process (i.e., “random walk”).

Vanilla options models, as simple as either binomial trees or Black-Scholes, can help deal with stochasticity of underlying revenue/cost drivers. By explicitly assigning a probabilistic likelihood to a given estimate for future cash flows based on the expectations of commodity price behaviors, vanilla options models circumvent (or, at least, explicate) the most pressing problem associated with implementations of straight discounted cash flow analyses.

Real options analysis can help managers and investors accounts for managerial flexibility. Unlike vanilla options, real options are tied to actual decisions open to managers at different points during the project’s life. These decisions tend to enhance asset value since more information becomes known over time which can increase upside and/or limit downside.

The nature of the upstream business, in which internal business decisions are largely reactive to external market forces (i.e., commodities prices), lends itself to valuation by a synergistic combination of vanilla and real options analysis. If an underlying revenue stream is characterized by uncertainty (revenues of upstream assets are), financial options can model commodity price risks whereas real options can represent the optionality embedded within business decisions and/or non-systemic risks4.

Real option valuation analysis is especially useful in evaluating distressed and marginal assets. Due to boundary and terminal conditions on contingent pay-offs5, straight discounted cash flow analyses will tend to assign less value to a stream of risky cash flows relative to a comparable options analyses. As a rule of thumb, discounted cash flow analysis will undervalue non-marginal upstream assets by around 20-40%. The disparity between valuation methods increases asymptotically for marginal projects and investments (i.e., those which have are estimated to have depressed and/or null values under flat price assumptions). This intuition runs counter to conventional wisdom that discounted cash flow analyses tend to overvalue equity. Yet, it is backed by anecdote: many upstream projects are not likely to produce positive cash flows, but yet they are often highly valued for their upside potentials. As with anything else, the quality of an assessment on value is a function of the underlying data quality. Garbage in — garbage out.

A simple way to think about options is this: when a payout which is contingent upon an underlying process which is stochastic (i.e., commodities prices), you would not evaluate the derivative function at only one given price for the underlying. Rather, you would evaluate the range of outcomes and their respective likelihoods to find an expected payout. Why is it common, then, to evaluate natural resource values according to a single price deck?

Figure 17: Simplified Real Options Analysis Framework for Upstream Oil and Gas

Source: author

The Winter 2003 edition of Oilfield Review more completely articulates the arguments for utilizing real options for valuing oil and gas investments6.

Through the lens of modern portfolio theory (MPT), the quest for the right metric is also shaped by the intuition that diversification preserves returns while reducing risk. This intuition holds as long as returns of distinct investments/projects are uncorrelated7. Still, MPT is based on singular linear regression which assume returns are commensurate with risk. Furthermore, most interpretations of MPT equate risk with asset price standard deviation (i.e., financial volatility). These assumptions are contradicted by empirical equity returns in which low volatility and low beta portfolios outperform. Even Post-Modern (Post-Mortem?) Portfolio Theory, which incorporates multiple return factors, suffers from fallacies of hindsight and assumed linearity.

Bet sizing according to the Kelly criterion (“fortune’s formula”) is generally superior to MPT for allocating capital in way that optimizes long-term expected returns without artificially differentiating between generating returns and managing risk. Optimal Kelly betting strategies minimize the expected number of bets required to double the bankroll. Fractional Kelly betting, however, is often prescribed to minimize the risk of ruin while preserving proportionally greater expected returns. For example, half Kelly decreases the bankroll volatility by 50% and expected growth by only 25%8. The Kelly criterion is fairly simple to calculate for a two-state probabilistic function. For a real-world portfolio with continuous and stochastic pay-offs, the arithmetic expected return over-states the long-run expected return of a risky payoff. This observation leads to convergence between Kelly betting and stochastic portfolio theory (SPT) under the special case that long-term expected returns are optimized with respect to the logarithmic utility function. Edward Thorp is the foremost pioneer in employing the Kelly criterion to capital markets; he is also an editor of the definitive treatise on fortune’s formula: The Kelly Capital Growth Investment Criterion.

The modeling techniques which one employs should foremost be commensurate with the modeler’s proficiency level. Modelers should also consider the desired end-state, time constraints, the quality of assumptions and data, and the accuracy expected of the end result. As Aristotle correctly pointed out 2400 years ago, the wise man demands no more precision than the nature of a given subject permits. Restated in Keynesian terms, “it is better to be approximately right than precisely wrong”. In any case, the more layers we heap up in an attempt to distill one correct solution, the more elusive it becomes. Metrics should be straightforward; models, not necessarily.

The sequence in which information has been presented thus far throughout the Drilling for Value series is admittedly somewhat haphazard. Yet, my intent has been to instill a “fundamentals first” mentality in order to better conceptualize how the pieces fit together. Future appendices will dive deep into more domain specific topics (e.g., accounting practices; resource estimation and production modeling, corporate finance, and applied valuation). But in case I’ve missed anything important, here are some online resources which address some of these topics:

Footnotes   [ + ]

1. Yanbo Jin, and Phllipe Jorion. Firm Value and Hedging: Evidence from U.S. Oil and Gas Producers. December 2004
2. EIA’s assessment was based on an original report by Advanced Resources International (ARI) for EIA
3. DPIR is also the valuation and/or performance metric of choice for Raymond James’ E&P analysts.
4. Long equity is analogous to a long call option on asset value, whereas long debt is analogous to a short put on asset value. The rule of put-call parity supports the analogy: a long call combined with a short put is a synthetic long position in the underlying security. Likewise, being long both the equity and debt of a company is a long position on the firm’s total asset value. Boundary conditions of call options implies that, all things equal, an increase in volatility of the underlying asset values apportions more value to equity relative to debt. How can making something more risky increase its value? The solution is usually because canonical probability distributions used in models do not reflect our internal utility functions which tends to: a) discount hyperbolically with respect to time; and, b) discount negative outcomes moreso than what is implied by symmetrical probability distributions. There is also the fundamental problem of assuming that the distribution of outcomes is known “normal” and/or even known.
5. Option values are essentially functions of expected payouts. These payouts are usually represented as boundary conditions whereby losses are limited to investment principal and whereby upside is unconstrained. As applied to equity valuation, boundaries represent two fundamental properties: 1) assets and/or equity cannot have negative values; and 2) projects will terminate when they are no longer economic
6. William Bailey; Benoît Couët; et al. Unlocking the Value of Real Options. Oilfield Review. Winter 2003.
7. C.R.K. Moore advocates the use of risk-adjusted values (RAV) as a means to adjust asset values within a portfolio to account for the benefits of diversification. While the RAV method is shaped by similar intuitions as MPT, it is not widely used by industry professionals.
8. Not shockingly, the best explanations of Kelly betting are found on gambling sites, such as Wizard of Odds.