U N I V E R S I T É P A R I S 1 P A N T H É O N ‐ S O R B O N N E Thesis Analysis of Options in FPSO Leasing Contracts RÉDIGÉ PAR : GWENC’HLAN CATHERINE DIRECTEUR DU MÉMOIRE : OLIVIER BRANDOUY INSTITUT D’ADMINISTRATION DES ENTREPRISES DE PARIS International Master in Business Administration MBA Promotion 2010 - 2011
54
Embed
MBAThesis_GwenCatherine_Analysis of Options in FPSO Leasing Contracts
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
U N I V E R S I T É P A R I S 1 P A N T H É O N ‐ S O R B O N N E
Thesis
Analysis of Options in FPSO Leasing Contracts
RÉDIGÉ PAR :
GWENC’HLAN CATHERINE
DIRECTEUR DU MÉMOIRE :
OLIVIER BRANDOUY
I N S T I T U T D ’ A D M I N I S T R A T I O N D E S E N T R E P R I S E S D E P A R I S
International Master in Business Administration MBA
Promotion 2010 - 2011
2
l’Université n’entend donner aucune approbation ni improba‐
tion aux opinions émises dans ce mémoire : ces opinions doi‐
vent être considérés comme propres à leur auteur.
3
Acknowledgement
To Olivier Brandouy for his lively and exciting corporate finance classes which gave me the
desire to discover more on the concept of Real Options, for accepting to supervise this thesis
and over all for his invitation to attend the Master 2 “Decision Sciences and Risk Manage‐
ment” at ENSAM Paris and IAE‐Paris.
To Damien Mourey for his inspiring management accounting classes, his in‐depth analysis of
managerial issues which always raised my interest and developed my curiosity, and last but
not least for his paramount contribution as Director of the MBA to always deliver a high level
education.
To Delphine Dion for her valuable inputs on the importance of methodological aspects in
research and specifically for this thesis.
To Elizabeth Minik for her kind support to all students of the MBA and her always positive
attitude and
To Adeline Leclercq for her great help and expertise in modeling stochastic phenomena.
Production activities in the oil business are by nature subject to large uncertainties and sig‐
nificant risks. Through contractual provisions in the lease contracts, technical and commer‐
cial underlying risks are split between the Contractor and the Oil Company. While both par‐
ties face risks of various origins, the present chapter presents only the risks having significant
direct impact on the economic results of the both parties.
2.6.1 RISKS BORN BY FPSO CONTRACTOR
Cost overrun during FPSO engineering and construction
Once the leasing contract is signed with the Oil Company, the maximum income of the FPSO
Contractor is fixed. But the expenses to be incurred as part of the investment in the FPSO
construction are only estimated. The actual cost could prove to be higher than the initial
estimate. The direct impact is a lower than expected financial results.
Schedule slippage during FPSO engineering and construction
Schedule risk has some similarity with the risk of cost overrun. Both schedule and cost are
estimated prior to the award of a lease contract. As a matter of fact, the actual duration and
cost of the engineering and construction phases can be greater than original estimates. In
theory actual duration and cost can also be smaller than initial estimates, but it is rarely the
case. Both risks take place in the early phase of the project and have direct impact on project
NPV. A cost overrun corresponds to an increase of the initial investment. With income re‐
maining the same, it leads to a decrease of the project NPV. The impact of delay can also be
understood as a cost impact. A delay in the FPSO completion implies a delayed start of the
lease period and additional financial costs. Considering the example of 2.5, one year delay
leads to an increase of financial costs by 96 Million USD.
FPSO availability risk: commercial uptime
By contract the FPSO Contractor is paid the full rate only if the FPSO is fully available for pro‐
duction. In case of breakdown diminishing oil production, the day rate is not paid or only
partially. This scheme works as an incentive for FPSO Contractor to maintain the FPSO in
good conditions during the lease contract. During the FPSO early life, design or construction
issues may still need to be fixed. The impact on NPV is even greater as it affects cash flows
discounted with factors close to 1. The Graph 6 represents the NPV of the same lease con‐
tract as in 2.5 with a 95% uptime only throughout the lease period.
17
GRAPH 5 ‐ NET PRESENT VALUE WITH LOWER COMMERCIAL UPTIME
Disposal value: Scrap/redeployment value
The shorter the leasing contract duration is the more residual value normally remains at the
end of the contract. This value is estimated prior to entering into the leasing contract. When
comes the end of the contract, it has still to be materialized either through a sale of the unit
or through a new contract. This may not always be an easy task. The example of the FPSO
Falcon (source SBM reports from 2009‐2011) clearly illustrates this. It was laid‐up for more
than 3 years waiting for a new contract. The Graph 5 represents the impact on NPV of an
actual residual value lower than the one initially estimated.
GRAPH 6 ‐ NET PRESENT VALUE WITH OVERESTIMATED RESIDUAL VALUE (SOURCE: THE AUTHOR)
18
2.6.2 RISKS BORN BY OIL COMPANIES
Technical uncertainties: reservoir, recovery rate, field life
The quality of oil field is by nature difficult to predict. The geology of the reservoir and its
geometry are complex. Various physical and chemical parameters may affect the perfor‐
mance of an oil well. An Oil Company tries to assess how much oil a well will produce based
on the Seismic surveys and test wells drilled during the exploration phase. Some oil fields
were believed very promising but failed to deliver, for example the Golfinho field in Brazil.
On the contrary, some fields demonstrated much better production than expected prior to
the start of production. This technical uncertainty of the reservoir performance is perfectly
illustrated by the example of the Jubilee field operated by the Oil Company Anadarko. Prior
to the start of production, 4 different production profiles were considered possible scenarii
by Anadarko as shown on Graph 7.
Commercial risk = Oil price volatility
Beyond the quantity of oil recovered, the Oil Company supports the variations of the oil
price during the project whole life. The decision to develop an oil field is taken knowing the
oil price history. But the income of the Oil Company depends on the future oil prices during
the whole production period. Considering as an example a field where production starts 3
years after the project sanction. The expected field life is 17 years. Overall, the Oil Company
income is subject to market volatility for the next 20 years. This commercial uncertainty
should be seen both as a risk and an opportunity. Considering the same oil field, investment
decision had been taken on an average price of crude oil of 40 USD per barrel. But 5 years
after, oil price reach 100 USD per barrel. This results in much higher profits for the Oil Com‐
pany.
Environmental risk: oil pollution
Oil pollution is inherent to the production activities. Breakdowns on critical part of the facili‐
ties could lead to oil spillage. Even though most incidents remain limited in magnitude, a
large pollution is one of the most feared risk for Oil Companies. The most example is the
Macondo accident in the Gulf of Mexico during year 2010. The fines and various compensa‐
tions amounted to more than 20 billion to BP and its insurers.
3 OPTIONS IN FPSO LEASING CONTRACTS
The previous section describes the fundamentals of developing an oil field with a leased
FPSO both from the Oil Company and the FPSO Contractor perspectives. As Oil Companies
face uncertain environment the options included in the leasing contracts offer interesting
tools for managing this uncertainty. This section first describes the most commonly used
options and analyses how they impact the Company and the Contractor. Then actual data
19
related to the exercise of options by Oil Companies provides interesting results on potential
strategies for FPSO Contractors.
3.1 OPTIONS TO EXTEND
3.1.1 DESCRIPTION OF THE OPTIONS
Most of the FPSO leasing contracts include provisions to extend the leasing of the FPSO be‐
yond the fixed period set in the contract. The number and the duration of the options re‐
quested are given by the Oil Company as part of the tender documentation sent to FPSO
Contractors.
Should the Company exercise the option to extend, the Contractor will continue to make
available for production its FPSO. During the option to extend, the Company will pay to the
Contractor a day rate under the same contractual provisions. The day rates associated with
the options to extend are part of the initial contract. It may happen that a Company requires
the day rate for the options to be the same as the one for the firm period. But quite often
different day rates are possible.
Beyond the options offered to extend the lease, the Company may be interested in continu‐
ing the production with an FPSO. In such case, the initial lease contract does not apply. The
Oil Company shall seek for a new agreement with the Contractor to continue the lease. Dif‐
ferent terms and conditions may apply, reflecting the balance of powers between Company
and Contractor at this time.
3.1.2 EXAMPLE OF THE JUBILEE FPSO
In order to illustrate the concept, the case of the Jubilee FPSO is a good example. We were
personally involved in the bidding process for the lease of a new FPSO. All details here after
come from the documentation provided at that time. The Company, Anadarko, requested
bids for a firm period of 7 years and completed by 13 options to extend the leasing period by
one year. Overall the FPSO was requested to have a design life of 20 years. In its invitation
to bidders, Anadarko had included the following requests
“Extension of Charter Term (After the initial 7 year period). In case of additional charter periods, Company will compensate Contractor the Charter Rate per Day of:
Charter Rate Schedule for Optional Exten-sions
USD/Day
Extended charter period of 13 years
Year 8
Year 9
Year 10 (Etc.)
20
Extension of Services after 20 years. Company retains the option to have Contractor extend the Bareboat Charter for additional periods of time beyond the initial period of the Agreement and all of the extension periods contemplated, subject to mutual agreement between Company and Contractor.”
In this example, it is interesting to note that day rates can be different for each one year op‐
tion. This gives opportunity to the Contractor to propose a customized scheme with varying
day rates that would better fit Contractor needs. Another comment is to highlight the rather
short firm period compared to the potential total leasing period: 7 years over a total of 20
years. This illustrates well the idea of Oil Companies aiming to reduce their risk with shorter
firm period and a high number of options to extend the duration.
While FPSO Contractors were requested for building a 20 year design life FPSO in the frame
of a firm leasing contract of only 7 years plus options, no technical information was provided
with regard to potential production profiles. Several production scenarios were considered
by the Company as illustrated in Graph 7 but this information was not disclosed to FPSO
Contractors.
The 4 scenarios account for different reservoir behaviors and the estimated quantity of re‐
coverable oil range from 362 Millions of barrels (mmbbl) in case n°1 to 593 mmbbl in case
n°4. Assuming the FPSO turn‐down point (the point at which production is no more profita‐
ble and FPSO is demobilized) corresponds to 30% of the FPSO nominal capacity or 40,000
barrels per day, this milestone is reached after 7 years only in case n°1, the pessimistic sce‐
nario. In the 3 other scenarios, this point is reached after 10 years or 15 years. The Oil Com‐
pany choose the most pessimistic scenario to determine the firm leasing period. The firm
period is equal to the most pessimistic time to turn‐down and not to the expected time to
turn‐down. This is a clear pattern of risk aversion from the Company.
21
GRAPH 7 ‐ JUBILEE FPSO PRODUCTION PROFILES SCENARIOS (SOURCE: ANADARKO)
3.1.3 IMPACT OF OPTION EXERCISE ON FPSO CONTRACTOR ECONOMICS
Until now, the NPV were calculated only on the firm duration of the leasing contract. This
contract may however be extended by the Oil Company exercising the yearly options. In
such case, the NPV can be improved as illustrated in Graph 8. The red curve represents the
NPV of a project having an IRR of 8% (same as the Contractor WACC) calculated on its firm
period and 3 options to extend by one‐year exercised by the Company. After 11 years, the
NPV is positive although the NPV was negative after the initial firm period. The exercise of
options by the Company ultimately increases the project NPV of the Contractor.
22
GRAPH 8 ‐ NET PRESENT VALUE WITH LOWER IRR AND CONTRACT EXTENSION (SOURCE: THE AUTHOR)
At the same time, the risk on the residual value diminishes. Indeed, the FPSO residual value
decreases with the time. A lower value has therefore to be considered after some more
years of operations on the field. The risk taken by the FPSO Contractor to monetize this
amount is de facto lower. This can be observed graphically on Graph 8. The positive NPV of
the base case (blue curve) is only obtained through the sale of the FPSO for its residual value
which is depicted in Graph 8 with a NPV step between year 7 and year 8. In the case of ex‐
tended lease, no such step exists.
3.2 OPTIONS TO PURCHASE
In addition to the options to extend, Oil Companies normally request options to purchase
the FPSO in lease contracts. This gives a second level of freedom for Oil Companies in man‐
aging their field.
An example of the contractual provisions associated with such options is given hereafter
from the Jubilee FPSO case:
“Purchase of FPSO. At any time after achievement of Schedule Milestone M12B (Reference
Exhibit “F” – Contract Schedule) during the Term of the Agreement, Company retains the op‐
tion to Purchase the FPSO. Such purchase would include all components provided under the
Bareboat Charter, with condition of the FPSO being that in existence at the time of the exer‐
cise of this option by Company.
23
In case of purchase of the FPSO by Company, Company will compensate Contractor on a Lump Sum basis subject to agreed Terms and Conditions of this Agreement, and the Purchase Price given in the following Table, and no further Charter Rate payments shall be made by Company after the date of purchase.”
The analysis of historical data shows however such options are only rarely exercised by Oil
Companies. Among the 40 FPSO lease contracts considered in 3.3, only 3 FPSOs have been
purchased after the firm period duration.
Having such purchase options is however a must for Oil Companies to easily take over from a
FPSO Contractor and continue producing the field after the purchase of the FPSO. Without
such option, the FPSO Contractor bankruptcy could put the Oil Company in a very difficult
position and ultimately cause to stop oil production.
3.3 ACTUAL EXERCISE OF OPTIONS BY OIL COMPANIES
Options being present in almost all contracts, we have gathered factual data on the actual
exercise of those options by Oil Companies. The aim is here to derive information on how Oil
Companies treat options from the analysis of actual recent data. Three types of sources
were used: websites of FPSO Contractors, annual reports and presentation to investors. The
dataset includes a total of 40 FPSOs owned and leased by the 3 leading FPSO Contractors:
SBM, Modec and BWO. It represents more than 50% of all the FPSOs currently leased or with
a signed leasing contract. Information on FPSO leased between 1980 and 2000 is not easily
available. This concerns however a very limited number of FPSOs. An extract of the data is
presented in Table 2 for illustration. The full dataset is available in Appendix 2.
ing the total duration of fixed period and extensions offers another perspective. The sum of
the fixed period for the 17 FPSOs amounts to 80 years and the total of the extensions repre‐
sents 50 years. So the actual duration of the lease contracts increased by an average of 60%
compared to the firm period included in the contracts.
In order to complete this analysis, it is worth reviewing what happened to the 4 FPSOs
whose lease contract was not extended beyond the firm period. Three of them were pur‐
chased by the Oil Companies which exercised options to purchase included in the contract.
The last one, Xikomba FPSO, was leased initially to ExxonMobil. Shortly after the end of the
firm period, the FPSO Contractor, SBM, entered into a contract with ENI to redeploy the unit
on another field nearby. ExxonMobil being a minority partner of ENI in the new field, the
redeployment alternate was considered more attractive for ExxonMobil than keeping it pro‐
ducing at its original location. But one could imagine that without such alternate plan, the
FPSO would have been kept on site and ExxonMobil would have exercised its options to ex‐
tend.
The analysis of the dataset leads us to the following conclusions. Options to extend are very
frequently exercised by Oil Companies. In the few cases where the FPSO lease is not extend‐
ed, the Oil Companies exercise their option to purchase the unit. The probability for an FPSO
Contractor of having an idle FPSO at the end of the firm period is therefore deemed very
low. Beyond this first conclusion, one may hypothesize that the reason why options are mas‐
sively exercised by Oil Companies is because they are so cheaply priced by FPSO Contractors.
Further to this analysis carried out at a macroscopic level, it would be interesting to revisit it
at a finer level. The exercise of options could be analyzed at the level of Oil Companies. This
could possibly lead to conclusions on the behavior and the risk aversion of the various Oil
Companies.
3.4 BEYOND CONTRACTUAL OPTIONS
The options included in FPSO leasing contracts offer significant flexibility to Oil Companies. It
allows in particular to adapt the duration of field production to the actual conditions faced:
oil prices and reservoir quality. Oil Companies have a prudential approach when fixing the
firm period duration. It generally represents the duration during which production is be‐
lieved to be profitable considering the most pessimistic case scenario envisaged prior to the
field development. The Jubilee FPSO case is a clear illustration of this approach. Similarly the
Contractor BWO states in its 2009 annual report: “It is common for customers, i.e. the oil
25
companies, to contract the firm period for the FPSO lease corresponding to the expected pro‐
ducing life of the reserves.”
There is however one case where the oil field proved to produce less oil than forecasted by
the Oil Company in its pessimistic scenario. On the field of Golfinho in Brasil, the Oil Compa‐
ny Petrobras had concluded a contract with SBM for leasing the FPSO Capixaba. The firm
duration was 7 years with 3 options of one year. The oil field output proved however to be
very disappointing after production started with Capixaba FPSO. Petrobras therefore pro‐
posed SBM to relocate the FPSO to another field within the framework of the same contract
but with revised commercial conditions to account for the FPSO upgrade costs.
Another case of direct agreement between Company and Contractor is the FPSO Falcon.
ExxonMobil had contracted with SBM for a firm period of 6 years. No option for extension of
the lease was included in the contract (source SBM presentation of 2004 Full Year Results).
Later on, ExxonMobil negociated to extend the lease for one year until December 2009.
It comes out from the 2 above examples that Oil Companies have opportunities to adapt the
FPSO lease beyond the initial contract. But the commercial conditions of such changes are
most probably more expansive than it would have been under a competitive bid.
4 VALUATION OF OPTIONS TO EXTEND
Given the success of options in Leasing contracts, having proper tools to value those options
is a must for well managed FPSO Contractors. The traditional pricing FPSO lease contracts
with DCF models remains a cost driven only approach. Assessing the value creation potential
of the options for the Oil Company is a new interesting approach that we propose. For the
FPSO Contractor, it should be considered as a major tool for defining improved pricing strat‐
egies. This section presents a new framework for valuing options to extend. It discusses the
parallel with the theory of financial options and analyses the main differences with the usual
real option cases through the use of simple cases. A concrete example of a valuation model
is then proposed with numerical simulations. Finally the new valuation method is comment‐
ed from a managerial perspective and future development are considered.
4.1 A MOVE TOWARD VALUE FOR FPSO CONTRACTORS
The traditional approach used by FPSO Contractors for pricing lease contracts is to apply the
DCF method. This cost driven approach is based on the cost estimate of capital expenditure
required to provide the FPSO. We believe a value driven approach should be considered to
26
complement the classical DCF analysis. Offering options to Oil Companies creates value for
them. Through the options they can benefit from positive conditions (high oil prices, in‐
crease of recovery rate, etc…) during field production while limiting their risks exposure. Op‐
tions to extend can thus be seen as creating value for Oil Companies and FPSO Contractor
have legitimacy to receive a share of this created value. Valuation of those options should
therefore be considered as a first step. Ultimately, the result of such analysis should allow
FPSO Contractors to build pricing strategies able to capture more from the value created for
Oil Companies.
4.2 LITERATURE REVIEW
We have searched the existing literature on the valuation of options in FPSO leasing con‐
tract. No previous work has, to our knowledge, already been published on the subject.
A number of authors have worked on the valuation of Real Options in Natural Resources
Development. Brennan and Schwartz (1985) proposed a famous model for the valuation of
mines in a production versus closure problem. Most of the papers consider however the
valuation of natural resource prior to the start of production. And more important, valuing
FPSO leasing contract and its options has only an indirect link with the value of the natural
resource, in our case the oil field.
The problem of real asset leasing has been treated first from the perspective of the lease
versus buy decision and the impact of taxes, with an important contribution from Myers, Dill
and Bautista (1976). Grenadier (1995) adopted a real option approach to value the leasing
contract: leasing an asset for a fixed period T is equivalent to a portfolio which includes the
asset and a European put option with a strike price of zero at the expiry T.
Recognizing the approach of Myers, Dill and Bautista (1976) was not suitable for valuing the
various operating options of leasing contracts, Trigeorgis (1996) proposed another approach
with contingent claims analysis and suggested a numerical approach. Options in lease con‐
tracts “can be seen as claims whose value is contingent on the future value of the lease as‐
set” (Trigeorgis, 1996). An option to buy the asset at the end of the leasing is equivalent to a
call option. An option to extend is similar to a lease contract with a longer fixed period to‐
gether with an option to cancel with no penalty.
It is interesting here to note that value of options in FPSO lease contract is contingent to the
future value of the oil field and not of the FPSO while Tigeorgis (1996) considered the future
value of the asset only. The main argument to link the value of the FPSO lease options to the
one of the remaining oil to be produced is the very high cost of mobilization of a new FPSO.
More recently, Jorgesen and De Giovanni (2010) use an approach similar to Trigeorgis (1996)
for the valuation of time charter contracts with purchase options in Shipping. The value of
the ship is derived from the charter spot rate using a mean reverting process for spot rate.
27
The various options are then valued based on the future ship value. Jorgesen and De Gio‐
vanni conclude their paper with proposing as possible future research a multi‐factor model
as opposed to their single factor model relying on the charter spot rate. In our case of FPSO
leasing, the value of the oil field being produced can be seen as a 2 factor model. The oil
price and the remaining quantity of oil to be produced are indeed the 2 main sources of un‐
certainty.
Given the present state of the research on the subject, we first illustrate the specificities of
FPSO leasing options with two simple models and decision tree approach. Then we propose
and comment a numerical method to approximate the value of options to extend based on a
Monte Carlo approach and a simplified stopping rule.
4.3 EXPLORING SIMPLE CASES
4.3.1 SINGLE ONE YEAR OPTION TO EXTEND
Let us consider here the simple case of a firm leasing period of n years and only one option
to extend the lease by one year.
The analogy with a financial call option is interesting:
Long = the Oil Company buy the right to extend
Call = right to produce with the FPSO for one year can be seen as the right to
buy an additional quantity of oil. Here the additional quantity is unknown as produc‐
tion for year n+1 is uncertain.
Underlying asset = the project to produce the oil field with the FPSO
Strike price K = rate for leasing the FPSO one year plus other Company OPEX
Expiration date = last day of year N
Payoff = zero or (actual value of the production of year N+1 – K)
At year n+1, we consider only two possible states :
The option is exercised and the FPSO leased during year N+1
The option is not exercised and production stops at the end of year N
28
GRAPH 9 – DECISION TREE FOR A ONE‐YEAR EXTENSION OPTION (SOURCE: THE AUTHOR)
We consider a simplified decision rule. The Company exercises the option if the expected
payoff is positive, said otherwise if the expected income for the year N+1 is greater than the
operating expenses:
It is interesting to note that the actual payoff can only be known after the option is exer‐
cised, once the quantity of oil produced during year N+1 is sold at the market price prevail‐
ing in year N+1. A consequence is that the actual payoff can be negative if the actual income
is less than the strike price K. The option payoff in case of exercise is therefore as shown in
Graph 10, but with a probability of negative payoff remaining low. It corresponds to the
conditional probability of getting an actual negative payoff for year N+1 whereas the ex‐
pected payoff for the same period was greater than 0 at year N.
29
GRAPH 10 ‐ PAYOFF OF A ONE‐YEAR EXTENSION OPTION (SOURCE: THE AUTHOR)
4.3.2 SEVERAL ONE YEAR OPTIONS TO EXTEND
Further to the first case, let us now describe a second case with the following characteristics:
a firm period of N years and two 1‐year options to extend. We consider the same framework
as previously and the same decision criteria.
Similarly as in the first case, we consider for the sake of simplicity only two possible states
for the oil field: expected benefit for the year is positive or negative.
GRAPH 11 – THEORITICAL DECISION TREE FOR 2 ONE‐YEAR EXTENSION OPTIONS (SOURCE: THE AUTHOR)
In B, the expected benefit for year N+1 is positive, the option to extend is exercised. The
year after, there are two possible states: D and E. In D the expected benefit is positive, the
second option to extend is exercised. On the opposite, the expected benefit is negative in C
and the FPSO is demobilized.
In C, the expected benefit is negative. The option is not exercised and the FPSO is demobi‐
lized. The further states F and G do not practically exist as shown in Graph 12 as the produc‐
30
tion cannot resume easily after the FPSO demobilization in C. In the real world, this corre‐
sponds to the high cost of remobilization of the FPSO and restart of the field.
GRAPH 12 ‐ DECISION TREE WITH POSSIBLE STATES ONLY FOR 2 ONE‐YEAR OPTIONS (2)
The simplified decision criteria of this case leads to considering only the superior branch of
the binomial tree. In practice, a Company can however continue leasing a FPSO even though
the short term expectation is a loss. In such case, the Company accepts a short term loss
against the potential for future benefits. The production profile of a field being generally
declining with the time, future benefits can be expected only if an increase of future oil price
is expected or if an upgrade of the FPSO to increase the production is considered. The simpli‐
fication made is deemed sufficiently realistic considering the exploratory nature of the pre‐
sent work. A more realistic decision criteria could be considered in a future research.
In order to understand how to value the 2 options to extend the easiest way is may be to
take an example. We consider here a strike price of 200 and the expected incomes are given
in the following table. Each path has the same probability of occurrence.
TABLE 3 ‐ EXPECTED INCOME FOR 2 ONE‐YEAR EXTENSION OPTIONS (SOURCE: THE AUTHOR)
31
At time N+2, Path 1 is in the money, i.e. the value of the underlying is greater than the strike
price. The option is exercised and the expected payoff is (230‐200=30). Path 2 is out of the
money. The option is not exercised and the payoff is 0.
At time N+1, Path 1 and Path 2 are in the money (point B of the above tree). The expected
payoff is (270‐200=70) and the option is exercised. Path 3 is out of the money. The option is
not exercised and the FPSO is demobilized.
Once the decisions to exercise the options are taken, the FPSO produces during the related
period and the actual income of the Oil Company are as follow.
TABLE 4 ‐ ACTUAL INCOME FOR 2 ONE‐YEAR EXTENSION OPTIONS (SOURCE: THE AUTHOR)
The actual payoff are then calculated as the actual income minus the strike price K.
TABLE 5 ‐ ACTUAL PAYOFF FOR 2 ONE‐YEAR EXTENSION OPTIONS (SOURCE: THE AUTHOR)
The earnings of the oil Company are cumulated year after year. The total earning of the
Company on Path 1 are 75 +10 after exercise of Option 1 and Option 2. The present value of
Path 1 earnings at time N are therefore with a discount rate R = 10%
175
1 0.110
1 0.1
Similarly for Path 2 and Path 3,
275
1 0.1
3 0
So the value of having the two one‐year options is
32
1 213
.
We can also calculate the value on having only a single option. In that case, we consider the
paths 1 to 3 until N+1 only,
175
1 0.1
275
1 0.1
3 0
.
The last step is to calculate the value of Option 2 alone. The value of Option 2 can be consid‐
ered from two perspectives: the value of Option 2 at year N and the value of Option 2 at
year N+1. The value of Option 2 at year N+1 is in fact the present value at time N of the ex‐
pected payoff of year N+2 discounted at time N considering the Option 1 has been exercised.
Indeed, Option 2 does not exist if Option has not be exercised. The value of Option 2 at year
N is the present value at time N of the expected payoff of year N+2. In the latter case, when
calculating the expected payoff, one shall consider the real probability of occurrence of each
state of the nature at year N+2 and not the probability of each state conditional to having
exercised the prior option.
,
1
With
.
.
The numerical results confirm the following equation
33
4.4 A FIRST ATTEMPT TO VALUE OPTIONS TO EXTEND
As discussed in 4.2, the value of an option to extend in a FPSO leasing contract is contingent
to the value of the remaining oil to be produced. Continuing on this idea, we analyze the
options value from the perspective of the Oil Company. The options to extend give flexibility
to the Company. We therefore propose to compare two cases, a first case with flexibility in
the duration (noted O) and a second case with fixed duration (noted F). The option value is
then considered as the difference in value between Case O and Case F.
The value considered here is the NPV obtained as the sum of DCF. The DCF analysis is a clas‐
sical approach used by Oil Companies for valuing their investment in oil fields “as the nature
of their business leads to a certain visibility on the stream of free cash flows” according to
Deloitte (2008). It is therefore a natural approach for our case.
We propose to use a Monte Carlo method to simulate a large number, n, of possible oil price
evolutions and production profiles for the duration of an oil field life. Considering one of the
simulation and the associated serials of average oil price and total production for each year,
the FPSO leasing contract is extended as long as it remains profitable to do so for the Oil
Company. The simplified decision rule described in 4.3 is applied to extend or not the con‐
tract each year after the end of the firm leasing period. The pay‐offs are computed each year
as the difference between the income, i.e. the quantity of oil produced during the year sold
at the average yearly oil price, and the operating expenses. Then NPVs can be computed for
the following lease contracts:
a contract with a firm period of M years, with M being the maximum field life
( / )
a contract with a firm period of M‐1 years plus 1 one‐year option ( / ),
…
a contract with a firm period of M‐k years plus k one‐year options / with k
<M
Now considering all the n simulations of oil prices and production profiles, the NPVs for each
lease contracts are calculated in each simulated case and recorded. The expected value of
NPVs are then obtained as the mean of the NPV calculated in each case i.
34
/ = mean { /
with / being the NPV calculated in the simulated case i for the contract
/
The value of the one‐year option to extend in year j is given recursively by:
/ /
/ /
….
/ /
The chapter 4.4.1 details how the uncertainties in the oil price evolution and the oil produc‐
tion profile have been modeled as stochastic variables. Then chapter 4.4.2 explains the sim‐
ple mathematical model considered to represent the FPSO leasing contract and the actions
of the Oil Company. Lastly, chapter 4.5.3 is the application of the proposed method to a case
study.
4.4.1 MODELING UNCERTAINTY WITH STOCHASTIC VARIABLES
4.4.1.1 OIL PRODUCTION PROFILES
Modeling the possible production profiles of oil fields based on the results drilling tests and
seismic surveys is a discipline in itself. Here our purpose is quite different as an FPSO Con‐
tractor does not have realistically all such data available. So we aim at modeling the oil pro‐
duction uncertainty in a very simple way based on the expected production profile derived
from the Oil Company assessment.
Lund (1999) proposed a simple zero dimensional reservoir model while McCardle and Smith
(1998) took a different approach by modeling the decline rate as a geometric Brownian mo‐
tion. Here we propose modeling the production profile with a combination of a plateau and
a decreasing exponential function. The production uncertainty is introduced on the duration
T0 of the plateau and on the coefficient p of the decreasing exponential function which are
assumed to follow non correlated normal laws.
Production = 1 for t < T0
= exp( ‐p .(t‐T0)) for t > T0
We found out our proposed model fit relatively with some real data provided by the Oil
Company TOTAL for its CLOV field. This dataset was given by TOTAL to FPSO Contractors for
35
the design of the CLOV FPSO. The coefficients best matching the CLOV data are p = 0.27
and T0 = 5.31.
Also our model fits relatively well to the data of the Jubilee FPSO. When the duration of the
plateau is reduced to 0, the production profile is similar to the pessimistic case envisaged for
Jubilee FPSO(see Graph 7).
GRAPH 13‐ PRODUCTION PROFILE OF CLOV FPSO MODELED WITH DECREASING EXPONENTIAL FUNCTION (SOURCE: THE AUTHOR)
The Graph 14 shows 15 simulations of production profiles with T0 and p having standard
deviations of 0.2 and 0.25 respectively.
GRAPH 14 ‐ SIMULATION OF STOCHASTIC PRODUCTION PROFILES WITH EXPONENTIAL MODEL (SOURCE: THE AUTHOR)
36
4.4.1.2 OIL PRICE VARIATIONS
As discussed in chapter 2.1.2, oil price is subject to large variations on the markets. Historical
data shows a drastic increase of variability during the last decade. Future values of crude oil
spot price cannot be directly derived from crude oil historical value. There is however a high
probability that future behavior would have statistical characteristics similar to the ones of
historical data.
Several attempts have been made to model the behavior of crude oil spot price. Geman
(2005) listed several models for oil prices and other commodities. The “jump‐diffusion mod‐
el” introduced by Merton (1976) accounts for impact of news on the oil spot market that
may generate jump upwards or downwards. Gibson and Schwartz (1990) proposed a two
state variable model. It considers a classical geometric Brownian motion and couples it to a
mean reverting model for the convenience yield. More recently Geman (2000) proposed a
three‐state variable model of the following form.
dSt k Lt– lnSt St σ t Std
Where:
St is the spot price of crude oil,
d ,
σ t
Geman comments as follow: “A positive drift µ in the second equation would translate to a
rise on average of the value Lt toward which the commodity spot price St tends to revert, while this spot price itself may fluctuate significantly around Lt depending on the arrival of positive or negative news about the situation of the world(…) reserves.”. The third equation
dictates the behavior of the spot price volatility with a second mean reverting model. This
allows for volatility clustering in oil spot prices.
In the present approach, we have considered for simplicity reason using an autoregressive–
moving‐average (ARMA) model. In statistics autoregressive–moving‐average (ARMA) models
are typically applied to auto‐correlated time series data. Given a time series of data Xt, the
ARMA model is a tool for understanding and, perhaps, predicting future values in this series.
37
The model consists of an autoregressive (AR) part and a moving average (MA) part. The
model is usually then referred to as the ARMA(p,q) model where p is the order of the auto‐
regressive part and q is the order of the moving average part. Although ARMA models are
known to have disadvantages compared to other models such as GARCH for modeling oil
prices (Sadorsky 2006), we selected it for our purpose due to its simplicity. It would be inter‐
esting in future research to consider more complex and realistic models.
The Brent index corresponds to the price of North Sea crude oil and is one of the most used
reference indexes. The Brent index daily prices are highly auto‐correlated as shown in Ap‐
pendix 4. An ARMA model is therefore appropriate. By an iterative process, we find the pa‐
rameters p and q that best match the Brent historical data: p= 2 and q=2. The coefficients of
the ARMA(2,2) model are calculated using the ARIMA function implemented in the statistical
software R. The relevance of our choice is then tested with positive results using the method
of the residues (see Appendix 4).
Graph 15 represents the Brent historical data from 2002 until 2011 in black and 5 simula‐
tions of future oil prices using the ARMA(2,2) model for the period 2012 – 2020.
GRAPH 15 ‐ SIMULATION OF STOCHASTIC OIL PRICES WITH ARMA(2/2) MODEL (SOURCE: THE AUTHOR)
4.4.2 MODEL OF OIL COMPANY OPERATING A LEASED FPSO
The second part of our approach relies on modeling an Oil Company operating an oil field
with a leased FPSO. The model shall reflect the behavior of the Company when taking deci‐
sions to extend the leasing contract. Given a particular evolution of the oil price and a pro‐
duction profile, the model output will be the value of the oil field from the Company per‐
spective as first introduced in 4.4 for different configurations of FPSO leasing contract.
The model is built on the following main features:
38
A FPSO is leased for a firm period with yearly options to extend.
The FPSO oil production for each year is an input to the module. It is given as a
production profile: daily oil production changes each year.
The Oil Company expenses are the FPSO leasing rate paid daily and the operating
expenditures made of FPSO OPEX rate and other Company operating expendi‐
tures.
The Oil Company incomes come from selling the produced crude oil of a given
year at the corresponding average yearly market price.
The oil market price is an input to the module as a serial of average values for
each year.
The Oil Company is rational: it produces the field until it is not profitable any‐
more. In effect, the Oil Company exercises a one year option if it expects the pro‐
duction to be profitable during the next one year period after completion of the
firm period.
The model main function is to determine whether the Oil Company exercises one or several
one‐year options as a function of two main inputs: the production profile and the serial of
yearly average oil prices. The outputs of the module are:
/ the NPV of a leasing contract with a firm period of j years and i one‐year
options
/ the NPV of a leasing contract with a firm period of j years and no options
The main limitations and assumptions of the model are :
The operating expenditures to be paid by Company and not included in the
FPSO leasing contract have to be estimated. This is used in the model for deciding
whether an option to extend is worth to be exercised or not. In practice, an FPSO
Contractor can only have a rough estimate of such costs. A bias is therefore created
in the decision making process of the model.
Only one field is produced by the FPSO. The potential for tying‐back a satellite
field to the FPSO at a future stage is not taken into account. In practice, Oil Compa‐
nies can have such option. This would increase the value of the FPSO leasing contract
for the Company and consequently the value of the options to extend would follow a
similar trend.
The NPV calculation are done considering a single discounting rate deemed to
be equal to the WACC of a typical Oil Company. It may be adjusted to better reflect a
specific Oil Company. Independent relatively small oil companies have higher WACC
than International Major such as ExxonMobil or Shell. As the value of an option to ex‐
tend decreases with an increase of the discounting rate, this parameter should be
39
taken into account in practical case. See in Appendix 3 the Valuation report of an oil
field acquisition by Deloitte for details on the current WACC of Oil Companies .
4.4.3 CASE STUDY : LEASING CONTRACT WITH A 7 YEAR FIRM PERIOD AND 13 OPTIONS
To illustrate our method, we consider the case of an FPSO with a daily nominal output of
120,000 oil barrels. The leasing contract has a firm duration of 7 years and includes 13 one‐
year options to extend. The overall maximum field life is 20 years as all facilities (FPSO, sub‐
sea structures, etc…) are designed for this duration. The day‐rate for leasing the FPSO is
600,000 USD. It is valid for both the firm period and the optional periods. The rate also in‐
cludes a portion related to the FPSO contractor operating expenditures (OPEX). Other OPEX
costs faced by the Oil Company are estimated to be 50,000 USD/day. The WACC used is 9%.
We simulate a large number (n=500) of production profiles and the oil prices are generated
with the Monte Carlo method described previously. For each simulation, the NPVs are com‐
puted and recorded.
The NPVs are calculated for the following cases:
firm duration = 19 years with 1 one‐year option
firm duration = 18 years with 2 one‐year options
…
firm duration = 7 years with 13 one‐year options
The results of the 13 cases with 500 simulations run for each are given in the table here af‐
ter. The value of the one‐year options are computed backward from the value of the option
to extend during year 20.
/ /
/ /
…
/ /
…
/ /
40
TABLE 6 ‐ RESULTS OF MONTE CARLO SIMULATION FOR VALUATION OF OPTIONS TO EXTEND (SOURCE: THE AUTHOR)
Firm du‐ration (Years)
Number of 1 year Option
NPVo
/ NPVf
/
Present Value of the first 1‐year
option
(1) (3) (4)= (1)‐(3)‐(4)
19 1 8 311 865 8 261 581 50 284
18 2 8 233 332 8 112 250 70 798
17 3 8 421 472 8 219 934 80 456
16 4 8 208 251 7 922 540 84 173
15 5 8 326 279 7 922 032 118 536
14 6 8 349 755 7 789 309 156 199
13 7 8 256 363 7 540 944 154 973
12 8 8 392 486 7 492 492 184 575
11 9 8 414 676 7 221 147 293 535
10 10 8 233 161 6 740 052 299 580
9 11 8 339 269 6 490 036 356 124
8 12 8 228 615 5 994 740 384 642
7 13 8 348 727 5 507 789 607 063
Total 2 840 938
The numerical results lead us to the following main observation: the value of the options
decrease with their time horizon. The value of the option to extend for the year 8 is indeed
greater than the option to extend for the year 11. Several reasons can explain this phenom‐
ena. First, the discounting factor applied to the option for year 11 is smaller than the one for
the year 8. Also, as production decreases with the time, the potential income for year 8 is
therefore generally greater than for subsequent years. Lastly, the value of the first option
includes the potential for subsequent options. So the option for year 19 includes in its value
the potential of the option for year 20. The same applies for the option for year 18 with the
value of options for year 19 and 20. Recursively, the option for year 8 includes the potential
for options of year 8, 9, 10,… till 20.
5 MANAGERIAL COMMENTS
At this point of the analysis, it is worth taking some distance with the approach and the
model proposed and spending some time on discussing the findings from a managerial point
of view.
The first main element highlighted is the proven risk aversion of Oil Companies. Options to
extend leasing contracts are by nature to be exercised should the oil fields deliver as per the
41
expectations. The management of an FPSO Contractor may therefore base its investment
scenario on the most probable case: at least some of the options will be exercised. The risk
associated with such investment scenario would be inversely correlated to the risk aversion
of the Oil Company. A highly risk averse Oil Company will consider a firm period much short‐
er than the expected field production duration. The risk of the lease contract being termi‐
nated after the firm period is therefore very limited. The higher the risk aversion, the smaller
the risk taken by the FPSO Contractor.
The value of options decreases with their time horizons. Options can be of importance when
the firm period of the contract is short and represent only a fraction of total potential dura‐
tion of the lease. In such case only, the proper valuation of options is paramount. The oppo‐
site situation is for example a lease with a 15 year firm period and 5 options to extend by
one year. The time horizon of the options being after 15 years, its present value for both the
Oil Company and the FPSO Contractor would be negligible.
Another important parameter is the oil price trend. Whatever the short term trends shown
by the Brent index, the current economic environment is believed to lead toward a long
term increase of oil prices. This may not be true at a 30 year time horizon as technological
jumps may change the central role of oil in the economy. But considering 10/15 year time
horizon, oil prices are believed to follow a trend toward higher values. Hence the value of
options should increase.
Coming back to the favorable case of a short firm period, 5 years for example, the options
carry significant value. The traditional pricing approach would be to propose a fixed rate for
the firm period and then declining rate for the subsequent years calculated from the non‐
depreciated part of the FPSO from an accounting point of view. It would be interesting how‐
ever to consider a different pricing strategy. The day rate for the firm period would be fixed
at a lower level. This would increase the competitiveness against the standard approach at
the cost of a higher risk in case of termination at the end of the firm period. The rates for the
options would however be increased to compensate for the decrease of the fixed rate. We
could expect the new pricing strategy to take advantage over the standard pricing approach.
Should a FPSO Contractor go this way, its risk profile would change. It may deliver higher
profits overall but with a higher risk.
The limits of the proposed model for valuing options should however be kept in mind. Im‐
portant assumptions have been made on the production profiles, the oil price and operating
costs faced by Company. Such information is normally not available to FPSO Contractors dur‐
ing a bidding process. At best it can only make a rough estimate of the production profiles
42
and of Company operating costs. The model should therefore be seen more as an interesting
tool for analysis than as a true valuation tool.
6 CONCLUSION
This analysis of the value of options embedded in FPSO leasing contracts recognizes the
complexity of the context characterized by large underlying uncertainties and high risk aver‐
sion of the participants. Our studies have shown that in certain situation, a proper valuation
of the options would provide interesting insights to FPSO Contractor for building new com‐
petitive pricing strategies. The limits of our studies and the possible implications for manag‐
ers of FPSO Contractors have been discussed in Section 5.
Among the several directions for future research mentioned throughout the paper, we
would recommend trying a real‐option approach with a more realistic decision criteria. An‐
other direction would be the study of Oil Company risk aversion and the potential applica‐
tion to bidding strategies.
43
Glossary
FPSO Contractor or Contractor: a firm owning one or several FPSO which are leased or avail‐
able for leasing to Oil Companies. Leading FPSO Contractors are SBM and MODEC as detailed
in Appendix 1.
Majors: term commonly used to name the biggest international Oil Companies such as Exx‐
onMobil, Chevron, Shell or BP
Oil Company: a firm performing exploration activities to identify new oil fields and producing
crude oil from the discovered fields. It includes large international companies such as the
ExxonMobil, Total, national companies such as Petrobras or Saudi Aramco , and small inde‐
pendent firms that operates in limited number geographical areas and only few fields.
Subsea structures: located on the sea bed of an offshore oil field, it includes all pipelines,
manifolds, and other systems to transmit the fluid from the wells to the production plat‐
form.
Abbreviations
CAPEX: Capital Expenditures
DCF : Discounted Cash Flow
FPSO : Floating Production Storage and Offloading unit
NPV : Net Present Value
OPEX: Operating Expenditures
USD: United States Dollars
WACC: Weighted Average Cost of Capital
44
Bibliography
Brennan M. and Schwartz E.,”Evaluating Natural Resources Investment”, The Journal of
Business, (vol 58, 1985)
Extract of “Acquisition of Shell’s North West Shelf oil interests” Independent expert’s report
by Deloitte ‐ 19 February 2008
Grenadier Steven R., “Valuing lease contracts, A real‐options approach”, Journal of Financial
Economics (n°38, 1995)
Heleyette Geman, Commodities and Commodity derivatives, modeling and Pricing fo Agricul‐
turals, Metals and Energy, Wiley Finance, 2005
John C. Hull, Options, Futures and Other Derivatives, Pearson, 2012
Jorgensen P.L. and De Giovanni D., “Time Chartes with Purchase Options in Shipping: Valua‐
Lombardo Lucia,"Overview of Floating Production, Storage and Offtake (FPSO) Services
Agreements", Australian Resources and Energy Law Journal (n°468, 2003)
Myers S.C., Dill D. and Bautista A., “Valuation of financial lease contracts”, Journal of Finance
(n°31, 1976)
Sadorsky, P., “Modeling and forecasting petroleum futures volatility”, Energy Economics,
(n°28 2006)
Trigeorgis L., “Evaluating leases with complex operating options”,European Journal of Opera‐
tional Research (n°91, 1996)
45
Appendices
Appendix 1: Structure of the FPSO Lease Market supply side
Appendix 2: Factual Data on Lease Contracts
Appendix 3: Extract from Deloitte’s report on Acquisition of Shell’s North
West Shelf oil interests (2008)
Appendix 4 : Oil Prices Model with Arma(2,2)
46
Appendix 1: Structure of the FPSO Lease Market supply side
The FPSO leasing market has experienced big changes over the last ten years. From an oli‐gopolistic situation, the market has evolved toward a competitive market with more than 16 of contenders. Back in year 2000, the few active players were SBM, BW Offshore, Modec and Bluewater. There was a relatively low competition between Contractors as each had a domi‐nant position in a particular geographical market. This was clearly stated in SBM Annual Re‐port for Year 2000 : “ In the market of FPSO’s on a lease and operate basis, there is still a limited but significant number ofcompetitors for the more complex systems. These include Modec of Japan, Bluewater of the Netherlands, Maersk Contractors (Denmark), PGS of Nor‐way (the owner of Golar Nor), and Nortrans.” Now the situation is rather different. In SBM presentation to financial analyst dated November 2011, 16 FPSO Contractors owning at least 2 FPSOs are listed. Although the four leading players have significantly larger fleets, other firms such as SAIPEM or Bumi Armada are part of the competition. In addition, the excess of market capacity compared to current demand highlighted in SBM presentation has an in‐creasing effect on competition.
47
Appendix 2: Factual Data on Lease Contracts
Data collected from annual reports of BWO, MODEC and SBM for the period from 2000 until