AIAR Vol 2 Issue 4 March 24 Final_0

8/9/2019 AIAR Vol 2 Issue 4 March 24 Final_0

1/73

Alternative

Investment

AnalystReview

WHAT A CAIA MEMBER SHOULD KNOWThe Case for Investing in Non-Traded REITs

Sameer Jain

RESEARCH REVIEWOil Price Drivers and Movements: The Challenge for Future Research

Hillard Huntington, Saud M. Al-Fattah, Zhuo Huang, Michael Gucwa, and Ali Nouria

CAIA MEMBER CONTRIBUTIONImpact of the Recent Regulatory Changes on the UCITS CTA MarketLouis Zanolin, CAIA

TRADING STRATEGIESStatistical Arbitrage and High-Frequency Data with an Application to Eurostoxx 50 EquitiesChristian Dunis, Gianluigi Giorgioni, Jason Laws, Jozef Rudy

PERSPECTIVESThe Resiliency of the U.S. Futures Industry

Hilary Till

CAIA MOMENTUM MONITORCAIA Momentum Monitor Alexander Ineichen, CAIA, CFA, FRM

Chartered Alternative Investment Analyst Assoc2014, Volume 2, Issue 4

Image Here


2/73

CEO’s Letter

As the calendar turns to 2014, we are reminded that the only constant in life is change. That is certainly true in the

alternative investment space where new products, wrappers, and a multitude of economic variables continue to

underscore the need for a perpetual commitment to education. The Alternative Investment Analyst Review wi

always seek to play an important role in this quest for our Members and their clients.

This is also a year of change at the top of the CAIA house. Florence Lombard, who has been a founder, board

member, and leader of CAIA since our humble beginnings in 2002, announced her retirement over a year ago.

Practicing what we preach, the search process was thoughtful and thorough, with a lean toward nding a successor

who could bring some “uncorrelated” skills into the CAIA fold. I hope to live up to that credo in everything I do.

thank Florence for her leadership, vision and voice; our organization and our industry are in a better place as a resul

of those efforts.

In the coming months and quarters, I hope to get out and meet many of you at local chapter events or through ouafliations with our association and academic partners. I will promise to never lose sight of the fact that I work for you

our members, and your experience and expectations must shape, focus and evolve our mission to be a leader in

alternative investment education.

The current issue of AIAR contains a little something for everyone. There are ve articles covering a very broad range

of topics. In the rst article we cover non-traded REITs as an alternative asset class and the characteristics of its

distinct risk-return prole. This is followed by a broad review of the current economic research on the structure and

function of the world oil market. Our regulators always remain top-of-mind, and our third article updates us on the

impact of recent regulatory changes on UCITS CTAs. We also take a closer look at the US futures industry as a variety

of participants go there seeking to manage a multitude of risk. Staying with that risk theme, the nal section of this

issue discusses a “Momentum Monitor”, which is a tool to assist risk takers and investment managers in their overall risk

management process. The Momentum Monitor is produced by Alex Ineichen, CAIA, and will be a regular feature o

AIAR going forward.

I wish all of you a healthy and prosperous 2014. As always, we encourage your feedback, suggestions for future

content, and direct submissions from our Members. Your opinions matter to us.

Bill Kelly, CEO

Chartered Alternative Investment Analyst Association


3/73

Call for Articles

Article submissions for future issues of Alternative Investment Analyst Review are always welcome. Articles

should be approximately 15 pages, single-spaced, and cover a topic of interest to CAIA members. Additiona

information on submissions can be found at the end of this issue. Please email your submission or any questions

to [email protected].

Chosen pieces will be featured in future issues of AIAR, archived on CAIA.org, and promoted throughout the

CAIA community.

http://[email protected]/http://caia.org/http://caia.org/http://[email protected]/


4/73

Rise above the crowd with the CAIA Charter

The CAIA Charter offers you immediate credibility in

the complex world of alternative investing, along with

access to a global network of peers. Find out how you

can earn a better rate of return on your educational

investment at CAIA.org.


5/73

Table of Contents

What a CAIA Member Should KnowThe Case for Investing in Non-Traded REITs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

By Sameer Jain

ABSTRACT: Non-traded REITs (NTRs), like their publicly traded counterparts, makeinvestments in income-producing commercial real estate. They typically holdmultiple properties in a single portfolio and are ordinarily categorized by the typesof properties they own, such as retail, industrial, multifamily, ofce, and storage,among others. With interest rates at rock-bottom levels, safe yet high-yielding

assets remain scarce. Investing in NTRs offers a number of advantages; this noteexplores key aspects of NTRs and their contributions to a portfolio of assets.

Research Review

Oil Price Drivers and Movements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11By Hillard Huntington, Saud M. Al-Fattah, Zhuo Huang, Michael Gucwa, and Ali Nouria

ABSTRACT: This paper briey reviews what is currently known about fundamentalfactors inuencing the oil market and what the apparent research gaps are from

the perspective of a major oil exporting country. The discussion also includes variousbroad modeling approaches for representing these factors. It begins with factors

that may inuence the long-run world oil price path and then shifts to factors thatmay inuence the short-run price path. The paper also highlights the role of the

world’s largest oil exporter, Saudi Arabia, and its inuence in the oil market.

CAIA Member ContributionImpact of the Recent Regulatory Changes on the UCITS CTA Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

By Louis Zanolin, CAIA

ABSTRACT: This article provides an overview of the UCITS CTA market and explains

the consequences of a consultation paper published by the European Securitiesand Market Authority (ESMA) proposing changes in the way CTA strategies can

be replicated. We explain the changes in regulation, assessing the consequencesfor managers using an index structure and we examine various options available

in order to comply with the proposed new regulatory framework.

Trading Strategies

Statistical Arbitrage and High-Frequency Data with an Application toEurostoxx 50 Equities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

By Christian Dunis, Gianluigi Giorgioni, Jason Laws, Jozef Rudy

ABSTRACT: The motivation for this paper is to apply the statistical arbitrage

technique of pairs trading to high-frequency equity data and compare itsprot potential to the standard sampling frequency of daily closing prices. We

use a simple trading strategy to evaluate the prot potential and compare theinformation ratios yielded by each of the different data sampling frequencies.

The frequencies observed range from a 5-minute interval to a daily interval.

AIAR STAFFHossein KazemiKeith Black Edward Szado

Editors

Barbara J. Mack

Content Director

Lynne Feldman

Marketing

Jason Vachula

Creative and Design

U.S.

+1 413 253 7373

Hong Kong

+852 3655 0568

Singapore+65 6536 4241

Table of Contents

CAIA.org

https://caia.org/http://www.linkedin.com/groups/CAIA-Association-Chartered-Alternative-Investment-39863https://twitter.com/caiaassociationhttps://www.facebook.com/pages/CAIA-Association/126663167435874https://caia.org/


6/73

The analysis reveals that the extent to which daily data are co-integrated provides a good indicator of the

protability of the high-frequency pairs trading. For each series, the in-sample information ratio is a goodindicator of the future protability as well.

The results show that a statistical arbitrage opportunity exists when applying a novel diversied pair tradingstrategy to high-frequency data. In particular, even once very conservative transaction costs are taken

into account, the suggested trading portfolio achieves attractive information ratios (e.g., above 3 for anaverage pair sampled at the high-frequency interval and above 1 for a daily sampling frequency).

Perspectives

The Resiliency of the U.S. Futures Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59By Hilary Till

ABSTRACT: This article demonstrates how U.S. futures markets were “forged by … [decades of] of trial anderror [efforts] … and … were not the product of some designing intelligence.” These markets “represent a

distillation of what human experience has found” to work. In an era of optimistic proposals to fundamentallyredesign market structures, it may be useful for participants of mature nancial centers to be reminded of

these historical facts. In addition, participants from new emerging nancial centers, who are “crossing theriver by feeling the stones,” may also nd it useful to understand how U.S. futures markets evolved in a trial-and-error fashion, overcoming adversity through innovation each step of the way.

In demonstrating the trial-and-error development of futures markets, this article will briey cover how and

why modern futures trading started in Chicago; how the Chicago and New York futures exchanges havehad to constantly innovate in order to remain in business, including with speculative product launches,

demutualization, and the development of electronic trading; and why some futures contracts fail. Afterreviewing the history of U.S. futures markets, one does get a sense of the resiliency of these institutions, inconstantly responding to adversity, from their earliest days through well into the present.

Momentum Monitor Momentum Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

By Alexander Ineichen, CAIA, CFA, FRM

ABSTRACT: Risk is often dened as exposure to change. Spotting change, therefore, is important. There areessentially three approaches to change: 1. Displaying complete ignorance, 2. Having a wild guess as towhat it means, or 3. Measuring it in a systematic fashion with an applicable methodology and adapting to

it. The author recommends choice number 3.

Momentum can be perceived as a philosophy. The author discusses the Momentum Monitor (MOM) and

recommends it as a risk management tool. If risk is dened as “exposure to change,” then one ought tospot the change.


7/73


8/737Alternative Investment Analyst Review The Case for Investing in Non-Traded REITs

Te Case for Investing in

Non-raded REIsSameer JainChief Economist and Managing Director, American Realty Capital

Research Review

CAIA Member ContributionCAIA Member ContributionWhat a CAIA Member Should Know


9/738Alternative Investment Analyst Review

1. Introduction

Non-traded REIs (NRs), like their publicly tradedcounterparts, make investments in income-producingcommercial real estate. Tey typically hold multipleproperties in a single portolio and are ordinarily cat-egorized by the types o properties they own, such asretail, industrial, multiamily, office, and storage, amongothers. Certain o these REIs structure their under-

lying investments as net leases in which the tenant isresponsible or bearing real estate costs directly, suchas property taxes, insurance, operating expenses, andcapital items, in addition to rent and utility payments.By distributing essentially 100% o all rent payments re-ceived by tenants, these REIs provide a durable streamo monthly income to their investors and also offer thepotential or long-term capital appreciation throughproperty value growth.

Te term private REI commonly attached to NRs is amisnomer, as these REIs are publicly registered and arethus subject to many o the same public reporting, taxqualification, SEC regulation, and governance require-ments that an exchange-traded public REI must meet.Unlike traded public REIs, these vehicles are not sus-ceptible to exchange-traded supply and demand-drivenprice volatility. Rather, price discovery happens throughperiodic valuations, much as in private equity real estateinvesting. NRs generally provide little or no interimliquidity and usually have a five to seven-year liespan.

However, unlike public REIs, NRs are only availableto investors that meet established net worth and incomestandards. With interest rates at rock-bottom levels,sae yet high-yielding assets remain scarce. Investing inNR offers these advantages:

• Potential or superior risk-adjusted returns.• Higher dividends than traded REIs.• Avoiding the potential inflated valuations in the

traded REI sector.• Illiquidity premium that can be captured by the

long-term investor.• Ability to raise and invest capital at opportune times

in the market cycle.• Valuations that are not subject to public market

volatility.

NRs have started to attract considerable investor in-terest. We estimate that around 70 NRs have raisedand invested $80 billion to $90 billion in equity duringthe past decade, with about $10 billion raised and de-

ployed annually in recent years. Te industry is heavilyconcentrated, with the top 10 NRs controlling a domi-nant share.

2. Determining ValuePublicly-held companies typically trade at a multiple oprice to earnings (or in the case o REIs, price to undsrom operations). Unlike traded REIs, where value is

tied to the price at which shares trade on an exchangeand is ofen influenced by emotions (such as ear andgreed) that drive public markets, shareholders o NRssee value equal to the cost o the asset at the time opurchase. Tereafer, the property is subsequently re- valued according to conventional real estate valuationmethodologies, including comparable sales analysisdiscounted cash flow analysis and replacement costanalysis. Moreover, NRs are largely insulated rombroader exchange-traded fluctuations as their net asset

value, not market sentiment, drives pricing.

Exchange-listed REIs commonly trade at modest pre-miums to NAV (largely due to the liquidity they offer)the mean premium since 1990 has been around 3%However, there are periods when public markets aregrossly overvalued and this premium may rise to dou-ble digits. Mean reversion is a generally consistent orcewithin financial markets (albeit difficult to time). Tere-ore, overvalued traded REI prices eventually revert tothe historic lower mean.

3. Risk-Adjusted Returnsraded public REIS are subject to market risk throughtrading volatility. NRs may not be subject to the samerisk. Prices o traded REIs may change in reaction tochanges in equity markets, not because o their unda-mental exposures to systematic equity risk, but becauseo the buying and selling pressures resulting rom broadportolio adjustments by managers. o the degree thatcorrelations between prices o traded REIs and broadequity markets do not reflect the systematic risk o thisasset class, the required return on NRs should con-tain a lower equity risk premium, which will makeNRs more attractive. Te historic equity risk premi-um over long periods o time is around 4%. Tis sug-gests that someone investing in publicly-traded REIsought to expect 4% greater returns or doing exactly thesame thing in NRs. Tus, assuming rational investorsshould demand higher returns rom investing in securi-ties with higher risk, to match a 6% return on a NRa traded public REI must have a return o 10% (6%


10/73

Practitioner PerspectivesWhat a CAIA Member Should Know

9Alternative Investment Analyst Review The Case for Investing in Non-Traded REITs

plus 4% market risk premium). Tis suggests that giventhe exact same underlying investment (in property orin any other asset), the expected return or holding it ina publicly-traded structure as opposed to a non-tradedstructure ought to be at least 4% higher.

O course, the illiquidity o the NRs is a double-edgedsword. While this may lead to a reduction in the equity

premium demanded by investors or investing in NRs,it could lead them to demand a higher expected returnin order to compensate them or NRs’ lack o liquidity.As discussed below, not all investors are equally averseto illiquidity, and, thereore, they may demand differentpremiums or investing in illiquid assets.

4. Liquidity Preferences Differ

Liquidity is measured by the ability to convert a se-curity quickly into cash without any price discount.

raded REIs provide liquidity by virtue o trading onpublic stock exchanges. Again, NRs are not traded onexchanges. Instead, limited liquidity or some NRs isavailable through interim redemption programs such astender offers with a ull cycle liquidity exit event pro- jected or attempted within five to seven years o pro-gram inception.

Liquidity preerences differ based on the investor’s ex-pected holding period. For the short-term speculator,the liquidity afforded by traded REIs is essential. How-

ever, or the long-term investor interested in unlocking

the long-term illiquidity premium ollowing a buy-and-hold strategy, the liquidity offered by a traded REI iso little benefit, or they arguably have no need or itMoreover, liquidity can exaggerate losses or publiclytraded REI investors. For example, i enough investorsflee a traded public REI, the share price can drop be-low the value o the underlying real estate. Te loss flooron a NR investment by contrast is moderated by the

inability o investors to panic sell their securities. Tissuggests long-term investors, or the reasons outlinedabove, may derive unique benefits rom owning real es-tate in NR ormats rather than in traded REI struc-tures. Tough long-term investors may demand a rela-tively small premium or the illiquidity o NRs, theseinvestments will need to offer a premium to compensatethe long-term investors or illiquidity risk. 5. Capital-Raising Cycle

raded REIs are subject to market movements andmarket volatility in their ability to raise capital. Gener-ally, traded REIs have difficulty attracting capital whenthe real estate sector is out o avor (i.e., prices are low)or when there is a general perception o greater oppor-tunities available elsewhere. Tey have typically raisedand invested capital when real estate markets were in a- vor and capital markets activity buoyant—a time whenproperty prices are also generally at their highest. NRsby comparison, cater to long-term investors who areusually agnostic to cyclical activity o capital markets

and are seeking superior risk-adjusted yields with low

$33,477 $34,547

$20,657$22,360

$18,804

$8,439

$18,648

$10,284$8,358 $8,107

$6,105

$9,585

$0

$5,000

$10,000

$15,000$20,000

$25,000

$30,000

$35,000

$40,000

2013 2012 2011 2010 2009 2008

REIT Equity Off erings

Public Off erings ($MMs) Non-Traded REIT Off erings ($MMs)

Exhibit 1 REI Equity Offerings Source: SNL Financial


11/73

What a CAIA Member Should Know

10Alternative Investment Analyst Review

traded market correlations. Tis suggests traded REIs,unlike NRs, are generally unable to raise and investcapital at times when it is most opportune to do so.

6. Managing for the Long-ermraded REIs must contend with the demands o ana-lysts and speculative investors. Te pressure to meet andbeat analysts’ orecasts, where missing earnings expec-

tations may lead to significant stock price declines, ofenprompts management to ocus on short-term quarterlyearnings. Consequently, resources better spent on developing, acquiring, and managing properties are em-ployed to manage earnings and meet short-term marketexpectations. By contrast, and to the betterment o theportolio, NR managers are afforded the advantage oconcentrating on longer-term real estate investing op-portunities because their investors share a commonlong-term return objective. In short, traded REIs must

contend with the pressures or quarterly perormanceinherent within the exchange-traded market. Yet suchquarterly measurement is ofen a mismatch with the investment characteristics o real estate as an asset class.Te absence o this pressure in the private markets sug-gests managers o NRs are typically more readily ableto ocus long term on real estate investment and meet-ing investors’ long-term investment objectives, ratherthan managing quarterly earnings expectations.

7. Conclusion

NRs typically invest in sector-specific real estate pro-grams, targeting stable, ully occupied properties subject to long-term leases to strong credit tenants. Teyare thus able to generate immediate, durable, rent-driven cash flows rom the inception o the investmentas capital is deployed without a cash drag. Much liketraditional private equity core real estate investing, theyaggregate property through acquisitions and build di- versified portolios by tenant, geography, industry, andlease duration. Tey return value rom these aggregatedportolios via asset sales, public listings, or mergers,usually over a five to seven-year timerame. In the cur-rent environment, they afford a way to make more tacti-cal investing calls because they allow investors to profitrom the current historic high spread between low costfinancing and high acquisition cap rates.

NRs may hold a number o investment advantagesover their publicly traded counterparts: notably, a valu-ation o shares reflecting the intrinsic underlying valueo owned real estate, avorable risk-adjusted returns, ap-

propriate liquidity characteristics or long-term inves-tors, superior capital-raising and deployment dynamicsand a heightened management ocus on maximizinglong-term investment opportunities.

EndnotesO course, some o the actors that affect public markets couldeventually affect the net asset values o NRs. For evidence on

the size o illiquidity premium see Khandani and Lo (2009), “Il-liquidity Premia in Asset Returns: An Empirical Analysis o HedgeFunds, Mutual Funds and U.S. Equity Portolios,” http://web.mit

edu/Alo/www/Papers/liquidity4.pd

Author Bio

Sameer Jain has executive management cross-unc-tional responsibilities at American Realty Capital, in-cluding heading risk management, firm strategy, and

direction development, as well as

alternative investments. He has 18years o investing experience, wherehis responsibilities have included theormulation o investment strategythe development o risk managementpractices and asset allocation models

creating thought leadership, and the assessment and en-gagement o real estate, private equity, and hedge undmanagers. Prior to joining American Realty CapitaMr. Jain headed Investment Content & Strategy at UBSAlternative Investments, where he was also responsible

or all illiquid investing across the platorm. Prior toUBS he was at Citi Capital Advisors, Cambridge Alter-native Investments, and SunGard System Access. Hehas written many academic and practitioner articles onalternative investments, many o which are available inthe public domain at ssrn.com. Mr. Jain is a graduateo Massachusetts Institute o echnology and [email protected] Realty Capital, 405 Park Avenue, 12th floor

New York, NY 10022, 212-415-6500Acknowledgements and DisclaimerTe rameworks, methodologies, and concepts outlined in this manuscriphave been developed, based on the practitioner and academic literatureincluding by the author in the past. Tey are in all such cases the intellectual property o such sources. Te author reserves the right to reproducethis manuscript in part or whole. Te views and opinions when expressedin this article are those o the author only, and do not represent the viewand opinions o American Realty Capital, any affiliates and employees. Tauthor makes no representations or warranty, either expressed or impliedas to the accuracy or completeness o the inormation contained in thiarticle, nor is he recommending that this article serve as the basis or anyinvestment decision.


12/7311Alternative Investment Analyst Review Oil Price Drivers and Movements: The Challenge for Future Research

Research Review

CAIA Member ContributionCAIA Member ContributionResearch Review

Hillard HuntingtonEnergy Modeling Forum, Huang Engineering Center, Stanord University

Saud M. Al-FattahKing Abdullah Petroleum Studies and Research Center (KAPSARC) and Saudi Aramco

Zhuo HuangNational School o Development, Peking University, Beijing, China

Michael GucwaManagement Science and Engineering, Huang Engineering Center, Stanord University

Ali Nouri DarianiManagement Science and Engineering, Huang Engineering Center, Stanord University

Oil Price

Drivers andMovements:TheChallenge

for FutureResearch



1. Introduction

Te complexity o the world oil market has increaseddramatically in recent years and new approachesare needed to understand, model, and orecast oilprices today. In addition to the commencement o thefinancialization era in oil markets, there have beenstructural changes in the global oil market. Financialinstruments are communicating inormation about

uture conditions much more rapidly than in the past.Prices rom long and short-duration contracts havestarted moving more together. Abrupt changes in supplyand demand, influenced by such events and trendsas the financial crisis o 2008-09, uncertainty aboutChina’s economic growth rate, the Libyan uprising,the Iranian Nuclear standstill, and the DeepwaterHorizon oil spill, change expectations and currentprices. Although volatility appears greater over thisperiod, financialization makes price discovery more

robust. Most empirical economic studies suggest thatundamental actors shaped the expectations over 2004-08, although financial bubbles may have emerged justprior to and during the summer o 2008.

With increased price volatility, major exporters areconsidering ways to achieve more price stability toimprove long-term production and consumptiondecisions. Managing excess capacity has historicallybeen an important method or keeping world crude oilprices stable during periods o sharp supply or demand

shifs. Building and maintaining excess capacity incurrent markets allows greater price stability whenAsian economic growth accelerates suddenly or duringperiods o supply uncertainty in major oil producingregions. OPEC can contribute to price stability moreeasily when members agree on the best use o oilproduction capacity.

Important structural changes have emerged inthe global oil market afer major price increases.Partially motivated by governments' policies, majordevelopments in energy and oil efficiencies occurredafer the oil price increases o the early and the late1970s, such as improvements in vehicle uel efficiency,building codes, power grids, and energy systems. Onthe supply side, seismic imaging and horizontal drilling,as well as avorable tax regimes, expanded productioncapacity in countries outside OPEC. Afer the oil priceincreases o 2004-08, investments in oil sands, deepwater, biouels, and other non-conventional sources oenergy accelerated. Recent improvements in shale gas

production could well be transerred to oil-producingactivities, resulting in expanded oil supplies in areas thatwere previously considered prohibitively expensive. Tesearch or alternative transportation uels continueswith expanded research into compressed natural gasbiouels, diesel made rom natural gas, and electric vehicles.

In spite o these advances, some aspects o the worldoil market are not well understood. Despite numerousattempts to model the behavior o OPEC and itsmembers, there exists no credible, verifiable theory abouthe behavior o this 50 year-old organization. OPEC hasnot acted like a monolithic cartel, constraining suppliesto raise prices. Empirical evidence suggests that atsome times, members coordinate supply responses andat other times they compete with each other. Supply-restraint strategies include slower capacity expansions

as well as curtailed production rom existing capacityRegional political considerations and broader economicgoals beyond oil are influential actors in a country’s oildecisions. Furthermore, the economies and financialneeds o OPEC members have changed dramaticallysince the 1970s and 1980s.

Tis review represents a broad survey o economicresearch and literature related to the structure andunctioning o the world oil market. Te theoriesand models o oil demand and supply reviewed here

although imperect in many respects, offer a clear andwell-defined perspective on the orces that are shapingthe markets or crude oil and refined products. Muchwork remains to be done i we are to achieve a morecomplete understanding o these orces and the trendsthat lie ahead. Te contents that ollow represent anassessment o how ar we have come and where weare headed. Around the world governments, businesesand consumers share a vital interest in the benefits thatflow rom an efficient, well-unctioning oil market. It ishoped, thereore, that the discussion in this review wilfind a broad audience.

2. Price Volatility and Uncertain Conditions

Oil prices have fluctuated widely since 2004. Brentcrude oil prices rose rom $29 to $38 per barrel (annualaverages) between 2003 and 2004. Tey rose steadilyuntil 2008, reaching a record near $147 per barrel in July2008. Tis price spike reflected extremely strong Asianeconomic growth, combined with certain geopoliticaevents. Prices collapsed below $33 within the next ew


14/73

Practitioner PerspectivesResearch Review

13Alternative Investment Analyst Review Oil Price Drivers and Movements: The Challenge for Future Research

months as the world economy spun downward intofinancial disarray. Tey spurted back to levels above$80 per barrel in 2010, as the economies in Asia andelsewhere recovered. Additional price increases in 2011beyond $100 per barrel were prompted by continuedAsian growth and supply uncertainty mounted with theArab uprising and the Libyan disruption. Continuedears about the financial system and uture economic

growth lingered in August 2011, causing world oilprices to begin their retreat once again.

Tese conditions have created massive uncertaintyabout where uture oil prices will be headed and whatactors create these dramatic price movements. Peakoil arguments abound during an era when non-OPECoil production has increased only modestly despite therecord-high prices. urmoil dominates the politicallandscape in the Middle East, ueling additional concerns

about the security o oil supplies. Most disconcerting toboth oil producing and oil-consuming nations has beenthe financialization o oil, where financial motives andtrading permeate oil transactions and make physicalmarkets appear less important.

Tis uncertainty creates very significant problems ormajor oil-consuming countries that are trying to recoverrom financial disintegration, as well as investors whoare considering long-term allocations to commodities.It also raises important concerns or major oil-

producing countries with ample resources. Shouldthey expand capacity to supply growing economiesand at what rate? How much spare oil capacity shouldbe maintained to offset sudden oil-market surprises—unexpectedly higher economic growth, political unrestin oil-producing regions, or major oil spills in offshoredrilling areas? Fundamental actors should be importantor both capacity decisions, but these uncertainties haveeroded the belie that these actors still operate in thesame way that they have in the past.

Capacity expansion influences both short and long-term market operations. First, greater capacity allowsmore uture production to meet growing demand.Tese decisions require an understanding o long-term market conditions. Second, additional capacitycan also build surplus capacity or market imbalances.Tese decisions require an understanding o short-termmarket conditions. Although the distinction betweenthe short and long term can be ambiguous, we definethe short-term to include horizons o three years or less.

Oil markets are not easy to understand and projectionso uture oil prices have not been accurate consistentlyI undamental supply and demand analysis and oilmarket modeling have any benefits, it would appearto be in their ability to organize complex inormationefficiently and to provide better understanding o howoil markets perorm. For this reason, it is sensible toemphasize these characteristics, rather than to ocus on

their suitability or orecasting. 3. Long-Run Oil Price Drivers and Models

Oil represents a substantial proportion o global energydemand. As the world’s most highly traded internationalcommodity, oil will continue to play a large role inmeeting energy demand in the uture. Over the longrun, the price o oil will be influenced by our majortrends: (1) global economic growth, (2) demand-sidetechnological progress and efficiency gains, (3) new

alternative energy sources, and (4) the changing costs oproduction. Te depletion o easily extracted resourcesis pushing production into more technologicallydemanding fields, lower-quality crudes, and higher-costoperating environments. At the same time, dramaticimprovements in technology are expected to continueto reduce the cost o finding and producing oil romsuch reserves. Government policies will have importantimpacts on the costs o both petroleum products andcompetitive energy sources. Understanding howproduction, consumption, and the price o oil will

change over the coming decades is o vital interest toboth oil-producing and oil-consuming nations, withstrong implications or energy policy, economic growthclimate-change policy, and international stability.

3.1. Oil Demand: Drivers and rends

Generally speaking, when the world economy as awhole experiences growth, oil demand will increaseTe existence o this undamental relationship isuncontested, but its strength varies between regions andwill be moderated by many actors with the potentialto curb demand, such as uel-saving technologies, uel-switching to different orms o primary energy, andpolicies designed to constrain carbon dioxide emissions

Much o the recent growth in global oil consumption(which rose by 1.5% per year between 1985 and 2008)occurred outside the OECD nations. As a percento world consumption, the emerging nations’ sharehas grown rom 37.6% to 44.5% over this periodDeveloping economies are expected to continue being


15/73

Research Review


the primary drivers o the growth in global oil demand.Te hypothesized energy and environmental Kuznetscurve, which views investments in energy efficiencyas a luxury good that become more affordable andwidespread as developing economies mature andprosper, teaches us that continued strong economicgrowth in China and elsewhere may work paradoxicallyto restrain the growth rate o demand—i only in the

longer run.

3.1.1. Growth and Industrialization

Per-capita oil demand grows at the same rate as theeconomy in many emerging economies, so long as otheractors like prices do not change. Many countries areexperiencing rapid increases in vehicle penetration andownership rates as incomes rise. Based on estimates as o1973, oil income elasticities exceeded unity throughoutthe developing regions o the world, and approached a

level o 2 in the poorest nations. Tis implies that oildemand should increase at least as ast as GDP in thedeveloping world, holding constant energy prices andtechnological progress. In the poorest Asian nations,oil demand should expand nearly twice as ast as GDP(Medlock and Soligo 2001 and Van Benthem andRomani 2009).

In contrast, per-capita oil demand grows more slowlythan GDP within the OECD, even beore the impact opotentially rising prices is actored in. Vehicle ownership

per person has stabilized and consumers are beginningto purchase alternative-uel vehicles in these countries.Gately and Huntington (2002) estimate that the long-runincome elasticity to be 0.55 in the more mature OECDcountries, implying that oil demand may increase onlyabout hal as ast as GDP in the industrialized portionso the world (again, abstracting rom the impact opotential changes in prices, regulation, and technology).

3.1.2. Oil Demand and echnical ProgressWhereas pure price-substitution implies reversibility,technological progress that is induced by price increasescreates an irreversible and unidirectional effect that isnot easily unwound, even when prices return to previouslevels. Several distinct processes drive technical changesthat influence oil demand. Te first is exogenous changethat is largely unrelated to specific changes in the priceo oil or economic conditions. For example, airplanedesigns incorporated significant improvements in uelefficiency, even prior to the price shocks o the 1970s. Inan endogenous process, rising oil prices are the specific

incentive that drives technical change. Automobilecompanies, or example, revamped their vehicle fleetsafer the 1970s to make passenger cars more ueefficient; even when oil prices declined afer 1985, thosedesign innovations were never eliminated.

3.1.3. Alternative Vehicles and Competitive Fuels

Limited historical evidence exists by which to measure

the strength and potential o inter-uel substitutionamong competing uels. In many countries, petroleum-based uels appear to have no strong or viable competitoror powering transportation. Tat may be changingas countries have begun to make commitments to vehicles ueled by compressed natural gas, biouels, andelectrification. Additionally, companies may increasinglypursue gas-to-liquid processes as a technological optionthat substitutes relatively inexpensive natural gas or oiin the production o diesel uels. Energy security and

climate mitigation policies may accelerate these oil-reduction trends.

3.1.4. Demand Response to Oil PricesI uture oil supplies are expected to be scarcer thantoday, uture oil prices will rise and curb some o thegrowth in demand; but, by how much? Tis questionhas probably attracted more o the attention o energyeconomists and commodity investors than any otherissue during the last ew decades. A major conclusionconsistent with the findings o most studies is that

the longer-run demand response to any gasolineprice increases occurring over the next twenty yearsis likely to be several times larger than the short-termresponse that is initially apparent (Dahl and Sterner1991 and Goodwin, Dargay et al. 2004). Te responseo consumption to price is the combined effect o manydifferent decisions. Utilization decisions impact thegasoline market by reducing traffic activity and thenumber o miles driven by households. Over a longerperiod, household response to higher prices is alsomagnified as the vehicle fleet is retired and replaced.

Te price elasticity o oil demand seems to be declininglately within the United States and perhaps morebroadly within the OECD. Many countries outside othe OECD maintain large uel subsidies that imposea wedge between crude and product prices (Arzedel Granado et al. 2010). Removal o those subsidieswhich have become quite expensive to maintain, wouldincrease uel prices to the end-user and thereby reduceuture oil demand. Te lack o data and estimates or


16/73



the emerging countries limits our ability to oresee howthese changes will influence oil markets in greater detail.

3.2. Oil Supply Availability and CostsDespite significant gains achieved via enhanced oilrecovery technologies, conventional oil supplies arediminishing in many fields located outside o theMiddle East. Development o unconventional resources

to offset this decline will be very important, but the cost,availability, and scale o resources such as Alberta's oilsands are as yet unknown. At the same time, oil supplyprospects, even rom conventional resources, mayimprove in certain areas. New oil may be discovered inrelatively unexplored regions and reserve appreciationin known resource basins remains an important sourceo new additions. echnical progress will probablycontinue to reduce exploration and developmentcosts significantly, as well as to enhance the saety and

security o operations that extend urther into rontierareas. Governments may reduce oil supply barriersby rolling back production royalties and taxes, and byeasing constraints on leasing and acreage.

3.2.1. Resources and Geological Availability

Oil resources are scattered across the globe in ormationswith very different characteristics. Based upon its worldoil assessment o 2000, the United States GeologicalSurvey (2003) estimated that there were 1,898 billionbarrels o remaining conventional oil and natural gas

liquids, excluding cumulative volumes that had alreadybeen produced. Tese geological estimates are basedupon likely discoveries, given the prevailing oil pricesand available technologies present in 2000.

Tese conventional resources are supplementedby considerably larger volumes o unconventionalresources—heavy oil, oil sands, and oil shale—thatrequire specialized extraction technologies andsignificant processing beore the oil can be sold. Aguileraet al. (2009) estimate that the combined volume oconventional and unconventional oil would last or 132years i production increased by 2% per year.

3.2.2. Resource Costs

For economists evaluating market conditions, resourcecosts, rather than total reserves, determine whetherscarcity prevails. Many geological estimates do notdistinguish between resources that are inexpensive toextract and those that are much more costly to developand produce. o fill this gap, a useul concept is the

resource availability curve—a schedule that representsthe total known resource base that could be developedat each successively higher-cost level.

Aguilera et al. (2009) derive an availability curve orconventional and unconventional petroleum resourcesTey estimate 7 BOE (trillion barrels o oil equivalent)o conventional resources and 4 BOE o heavy oil

5 BOE o oil sands, and 14 BOE o oil shale withaverage production costs usually considerably higherthan the comparable costs or conventional oil. Te costestimates or these unconventional petroleum resourcesare very uncertain and subject to change. o be useul,any long-run cost estimates should reflect productionexpanded to scale and the considerable learning thatwill accumulate through experience in developing theseresources. Oil prices may well overshoot these long-runcost estimates during intervening years when additional

unconventional sources are not yet large enough tomeet growing demand.

3.2.3. Oil Supply from Competitive RegionsProducers outside the major exporting countries aregenerally considered as competitive price takers. Marketprices must cover the marginal cost o producing the lastunit o these supplies, including both the direct expensesand the firms’ opportunity cost o drilling or oil, ratherthan engaging in another economic activity. I resourcedepletion is a actor, each supplier will also consider the

opportunity cost o current extraction relative to utureproduction. At higher prices, firms can justiy exploringor and extracting more costly resources, and doing soearlier rather than later.

wo major trends are driving oil supply rom regionsoutside o OPEC: the depletion o reserves that are easyto extract and the improvement o oil exploration andproduction technologies. Te combined effects haveled to an increase in mega-projects aimed at resourcesthat were ormerly inaccessible, either commercially ortechnically. Such projects include the Alberta oil sandsthe deep water resources o the Gul o Mexico, and thepre-salt deposits offshore o Brazil.

3.3. OPEC

Te major oil exporters are sufficiently large to influenceas well as to respond to price. Tey have market powerHowever, the extent to which market power hasbeen exercised is less certain. Te previous empiricalliterature leaves many questions regarding the impact


17/73

Research Review


o decisions and actions taken by OPEC. Te datatends to support multiple competing theories, withoutdefinitively excluding any particular behavioral model.Analysts choose their avorite hybrid; they seldom testall versions. Tere is clearly room or additional researchon the nature o OPEC and its evolution.

3.4. Long-term Models

Te long-run behavior o the oil market has receivedconsiderable study through the application o computermodels. Models can be classified by many differentcriteria, but we find it helpul to distinguish structuralmodels rom computational models. Both approachestake undamental microeconomic theories aboutthe objectives, constraints, and behaviors o marketactors into consideration at their core. Tese theoriesare distilled into a mathematical structure, allowingor interaction between the actors within a specific

market context. Te primary distinction between thetwo categories is the level o complexity and detail;computational models have significantly more detailedrepresentations o the market at the cost o model runtime. Tey also have increased data requirements andmay offer less straightorward interpretations.

Research into the ormal modeling o the oil marketbegan largely as a response to the oil crisis o 1973. Teinitial goal was to understand the role o OPEC decisionmaking and its impact on the market price. Since that

time, as the oil market has changed, and the researchcommunity has become more international, structuralmodels have been applied in analyzing a wide range oissues involving oil. Te major structural approachesinclude simulation, optimization, and game-theoreticrameworks.

In simulation models, the behavior o actors in themarket is represented by a specific unction contingenton market conditions. Tis unction can be basedeither on some rule-o-thumb (such as a target priceor target capacity utilization rule), or on historicaleconometric estimates o past behavior. Depending onthe researcher’s ocus, the behavior o different agentsmay be described in various levels o detail. In general,OPEC is given more complex behavior, while non-OPEC producers ollow a simple supply curve, ofenone that exhibits constant price elasticity. O course, theresearcher’s goal is to develop rules or unctions that aredescriptive o actual behavior.

In an optimization model, at least one agent activelychooses its behavior to maximize an objective unctiontypically related to profit or welare. For models o theoil market, the optimizing agent is generally assumed tobe OPEC, or some subset o that organization. OPECchooses a level o production to maximize the present value o profits, while taking how the resulting pricewill influence the decisions o competitive producers

and consumers into account. While some models mayhave sophisticated representations o the limits to theknowledge available to the optimizing agent, in manycases, the optimizer is given complete oresight othe uture path o the market. With the optimizationapproach, the researcher seeks to understand what amarket player must do to obtain his best outcome.

In game-theoretic models, two or more agents areassumed to have market power, or at least some influence

on each other’s welare. Tey attempt to take actionsthat are optimal, given their anticipation o what theother agent will do. Each agent is also assumed to takeinto account the strategic behavior o other actors in themarket. A game-theoretic approach may be useul whenit is necessary to explicitly consider the consequenceso rivalry and competition between different largeplayers in the market—or instance, when evaluatingthe incentives or individual OPEC members to deviaterom established production quotas.

Computational models share many attributes withstructural models and are largely distinguished by thesheer number o details included. Te complexity o themodels makes them costly to build and maintain and thelevel o detail makes it difficult to establish the impact oany one model choice. However, computational modelsacilitate certain types o analysis that are impossible witha structural model: detailed impacts upon individualstakeholders, specific technological scenarios, andull policy analysis. Moreover, one approach tocomputational modeling, so-called computable generaequilibrium models, has been used extensively toinvestigate uel substitution opportunities and thebroader energy sector impacts o global greenhouse-gasemissions policies. Computational models also acilitatethe division o labor in the modeling effort by dividingthe project into distinct sub-modules.

Due to their cost and complexity, computational modelsare relatively scarce, but with cheaper computer power


18/73



they are becoming more common. Still, computationalmodels are typically confined to institutions suchas the International Energy Agency and the EnergyInormation Administration.

3.5. Research Gaps in Long-term Oil MarketsWhile there has been significant effort representing thelong-term behavior o oil markets using models o all

sorts, a great deal o work remains to be done. As newtechnologies are developed, demand grows, new kinds oresources are exploited, and relationships in the marketchange; the theories and models we apply to the oilmarket need constant re-evaluation. A ew topics standout as significant open questions in our understandingo the long-term behavior o the market going orward:demand behavior, modern OPEC behavior, producerwelare, and resource depletion.

3.5.1. Demand BehaviorWith the exception o the large computational models,most oil models do not have a very sophisticated ordetailed representation o the demand side o the market.Understanding demand dynamics would be useul notonly in explaining recent price movements, but also inexploring the impacts this degree o demand variationhas on oil-producing nations. Marked variation, butespecially unpredictability, o demand presumablyaffects the welare o producers, not just consumers,and may change the nature o their capacity investment

decisions. Tree major topics in demand behavior standout as top candidates or urther exploration: (1) thehigh rate o demand growth in developing countries,(2) the asymmetric response o oil demand to pricechanges, and (3) the role o technology in altering theenergy intensity o oil-consuming activities.

3.5.2. Security and Climate PolicyClosely aligned with demand issues is the inclusiono other energy market dynamics that produce viablesubstitutes or oil-based products or transportation.Governments are adopting policies to accelerate theshif consumption away rom oil through mandates,taxes, and subsidies—all in response to concerns aboutenergy security and global climate change. o derivemeaningul results, the broadening range o availablesubstitutes or petroleum-based uels requires thesimultaneous evaluation o multiple uel markets, ratherthan oil-only analysis.

3.5.3. Modern OPEC Behavior

In the late 1970s, OPEC was modeled by many to be amonopolist in the world oil market. One author oncereerred to it as a ‘clumsy” cartel (Adelman 1980). Modelsdeveloped in late 1970s and early 1980s examined anumber o different theories regarding OPEC’s behaviorand market power. However, OPEC and its membershave evolved through time and observations gleaned

rom the 1970s are now outdated.

In the most recent two decades, the global view o OPEChas changed. OPEC is no longer considered definitivelyas a cartel that exercises market power by regulatingoutput. Smith (2009) suggests that OPEC has beenrestraining investment in new oil production capacityin recent years and thereby has contributed to higherprices in a market with very rapid demand growthAlthough research efforts to study OPEC’s behavior

either econometrically or theoretically have diminishedcompared to prior years, there remains a need or newtheoretical models describing OPEC; these modelsshould be tested with detailed data culled rom recentyears.

3.5.4. Producer Welfare

Many o the market-power models treat OPECproduction decisions as i they were made by a profit-maximizing firm or cartel. When trying to understandthe impact o OPEC production decisions on global oil

prices and consuming nations, such a ormulation maybe an adequate approximation o the decisions madeby the organization. However, in reality, as sovereignnations, both political and economic concerns drivedecision making. Oil-producing nations may constrainprices in order to maintain avorable relationships withother nations, or they may sell oil at a discount in theirhome market to benefit domestic consumers. It maymake more sense to view the nations as maximizingwelare rather than maximizing profit.

Unortunately, when moving rom models that considerprofit to ones that try to measure welare, modelingtechniques increase in complexity and require greaterinormation on the national economy as a wholeWhile some models (De Santis 2003) have previouslyapproached this important question, a great deal owork remains to be done in this area.


19/73

Research Review


3.5.5. Resource Depletion

Oil reserves are finite and production will become moreexpensive and perhaps eventually hit a peak (Hubbert1962). It remains unclear when such a peak will occurand whether it will be based on a lack o availableresources or the lack o sustained crude oil demand. Inact, the threat o peak oil has loomed over the horizonsince the dawn o the petroleum age, but consistent

resource discoveries, unconventional resources, andtechnological breakthroughs have so ar managed toexpand oil supplies and may continue to mitigate cruderesource scarcity or the oreseeable uture. As discussedelsewhere (Smith 2012), it is not even clear that a peakin the production o oil, i it does occur, would be aharbinger o impending scarcity.

4. Short-Run Oil Price Drivers and Models

Generally speaking, conclusions regarding the short-

run behavior o oil prices are even less certain than ourknowledge o the actors that drive long-term trends.In large part, this is due to the relatively short historyo investigation into short-run fluctuations, as well asrecent changes in the composition and liquidity o short-term oil markets that have only begun to be sorted out.

Modeling short-run changes in the oil market requiresdifferent techniques, depending upon the specific issueunder investigation. Financialization, in particular, hasmade the oil utures and other derivatives market more

liquid and perhaps more influential, while the numbero participants in financial markets has increasedbecause o hedging and investment opportunities. Teuse o high-requency data may be required to considerall the relevant details in short-term models, but mucho that data is not available in the public domain. Teprimary goal o short-term models is to provide a betterunderstanding o short-term price movements andto create short-term orecasts. In contrast with long-term models, short-term models do not usually seekto determine what the uture equilibrium path or oilprices will look like. Instead, they attempt to orecastprices or price changes that are expected to be observedin the near uture. Tis is typically attempted with thehelp o reduced-orm models that estimate parameterso statistical models that best describe short-run pricemovements without considering the undamental orceso supply and demand. Short-term models also use thepowerul financial theories concerning arbitrage andrisk-taking in an attempt to iner market expectationsrom observed uture prices.

4.1. Critical Observations

During the previous decade, the oil market experiencedsignificant short-term upsets, one the most important owhich was the boom-bust price cycle during 2008. Tatparticular episode challenged the ability o conventionalmodels to provide adequate explanations and orecastso oil prices. Many studies have looked to find structuraexplanations, but there still is no consensus on the

underlying economic causes. In addition to the highlevels o price, a higher level o volatility has beenobserved in the oil market in recent years. For the firsttime, a change o $100 per barrel in only our monthswas observed in oil prices rom July to November 2008

Tese trends are not limited to the oil market; financiaactivity and turmoil in commodity markets ingeneral have increased. Te volume o investment incommodity index unds, overall utures market trading

activity (as revealed by the open interest in all contractmaturities), and correlations among commodity pricesas well as between commodities and equities, haveincreased by varying degrees. Forward curves havebecome substantially flatter at times, indicating thatutures prices at varying maturity dates are now movingmore closely with each other and also with spot pricesFinancialization may act as a double-edge sword; itincreases market liquidity and acilitates price discoveryand risk management. However, it also creates moreopportunities or some traders who would attempt to

distort and manipulate utures prices.

Against this backdrop, there appear two overridingchallenges or the modeling community. First, is theneed to examine whether utures trading causes artificiamovements in the spot price o oil or not, and, i so,to trace out the expected remedial effect o alternativeregulatory reorms. Second, is to assess i and howfinancial variables can be used to orecast uture pricepaths more accurately than methods that are based onundamental analysis alone.

4.2. Fundamental Drivers

Certain economic actors have played a undamentarole in recent price changes. Supply and demandshocks, together with the continuous flow o newsand uncertainty that surrounds them, are the primarydrivers underlying short-run oil price dynamics. Teimpact o these shocks is magnified by the low elasticitieso both short-run oil supply and demand. Hamilton(2009) demonstrates that under specific assumptions




about the elasticities o supply and demand, one canexplain the 2008 boom-bust price cycle just by usingundamental supply and demand actors. A caveat isthat the price predictions drawn rom such models areextremely sensitive to the specific magnitude o thepresumed elasticities. Nevertheless, short-term supplyand demand drivers are believed to be able to describemost o the observed price changes. Tis section reviews

these undamental drivers.

4.2.1. Short-term Supply Drivers

During 2005-08, available inventories were depletedwhile major oil-producing countries held low levelso spare capacity. Considering the inverse relationshipbetween spare capacity and spot oil prices, and theinelastic supply in the short-term, this has led to higherprice levels.

Short-term supply shocks have also influenced the oilmarket. Te Deepwater Horizon oil spill in the Gulo Mexico and the revolution in Libya are two recentexamples o such events. Consider first the deep waterHorizon spill. Te methods o obtaining liquid uels arebecoming increasingly reliant on advanced and capitalintensive technologies. From deep water drilling, tothe processing o oil sands, to advances in refining, theoil market is changing its risk profile. While engineersare constantly working to perect control systems andreduce the chance o ailures, the potential or damage

rom any single catastrophe is increasing. Furthermore,with the increasing lumpiness o production rom thetrend towards more complex megaprojects, the supplyimpact o a single outage (or addition) is increasing,potentially leading to greater price volatility (Skinner2006).

4.2.2. Short-term Demand DriversTe short-run demand or oil is also relatively priceinelastic. Tere are our main reasons or this. First, oilconsumption levels cannot change quickly, due to theexisting stock o vehicles and other equipment that usesoil. Second, in the OECD countries, oil consumptionis less responsive to price changes because the share oconsumers’ energy expenditures as a raction o theirtotal incomes is relatively low. Tird, oil demand indeveloping countries is largely driven by steady incomegrowth and industrialization. Fourth, the demandimpacts o crude oil price changes are in many casesoffset by government subsidies or taxes.

Macroeconomic news also influences oil prices. Asincomes increase and economies expand, more energywill be used or transportation, heating, and coolingHicks and Kilian (2009) utilize a direct measure oglobal demand shocks based on revisions o proessionaorecasts o real GDP growth. Tey show that recentorecast surprises are associated primarily withunexpected growth in emerging economies. According

to this line o research, markets have been repeatedlysurprised by the strength o this growth.

Finally, U.S. oreign exchange and interest rates exert aninfluence on the price o oil. Te price o oil (in USD)increased by more than 600% rom January 2002 toJuly 2008. Te same increase in terms o the Euro wasless than 300% as the Euro gained strength during thatinterval. As this example suggests, depreciation o theU.S. currency may either lead or at least contribute to an

increase in oil prices (which are typically expressed inU.S. dollars). Fluctuations in interest rates influence the value o oil in the uture relative to its value today, whichcan lead to changes in production, consumption, andstorage decisions. In addition, changes in interest ratesprompt changes in the prices o financial derivativesaccordingly.

4.2.3. News and Information Signals

In financial markets, the price is believed to reflectall publicly available inormation. Newly released

inormation about uture events will have a proportionateimpact on today’s price. All kinds o news are relevantinormation regarding the economic growth o differentcountries, the prices o other commodities, currencyrates, major countries’ stock market movements, signso geopolitical unrest or uprisings, unexpected severeweather conditions and natural catastrophes, and manyother actors. Te flow o inormation can change pricesrequently and sharply. However, to have any impactthe news must be credible.

Previous research shows that not all announcementsmade by major players in the oil market (OPEC, IEAetc.) are credible. o better understand short-run pricemovements, it is important to distinguish betweenrelevant, credible announcements and ones that areignored by the market. An important step in conductingthis analysis is to consider the incentives o the issuerso inormation: specifically, whether those objectivesare aligned with the truthul revelation o inormationA signaling ramework and a orecast model can be



Research Review

used to simulate the effect o new announcements andanalyze their incentives.

4.3. Price Forecasting ApproachesMost short-term oil market models ocus exclusivelyon oil price and its statistical time series properties. Incontrast, structural models explicitly speciy and attemptto estimate the impact o changes in oil demand and/

or supply. Tis distinction means that short-term pricemodels are mainly limited to the task o orecasting,rather than providing economic interpretation o thesort required or policy analysis. Despite the ratherlarge number o recent short-term price models thathave appeared in the literature, significant opportunitiesremain or urther study.

4.3.1. Reduced Form Models

Reduced orm models take advantage o financial and

structural data and employ econometric tools to builda model and estimate its parameters. Tese modelsare usually applied to orecast a specific variable (e.g.,world oil price). Tey differ primarily in terms o thecomplexity o the statistical structure that is assumedto fit the data. Tree contrasting approaches have beenused to study oil price data: 1) time-series analysis,e.g. autoregressive moving average (ARMA), generalautoregressive conditional hetereoskedastic (GARCH),2) Structural Vector Autoregression (SVAR) and 3)non-parametric regression, e.g. artificial neural network

(ANN) (Jammazi and Aloui, 2012). Each approach hasits own advantages and, at the conceptual level, no oneapproach is superior to the others. Tereore, the choiceamong the models should be dictated by the observedstatistical properties o the time series involved in theanalysis.

I the reduced orm models are applied or purposesbeyond simple orecasting, however, serious problemsarise. Tese problems center on the concept andinterpretation o “causality,” a term that plays an ofenmisunderstood role in many short-run time-seriesanalyses. Causality is, o course, central to the study opolicy analysis. o be successul, policy makers must beable to anticipate the consequences o their actions. Willtrading limits cause volatility to decrease? Will producingrom the strategic petroleum reserve cause prices todecline? And so on. Te cause and effect relationshipsthat are implicit in these questions represent somethingstronger than the statistical tendency o two variables tomove together, which is not evidence that an exogenous

change in one variable will cause another resultingchange in the other variable. Tereore, it is essentialwhen contemplating short-term orecasting models tounderstand that a finding o “Granger–causality,” whichis based on patterns o correlation, neither proves nordisproves that a undamental causal relationship linksone variable to another. o the extent that a undamentastructure is added to the short-term approach in the

orm o a SVAR, it is also important to keep in mindthat the structure that is assumed to link the variablesin a causal chain is typically dictated by convenience, aswhen a diagonal pattern o variable exclusions is adoptedin order to permit the model to be solved recursivelyor, when a priori constraints are imposed on the sizeo key parameters to achieve identification o causalrelations. O course, i the constraints are untested andinappropriate, so may be the causal relations.

In summary, short-term statistical models will continueto flourish, based in part on the availability o additionalhigh-requency pricing data and in part due to increasedscrutiny o financial investors in the oil market. It will beimperative or both the producers and consumers o thisresearch to keep in mind the undamental limitations othese time-series methods and to tailor their inquiriesto questions that can properly be answered with thetools at hand.

4.3.2. Financial Models

Financial models are a more recent brand o oil pricemodels that extend statistical analysis to some o thenewer time series (utures prices and options) withguidance rom relevant hypotheses developed in thetheory o finance. Since options and utures contractsconvey inormation about the uture, they have beenconsidered as a first step in incorporating financial datain oil models. However, utures prices may include a riskpremium that varies through time, and thereore, theydo not represent a simple expectation about the pricethat will prevail in the uture. It has been shown, orexample, that “no change” orecasts are more accuratethan orecasts based on utures prices (Alquist, Kilianet al. 2010).

Pagano and Pisani (2009) document significant time- variation in the risk premium and use the degree ocapacity utilization in U.S. manuacturing and oilinventory levels as proxies or this variation. Teydemonstrate how one can find expected uture pricesbased on the combination o utures prices and the risk




premium. Tus, i one could model and orecast the riskpremium when combined with market data, it shouldbe possible to obtain estimates o uture expected prices.Given the potential value o this ability to producers andconsumers alike, urther research into determinants othe risk premium seems warranted.

4.3.3. Structural/Financial Hybrid Models

Hybrid models, combinations o structural andfinancial models, are motivated by the need to produceshort-term orecasts that are more consistent withsupply and demand rameworks. Tese models arecalibrated to base-case orecasts o a long-term model,with outcomes that are adjusted based on the flow onew market inormation and short-term economicresponses. Relevant new market inormation wouldinclude price observations rom utures markets,orecasts in demand growth, and supply shocks (e.g.,

the reduction in Libya’s production during 2011).Hybrid modeling requires estimates o both short-term and long-term elasticities with which to simulateprice responses. Te model takes inormation signals asinput and generates price and quantity paths as output.In light o the Efficient Market Hypothesis, the marketresponds instantly to the new orthcoming inormation.Further efforts to incorporate theories o commoditiesand storage might lead to models capable o orecastinginventory changes and the movement o utures pricesas well. (e.g., Routledge, Seppi, and Spatt 2000).

4.3.4. Modeling Volatility Market price volatility can be estimated either rombackward-looking historical data or rom orward-looking financial derivatives using implied volatility(Szakmary et al. 2003). Indeed, even or models where volatility is not o direct concern, a researcher mightneed to know how volatility and price shocks lead tochanges by consumers and producers. For example, theuse and production o oil are heavily tied to existingcapital stock and capacity investments. Price shocks,even over a relatively short time rame, can have lastingimpacts on demand and supply or years to comethrough their impact on capital investment. MonteCarlo methods and artificial neural network technologycould be applied to simulate supply and demand shocksand to estimate the benefits o major producers adoptingstrategies that stabilize prices, but that is all dependenton first developing an understanding o how the use oexcess capacity and stockpiles influences volatility.

4.4. Analytical/Teoretical Models and Insights

Financial aspects o oil markets are not well exploredStudies are still trying to confirm a range o theoreticahypotheses about the operation o the financial marketsand to identiy the most important financial driversTese include models that do not try to simulateor orecast the whole oil market. Instead, they usepartial equilibrium or econometric techniques in an

attempt to understand short-term market movementsmore accurately and to distinguish among competingtheoretical hypotheses.

During 2000-08, when oil prices were increasinginvestments in commodities markets also increasedsignificantly. Tis triggered the question o whether theprice rise o 2008 represented a financial bubble o somekind or not. Brunnermeier (2009) defines a speculativebubble as characterized by the ollowing elements: (1)

prices are higher than the undamental value, (2) agroup o investors buys the asset based on the belie orsentiment that they can sell it to others later at a higherprice, and (3) such belies or sentiments cannot besupported by undamental actors.

Studies on the role o financialization can be categorizedinto two groups: conceptual models and statisticatests. Te ormer type o analysis consists o deductivearguments or accepting or denying the hypothesis thatan increase in financial activity will cause prices to rise

more than what undamental actors would dictate. Telogical validity o these arguments rests solely on theunderlying assumptions independent o any empiricalevidence. Te latter type o analysis ocuses onquantitative relationships between trending variables tofind statistical patterns o predictability.

A ew studies cite conceptual arguments to advancethe claim that excessive investment in commodityindex unds might have played a role in creating thebubble. However, conceptual analysis alone cannotestablish the strength or magnitude o the effect. Tusadditional empirical research is needed to clariy thepicture. Certain conceptual relationships remainingso ar are still rather inscrutable, even afer they havebeen quantified. For example, ang and Xiong (2010)find a link between increased price correlations amongdifferent commodities and the growing volume ocommodity index investments. However, there isno indication, theoretical or otherwise, that highercorrelations are good or bad.



Research Review

In any event, elevated correlations are not evidence oa bubble. Regarding the possibility o a bubble inducedby financialization, it is useul to remember that pricemovements in utures markets with rising indexund investment have not been moving uniormlyupward (Irwin, Sanders et al. 2009). Moreover,Headey and Fan (2008) show that prices o manynon-financial commodities—commodities that were

not financialized—displayed similar dynamics as thefinancialized commodities, despite having no influx ospeculative financial investors. Tere is considerableroom or additional research into the price movementso all commodities, whether they are financialized ornot. Te parallel movements suggest the presence osome common actors beyond financialization andresearch is needed to identiy and measure the influenceo those actors. It seems likely that any progress in thisdirection will depend on a more complete appreciation

o the role o common demand shocks, inventories, andconvenience yields.

4.5. Statistical estsResearchers commonly perorm statistical causalitytests to describe the temporal relationship betweenspeculators’ trading activity, oil price movements, and volatility. As noted above, these tests using Grangercausality establish causality not in a structural sense,but only confirm whether the observed movement inone variable precedes the changes observed in another

variable. Te difference is important: although onemight move a picnic inside just beore it starts raining,moving the picnic doesn’t cause the rain. With that caveatin mind, and using non-public data, the Interagencyask Force on Commodity Markets (2008) studiedthe dynamic relation between daily price changes andchanges in the positions o various categories o traders.Tey ound that some trader positions can be predictedas a response to price changes, but not the reverse.Sanders and Irwin (2011) find evidence that larger longpositions by index traders Granger-cause lower market volatility. Tis result is contrary to the popular beliethat index traders’ activities increase market volatility.

Tere is some additional evidence on the other sideo the argument and the conflicting conclusions arean invitation to pursue urther both conceptual andempirical research into the causes o commodity pricemovements. One example is the group o studies thatevaluate whether or not statistical characteristics ooil price movements match the pattern o an explosive

bubble. In contrast with explosive bubbles thatprevailed or several months in the copper and nickemarkets, Gilbert (2010) finds only weak evidence or anexplosive bubble in the oil market, and that it appears tohave endured or only a ew days in July 2008. Even sodo we know what caused it, or why it subsided? Further,a ew studies report some evidence, conceptual as wellas empirical, that financial activities were driving the

oil price away rom its undamental value during 2008Although undamental actors are important in hisanalysis, Einloth (2009) suggests that speculators mayhave been building inventories rom March to July in2008 based upon evidence that spot prices rose urtherafer convenience yields had begun to all. Singleton(2011) finds significant empirical support that investorflows influenced excess returns rom holding oil uturecontracts o different maturities, afer controlling or anumber o other exogenous actors.

In summary, there have been many studies, but asyet no absolute consensus on the causes o the oiprice boom and bust o 2008. Although there existsonly limited statistical evidence that the price cyclerepresented a speculative bubble caused by an influxo financial traders, the matter remains the subject ogreat debate among researchers, policy makers, and thegeneral public. Te value o any urther work that helpsto clariy this issue would be substantial.

4.6. Prescriptive ModelsShort-term modeling is a relatively new approach. Somestudies have tried to build a theoretical ramework orinteractions between the financial markets and thephysical markets. Tese prescriptive studies usuallysimpliy the details o the actual market and examine various phenomena that would be expected to occurunder certain conditions. Te main goal o thisdeductive approach is to understand how the marketworks, rather than orecasting or simulating with highaccuracy.

For example, Deaton and Laroque (1996) andRoutledge, Seppi, and Spatt (2000) consider storageagents in the commodities' markets and determinehow the levels o inventories should change withuncertainty and how orward curves should behave insuch settings. Routledge, Seppi, and Spatt interpret theconcept o convenience yield as an option that storageagents will exercise at an optimal time. Allaz (1992)develops a generic commodity market model (1992) to


24/73



demonstrate that, depending on the relative strength othe hedging and strategic motives, a producer’s optimalposition in orward markets may be either short orlong. Brandts, Pezanis-Christou, and Schram (2008)study o Cournot (quantity) competition and Liski andMontero’s (2006) inquiry into a potential link betweenorward contracting and collusion are urther examples.Te stylized nature o modeling that lies behind all

o these studies invites extensions that explore therobustness o the findings to more realistic depictionso the agents who trade in these markets.

5. Te Role of Saudi Arabia in the Global Oil MarketTe influence o Saudi Arabia on the global oil marketis indisputable. Saudi Arabia’s role and decisionparameters since the discovery and production o oil inthe Kingdom have been determined by different actors.Al-Monee (2011) discussed this issue and highlighted

our important actors.

Te first actor is the size and production lie o SaudiArabia’s oil reserves. For the past fify years, SaudiArabia has had very large crude oil reserves, equivalentto 20% o the world’s proven reserves.

Te second actor is the diversity o Saudi Arabia’sexport outlets. Saudi Arabia is exporting to the U.S.,Europe, and the Far East. Tis diversity o outlets (andcrude types exported) offers Saudi Arabia marketing

flexibility and highlights the international consequenceso its policies. In addition, Saudi Arabia is exporting itsoil to the rest o the world rom two domestic terminalslocated on its eastern and western coastlines.

Te third aspect is the Kingdom’s large crude oilproduction capacity. Saudi Arabia maintains a largeexcess capacity that is available to ace supply disruptionsand demand surges. Saudi Arabia’s excess capacity in thepast three decades since 1980 averaged 60% o OPEC’s(and o the world’s) excess production capacities, whileits share in OPEC’s and the world’s production averaged32% and 12% respectively during the period. Tisunused capacity averaged 35% o Saudi oil productionduring the 1982-1990 period, 13% during the 1990s,and 14% in this decade. OPEC’s averages over thesethree periods were 17%, 6%, and 4%, respectively.

While the other OPEC members' excess capacities dependon market conditions, Saudi Arabia made an official policysince the mid-1990s, o maintaining 1.5-2 MBD excess oil

production capacity at all times. Saudi Arabia’s role hasbeen very useul to sofen the impact o major oil supplyinterruptions, such as the Iran-Iraq war, Iraq’s invasion oKuwait, the Venezuela crisis in 2003, Hurricane Katrina in2005, and the Libyan crisis in 2011. Tese actions helped inlessening oil market volatility and stabilizing oil prices.Te ourth acet is the role o oil in Saudi Arabia’seconomy. For the past three decades, oil has represented

35% o Saudi Arabia’s GDP, 84% o its governmentrevenues, and 90% o its merchandise exports. Teserates explain the high interdependence between theKingdom’s domestic and international oil policies.

Tese our actors have pushed Saudi Arabia to developits own oil industry through its national oil companySaudi Aramco. Te company was created throughthe purchase by Saudi Arabia o the assets o the ourAmerican companies operating in the Kingdom. Saudi

Aramco was entrusted with the tasks o managing anddeveloping the hydrocarbon resources o the Kingdomto achieve its development objectives, executing thegovernment energy policies, and developing thetechnical skills needed in that sector.

Saudi Arabia’s oil policies are geared towards efficiencyand sustainability, which involves stable oil markets andan efficient oil industry that is able to play a strong role inthe oil sector. In the ace o environment uncertaintiesSaudi Arabia is investing in research and development

projects such as research centers, universities, andcompanies. Regarding the role o Saudi Arabia in OPEC, it has beenas important or OPEC as OPEC has been or SaudiArabia (as suggested by R. Mabro, Oxord Institute orEnergy Studies 2001). Te roles o OPEC and SaudiArabia have evolved in line with market changesSuch changes include the diversity o market playersthe influence o the financial markets on the physicalmarkets, the energy policies o consumer countries, andclimate change, as well as energy security concerns.

Since the influence o the financial market on the physicaloil market is increasing, Saudi Arabia has acknowledgedthe new market reality and adopted a policy o urgingthe international community to exert some regulatoryoversight, as well as transparency measures, over themeans o transactions in such markets. In order tostabilize the market, Saudi Arabia has been collaboratingwith international organizations such as OPEC and


25/73

Research Review


IEA to reach better predictability. Te Kingdom is alsopromoting the strengthening o the producer-consumerdialogue, among other things, by strengthening the roleo the International Energy Forum (IEF), created in 2003and entrusting it to coordinate the Joint Oil Data Initiative(JODI) to enhance the flow o timely and accurate oildata worldwide.

In the early 1990’s, Saudi Arabia realized that thechallenges o climate change would add to oil supply anddemand uncertainties and so it integrated its climate-change policy with its oil policy. It also considersenergy security as a two-dimensional concern: supplysecurity (the availability, diversity, and reliability oenergy supplies at all times) and demand security (thepredictability, efficiency, and growth o energy demandin line with economic growth).

Saudi Arabia is expected to continue playing adominant role in stabilizing the global oil market. TeKingdom will continue its efforts to ensure sustainableoil supply to the world with stabilized long-term pricesat reasonable levels. At the same time, it will go onwith its investments in the oil and gas sectors to ensureadequate supplies and sustainable economic growth.It is expected to maintain an excess capacity o 1.5 to2 MBD to ace supply crises efficiently. Finally, SaudiArabia’s oil policy will be defined in dialogue with otherproducers and consumers to address the environmental,

investment, and price volatility challenges as a whole.

6. Conclusion

Te complexity o the world oil market has increaseddramatically in recent years and new approaches areneeded to understand, model, and orecast oil pricestoday. In addition to the rapid financialization o theoil market, many undamental structural changes haveaffected physical markets or oil. Financial instrumentsnow communicate inormation about expected changesin the underlying undamentals much more rapidlythan in the past, so the implications o both financialand physical developments are clearly linked.

Casual evidence o the closer relation between financialand physical markets may be ound everywhere. Teprices o long- and short-dated contracts have startedmoving more closely together. Sudden supply anddemand adjustments, including those related to China'seconomic growth, the Libyan uprising, and the DeepwaterHorizon oil spill have changed expectations in ways that

affect both current and utures prices. Although volatilityappears to have increased, financialization has arguablymade price discovery more robust and expectationsmore transparent. Most empirical economic studiessuggest that expectations regarding undamental driversand their uture trends shaped prices during the 2004-08cycle, although over-exuberant expectations cannot beruled out completely, based on available evidence.

With increased price volatility, major exporters arenow considering ways to provide more price stabilitywhich is needed to improve long-term production andconsumption decisions. Managing excess capacityprimarily within OPEC, but also in the strategic stockpilesheld by major consuming nations, has historically beenan important actor in keeping world crude oil pricesstable during periods o sharp demand and supply shifso what extent would the expansion o excess capacity

alter market expectations in the current environment?Would the result be greater price stability in the ace ouncertainties regarding, or example, the rate o Asianeconomic growth, the debt

AIAR Vol 2 Issue 4 March 24 Final_0

Documents

Transcript of AIAR Vol 2 Issue 4 March 24 Final_0