The Interplay of Fundamentals and Financials in NYMEX Crude Oil Price DeterminationUSAEE/IAEE North America ConferenceOctober 12, 2011Washington, D.C.William H. Brown IIIPresidentWHB Energy Research LLCDavid KnappEnergy Intelligence Group
Outline of Presentation • Overview • Conclusions • Methods of Analysis Tactical Strategic Fundamental Financial • Results • Price components in yet another “recovery” year: 2012 Broad fundamental view Non-commercial influence Seasonality • Decipher all factors into a price perspective for an uncertain 2012
Overview • There has been substantial debate over the last few years with regard to the role of financial players, i.e. so-called “speculators” or non-commercials, in determining the path of crude oil prices since 2004, and how these players have or have not distorted what many believe to be strong underlying trends in the fundamental, or physical, oil market. • We believe 2004 marks a watershed in this regard because in late 2003 a few major and influential pension funds were searching for ways to enhance returns, given the mediocre performance of fixed income portfolios. • These pension funds recognized that in the preceding years, Commodity Trading Advisors (CTAs) had realized attractive returns, and thus sought to diversify their portfolios by allocating incremental capital to commodities in general and oil in particular. • In contrast to “traditional” CTAs who had the option of being either long or short and thus could capitalize on market moves in either direction, however, most pension funds had to be long, and thus began to have a “one way” influence on oil prices as they increased their capital allocation to oil, to be joined progressively over the years since 2004 by hedge funds, other financial institutions, and individuals.
Overview (cont’d.) • We have studied the role of various players in both in the physical and paper markets on both a micro and macro level over the last few years. Given what we believe to be a “modified” profile of participants in the paper futures markets, understanding the factors that determine NYMEX crude oil prices is critical from the standpoint of price forecasting, identifying producer hedging opportunities, and U.S. energy policy. • We wish to emphasize at the outset that we regard non-commercial participation as neither “good” nor “bad.” We view it as a fact of life and one component of the equation, and do not believe that “speculation is evil” or that funds “have no right to hit a home run”. • We have believed ever since the 1980s that commodities represent a legitimate alternative asset class, but on both the long and short side. The fund influence on oil prices has played a major role in rebalancing demand with supply in an industry that is not characterized by discrete, annual, and predictable price increases. • Finally,we believe that legislation and/or rule making will have little or no effect on crude oil prices.
Conclusions • From both a qualitative and quantitative standpoint, both fundamental and financial factors have played significant roles in NYMEX crude oil price determination over the last few years, but neither group of factors is perfectly consistent and continuous in its “explanatory power”. • Rather, it is the interplayof both sides of the equation over time with varying degrees of influence, either independently or together, during discrete time periods that link to form a continuous chain, ultimately yielding the actual path of NYMEX crude oil prices. • We would note that as NYMEX/Brent crude oil prices exceed the upper $60s per barrel, our estimate of the underlying cost of marginal resource development, the larger the periodic, but once again not continuous, influence of financial factors. • On balance, we believe it is counterproductive to “blame” exclusively either one or the other side of the equation, particularly in the context of a debate on the appropriate path of U.S. energy policy.
Methods • Our approach to the problem, of necessity, was both qualitative and quantitative in nature. • From a qualitative standpoint, we are “in the market” on a daily basis on behalf of our clients, tracking NYMEX and ICE price movements and the factors that influence them hour by hour, which affords us a “tactical” perspective on price. • From a quantitative standpoint, the broad fundamental framework includes our detailed macro analyses of the international and U.S. petroleum industry which affords us an overall “strategic” perspective and a fundamental view of price. Our “strategic” perspective also includes a broad analysis of financial elements and players in the price equation. • A good example of our “tactical” perspective which strikes us as adding merit to our arguments and yet admittedly has no ability to be quantitatively proven, is an early month pattern we saw developing in 2010 illustrated as follows:
The “Tactical” PerspectivePerceived Month-by-Month Capital Allocation Patterns in 2010 Cumulative Gain/Loss, First Three Trading Days of the Month Prompt NYMEX CL Settlements Jan $3.82 Feb $4.09 Mar $1.21 Apr $3.08 May -$6.18 Jun $0.64 Jul -$3.65 Aug $3.52 Sep $2.68 Oct $2.85 Nov $3.26 Dec $5.08 For 2010 as a whole, the “average peak” in the first three trading days of the month was $81.25 per barrel. The average for remaining trading days of each month was $79.51 per barrel. Therefore, on average the difference was an early-month peak that was higher than the subsequent average by $1.74 per barrel, suggesting that early-month asset allocation played a role. A similar pattern often occurred in the equity market. Net Investment: Broad Economic Optimism Net “Disinvestment”: Euro Crisis and “Double Dip” Fever Net Investment: Ben’s QE II Discourse and Renewed Economic Optimism
The “Strategic” Perspective:Fundamentals • We have considered a variety of factors based on our detailed models of world oil demand and supply and U.S. refinery balances, the ultimate price “bellwether” of which would be the global and/or U.S. “discretionary” crude oil inventory position as measure in days supply. • In addition, taking into account the rising impact of East of Suez, primarily China and India, oil demand, we have attempted to quantify the price impact of what we would term the incremental “China pull” of light, sweet crudes out of their natural market, the Atlantic Basin. • Also, we have examined fundamental, seasonal patterns that manifest in crude oil prices within any given year which often “overlay” the underlying trends in the global oil picture.
The “Strategic” Perspective:Financials • With regard to financial factors, we have looked at a number of variables against NYMEX crude oil prices. • These variables include: The categories of the weekly CFTC Commitments of Traders Report Changes and/or expectations of changes in Fed policy The equity market as represented by the S&P 500 The dollar/euro rate • Of course, these factors are not mutually exclusive. If hypothetically there is a close relationship between oil prices and CFTC fund position data, those positions are established for a reason, e.g. Fed policy, positive/negative economic expectations, etc. which implies in turn a close relationship between oil prices and those catalysts for fund positions.
Results:Fundamental Relationship Between Price and Inventory Days Supply • Our research has concluded that from both a qualitative and quantitative standpoint, the global and U.S. fundamental picture plays a substantial role in crude oil price determination, as it should. • Prior to 2004, the most relevant variable in terms of the relationship to NYMEX crude oil prices was the days supply of U.S. crude oil as calculated from the inventory and refinery run data reported by the Energy Information Administration. The U.S. stock profile was more relevant than the global inventory position since the U.S. data are more timely and visible. The relationship also yielded intuitively reasonable results when considering commodity price behavior. That is, although not always perfect, generally speaking as the U.S. days supply of crude oil increased prices declined, and vice versa. • Although in theory one would look at inventory at Cushing, Oklahoma versus the NYMEX crude oil contract, we have found that the market responds just as much or more to changes in the total U.S. inventory picture. • However, attempting to quantify the relationship between U.S. days supply of inventory and NYMEX crude oil prices yields relatively poor results in the period post 2004 on average.Granted, since 2004 oil demand outside the OECD has provided the bulk of growth, with China, for example, “pulling” light, sweet crudes out of their natural Atlantic Basin markets, as cited above. • Also, one could make the argument that rising OPEC capacity utilization commands, by definition, a “security premium.” We believe, however, the abandonment of any intuitively reasonable relationship with the most visible and “data-timely” crude oil market in the world suggests additional, non-fundamental factors at play.
Results:An Historical Perspective U.S. Crude Oil Days Supply vs. Prompt NYMEX Crude Oil Settlements Annual Averages Days Supply Price 2000 19.4 $30.26 2001 20.3 $26.04 2002 20.6 $26.16 2003 18.4 $31.12 2004 18.8 $41.43 2005 20.9 $56.71 2006 21.9 $66.18 2007 21.6 $72.42 2008 21.0 $99.73 2009 24.2 $62.08 2010 24.2 $79.60 2011 to Date 24.3 $96.85 Intuitively Reasonable Relationships De-Coupling of Price From “Visible” U.S. Supply
Results:Financial Players and Variables • Our correlation analyses for 2010 and 2011 (as well as previous years) did not yield consistently strong relationships between NYMEX crude oil prices and the financial variables previously itemized on a continuous basis for the entire time periods under consideration. • However, as with fundamentals as gauged by days supply of crude oil inventory, there were a number of discrete time periods within the course of a given year when the relationships between NYMEX crude oil prices and variables such as Fed expectations, the S&P 500, and the dollar were closely correlated. • The “financial manifestation” of the non-commercial influence on crude oil prices is of course subject to much debate. The CFTC and others had previously concluded that speculation had little or no influence on oil prices. However, we suspect that CFTC analysts were basing their conclusions on the analysis of an entireyear or series of years which yield weak results when performing a Granger causality analysis. • Also, once the CFTC finally began disaggregating their weekly Commitments of Traders Reports for these times, there were substantially better and more appropriate data to compare with price movement, yielding more intuitively “reasonable” results. • We wish to strongly emphasize at the outset that the CFTC data are an imperfect, but useful, tool to gauge the price influence. Although the CFTC breaks down the weekly Commitments of Traders Report into 4 categories, we have found the two most “explanatory” to be Managed Money and Swaps Dealer. • Of the two, Managed Money has been progressively and substantially more statistically significant over the course of an entire year and particularly, within a given year. Managed Money is defined as a registered commodity trading advisor (CTA), a commodity pool operator (CPO), or an unregistered fund identified by the CFTC (e.g. hedge fund).
Results:CFTC Data: Explanatory or Not? Correlation Between the Prompt NYMEX Crude Oil Settlement And The Net Change in Managed Money Position 2007 0.42 2008 0.41 2009 0.72 2010 0.75 2011 to Date 0.67 • The relatively higher correlation in more recent history, on average, suggests that either Managed Money is concurrently reacting to short-term fundamental developments (read demand) or is helping to cause prices to rise or fall. This obviously begs the question of correlation versus causality. To believe that Managed Money is reacting to fundamentals, however, perhaps presupposes a real-time, continuous, and near-perfect knowledge of fundamental developments in the physical market, which we believe is unlikely given anecdotal evidence of the profile of the average oil trader these days who has come into the market as commodities became a more popular asset class.
Results:CFTC Data: Explanatory or Not, 2010? • Once again we would make the point that “Managed Money” has relatively strong explanatory power regarding NYMEX crude oil, but generally during discrete periods within a given year, with 2010 a good example of periodic “coupling” and “decoupling” of the influence.
Results:CFTC Data: Explanatory or Not, 2011? • Our point regarding the influence during periodic, discrete periods can also be illustrated in tabular form for 2011 to date. As indicated, earlier in the year the implied price impact was reasonably consistent, but for the period in early to mid August was much larger, implying in turn a fundamental as well as financial influence on price change as the global oil balance became “looser”.
Results: Prompt NYMEX Crude Oil vs. Days SupplySuch a transition from financial to fundamental influence, though of course not mutually exclusive, was in evidence late last year, as we illustrate in our graph of 2010. In the context of our previous discussion of QE II, crude oil price movements were not exclusively a non-commercial response to liquidity and inflation expectations.
Results:NYMEX Crude Oil Prices and the Equity Market • Aside from the influence of monetary policy, another observable, and more consistent financial influence on crude oil prices has been the S&P 500 as the barometer of economic growth. Although there are more energy consumers than producers in the S&P 500, the tendency toward highly correlated trades across asset classes due to the influence of hedge funds has inverted the relationship to one of a positive correlation. Not to oversimplify, but we believe, based on our anecdotal and statistical evidence, that funds in general function according to the following “equation”: Economy = Oil Demand = Oil Prices • That is, funds tend tolook only at oil demand and not the entire fundamental balance as embodied by the net global inventory position. The tendency toward this approach can be illustrated in the data below for the years 2008-2010.
Results:Prompt NYMEX Crude Oil vs. S&P 500 Correlation, S&P 500 Close vs. Prompt NYMEX Settlements 2004 .00 2005 .25 2006 .40 2007 .15 2008 .66 2009 .76 2010 .76 2011 to Date .41
Results:Prompt NYMEX Crude Oil vs. The Equity Market, 2010The correlation was “close enough for government work”, given the comparison of a 4:00 equity market close with a 2:30 NYMEX settlement
Results:The 2011 Falloff in Correlation • Of note is the falloff in the correlation in 2011 to date, and we illustrate this graphically below. We first would note that the correlation between the S&P 500 and Brent in 2011 to date is just as weak. Thus far this year there is a “breakaway” of crude oil prices from the equity market because of supply issues that for a good part of the year have overshadowed “demand discounting”.
Results:NYMEX Crude Oil Prices vs. the Dollar • Another financial influence on crude oil prices often cited by the general press is the dollar. • These days, few arguments are offered to defend this “trade”, merely periodically cited by the media as a presumed influence, but “originally” the idea was that as a dollar-denominated commodity, as the dollar declined it cheapened oil costs to non-dollar economies, thus stimulating demand. • Another argument was that as the dollar declines OPEC loses purchasing power, and thus the market discounted the likelihood that OPEC would take steps to increase prices in dollar terms. • A third argument was that loose money weakens the dollar, creating inflationary expectations, leading in turn to buying oil as a hedge against the very inflation that higher oil prices are creating. • In our view, based on anecdotal evidence via discussions with hedge funds, the inverse relationship between crude oil prices and the dollar was a self-fulfilling trade that “worked”, pure and simple. • As the original arguments emerged once again, some funds initially embraced the idea, leading to an increasingly favorable technical picture which induced more funds into the trade. It worked until it didn’t work, if you will, as illustrated by the following data. r
Results:NYMEX Crude Oil Prices vs. the Dollar Correlations, Dollar/Euro Rate vs. Prompt NYMEX Crude Oil Settlements 2004.04 2005 .43 2006 .02 2007 .93 2008 .73 2009 .87 2010 .22 2010, Apr-Nov. .60 2011 to Date .36 • We show both 2010 in its entirety and the period Apr-Nov of 2010 to demonstrate a good example of “on again, off again” correlation related specifically to the “first” Greece crisis and the market discounting QE II as we illustrate:
Results:A Return to Fundamentals: Seasonality • With very few exceptions, crude oil prices tend to rise in the second quarter from a first quarter trough. • Factors include: In Q1, lower refiner demand for crude in Atlantic Basin, U.S. stocks recover In Q1, market discounts end of winter Consensus invariably “mis-correlates” the first to second quarter downturn in global refined product demand with crude oil prices In Q2, refiner crude oil demand recovers In Q2, North Sea production begins to fall seasonally In Q2, funds buy gasoline, always believing the U.S. gasoline season will be strong • The exceptions to this pattern tend to lie with global macro developments that more than offset the seasonal first-to-second quarter price recovery. In the case of 2010, the seasonal pattern was well defined until the Greece crisis in early May. Summertime “double–dip” fever led to lower prices than those prevailing in the first quarter, also an exception. In 2011, the seasonal pattern began well-defined, compounded by “Arab Spring” and the Libya revolution. Thereafter, once again “double dip” fever has led to lower than first quarter prices at present.
Results:Fundamental Crude Oil Price Seasonality Monthly Crude Oil Price Seasonality Since 2004: Year=1.00, Except 2011
Price Components in Yet Another “Recovery” Year2012:Summarizing the Components 2012 NYMEX crude oil prices will once again be determined by a combination of fundamental and financial variables: • Fundamental factors: -Growth in world oil demand -Growth in non-OPEC supply -Geopolitical influences on supply -Global days supply of crude oil and product inventories influenced by: -Excess refining capacity -Excess producing capacity, primarily Arabian Gulf sour grades • Financial factors: -Need/desire by financial institutions to diversify invested assets. However, by the end of 2011 we believe there is a good chance funds will begin to rethink their long-term commitment to a long-only oil strategy, particularly v. equities -Global equity market strength/weakness -Value of the dollar, although only periodically a factor since this argument is starting to get “old” -Monetary policy expectations
Price Components in Yet Another “Recovery” Year2012:Adding It All Up • Fundamental factors are “straightforward”, but there is still a critical need to appreciate the importance of the financial factors and the “non-commercial” price influence. • This is particularly critical for producers (hedge timing) and equity portfolio managers. The non-commercial influence implies a greater volatility and velocity of crude oil price movement than if prices were determined solely by fundamentals. • Ever since 2004, the non-commercial influence on crude oil prices has been increasing, and has had a major impact, even during the global recession. We have tried various approaches to quantify the impact for the last four years. In our view, those who argued that fundamentals were primarily, if not exclusively, responsible for the price run up because there was little consistent, public, and quantitative “proof” of the non-commercial influence were guilty of the logical fallacy: Argumentum ad ignorantiam i.e. asserting that a proposition is necessarily false because it has not been proven true
Price Components in Yet Another “Recovery” Year2012: Conclusions • The overall fundamental global balance should remain competitive in 2012 with excess producing and refining capacity suggesting lower prices than $80.00 per barrel, basis Brent. • However, funds and other non-commercials may be intent upon allocating incremental capital to oil next year, if risk of global recession wanes. • Thus, a higher price “bar” will be established from which prices will rise seasonally as we move through the first half of next year from a first quarter trough. • Price elasticity of demand remains alive and well, however, suggesting that crude oil price strength in 2012 will extend only through summer. In the fourth quarter, concerns over the impact of higher oil prices on 2013 economic activity once again will lead to fund net length liquidation back toward, but not to, underlying fundamental value. Funds will limit their net length liquidation in order to retain some exposure to the possibility of generating “excess” portfolio returns. . • We are therefore suggesting a continued crude oil price cycle of mini “boom and bust” for the foreseeable future.