View
2.611
Download
4
Category
Preview:
DESCRIPTION
Citation preview
Risk Assessment in the Context of Portfolio Management and Decision Making
Graeme SimpsonBernie Vining
Fiona Macmillan
AAPGInternational Conference and Exhibition,
Perth, November, 2006
Outline
• Role of risk and uncertainty assessment in resource classification
• Geological versus economic chance of success
• Is chance of success an uncertain variable?
• What is the impact on portfolio management and decision making?
Noble plans $300m notes offering
Risk, Uncertainty and Resources … SPE/WPC/AAPG
DiscoveredCommercial
DiscoveredSub-Commercial
Undiscovered
Reserves
1P 2P 3PProved Prov+Prob Prov+Prob+Poss
Contingent Resources
Low Best High
Prospective Resources
Low Best High
Project Status
On Production
Under Development
Planned for Development
Development Pending
Development On Hold
Development not Viable
Prospect
Lead
Play
Low
Risk
HighUncertainty
P90 P50 P10
Risk really matters …
• Jason McVean (2000) in “The Significance of Risk on Portfolio Selection”, SPE 62966, quote:
“The definition of risk can significantly affect portfolio selection.”
Risk and uncertainty are defined as …
• Risk (or, better, chance) … the probability of a discrete event occurring (and so one minus the risk is the chance that the discrete event will not happen).
• Uncertainty … the (usually continuous) range of possible outcomes if the discrete event happens.
Noble plans $300m notes offering
Risk, Uncertainty and Resources … SPE/WPC/AAPG
DiscoveredCommercial
DiscoveredSub-Commercial
Undiscovered
Reserves
1P 2P 3PProved Prov+Prob Prov+Prob+Poss
Contingent Resources
Low Best High
Prospective Resources
Low Best High
Project Status
On Production
Under Development
Planned for Development
Development Pending
Development On Hold
Development not Viable
Prospect
Lead
Play
Low
Risk
HighUncertainty
P90 P50 P10
Risk and uncertainty are defined as …
• Risk (or, better, chance) … the probability of a discrete event occurring (and so one minus the risk is the chance that the discrete event will not happen).
• Uncertainty … the (usually continuous) range of possible outcomes if the discrete event happens.
• But it’s actually not quite that simple …
Figure 17, page 32, from Rose (2001)
Should risk measurement be Pg / GCoS or Pe / ECoS?
Risk and uncertainty
Hypothetical Prospect – GCoS = 30%
Sample volumes for economic analysis
Minimum Economic Volume
-100
-50
0
50
100
150
200
0 20 40 60 80 100
MMBbls
NP
Vx
Zero crossing on x-axis gives Minimum Economic Volume
Portion of curve above Minimum Economic Volume gives ECoS
ECoS = 20%
So a potential source of confusion is …
• A “discrete event” can be the achievement of at least a certain value on a continuous distribution
Noble plans $300m notes offering
Risk, Uncertainty and Resources … GCoS
DiscoveredCommercial
DiscoveredSub-Commercial
Undiscovered
Reserves
1P 2P 3PProved Prov+Prob Prov+Prob+Poss
Contingent Resources
Low Best High
Prospective Resources
Low Best High
Project Status
On Production
Under Development
Planned for Development
Development Pending
Development On Hold
Development not Viable
Prospect
Lead
Play
Low
Risk
HighUncertainty
P90 P50 P10
Approx GCoS %
100
100
100
100
100
100
10-50
0-15
N/A
Noble plans $300m notes offering
DiscoveredCommercial
DiscoveredSub-Commercial
Undiscovered
Reserves
1P 2P 3PProved Prov+Prob Prov+Prob+Poss
Contingent Resources
Low Best High
Prospective Resources
Low Best High
Project Status
On Production
Under Development
Planned for Development
Development Pending
Development On Hold
Development not Viable
Prospect
Lead
Play
Low
Risk
HighUncertainty
P90 P50 P10
Approx ECoS %
100
100
90-100
50-95
20-80
0-30
5-25
0-10
N/A
Risk, Uncertainty and Resources … ECoS
Noble plans $300m notes offering
DiscoveredCommercial
DiscoveredSub-Commercial
Undiscovered
Reserves
1P 2P 3PProved Prov+Prob Prov+Prob+Poss
Contingent Resources
Low Best High
Prospective Resources
Low Best High
Project Status
On Production
Under Development
Planned for Development
Development Pending
Development On Hold
Development not Viable
Prospect
Lead
Play
Low
Risk
HighUncertainty
P90 P50 P10
Risk, Uncertainty and Resources …
CoD
GCoS
ECoS
Is “Chance of Success” an uncertain variable?
• Yes, but does it matter?
• No, if uncertainty is symmetrical, for the distribution acts as if it is the mean, and the mean = the most likely (the mode)
• Maybe GCoS estimates ought to have symmetric uncertainty ranges, so using the most likely estimate is fine
• But ECoS is a function of economic assessment, that is multiplicative calculations involving independent variables
• So the Central Limit Theorem suggests uncertainty in ECoS should tend towards being lognormally (i.e. asymmetrically) distributed
• So the most likely does not equal the mean, and hence the most likely deterministic and the probabilistic methods give different results
Hypothetical Case
Deterministic GCoS = 40%
Probabilistic Symmetrical GCoS ML = 40%
Probabilistic AsymmetricalGCoS ML = 40%
Percentiles: Forecast valuesP100 0.00P90 0.00P80 0.00P70 0.00P60 0.00P50 0.00P40 0.00P30 217.22P20 245.51P10 276.26P0 505.38
Percentiles: Forecast valuesP100 0.00P90 0.00P80 0.00P70 0.00P60 0.00P50 0.00P40 0.00P30 216.91P20 244.50P10 276.45P0 481.46
Percentiles: Forecast valuesP100 0.00P90 0.00P80 0.00P70 0.00P60 0.00P50 0.00P40 207.96P30 233.96P20 256.35P10 286.08P0 445.28
Figure 17, page 32, from Rose (2001)
Should risk measurement be Pg / GCoS or Pe / ECoS?
Risk and uncertainty
Fixed, binary event
Floating, uncertain event
Alternative Definitions …
• GCoS is the probability of a Prospective Resource maturing into a Contingent Resource
• ECoS is the probability of a Prospective Resource maturing into a Reserve
• And the chance of a Contingent Resource maturing into a Reserve is usually less than 100%. We call this the Chance of Development, CoD
And the impact on Portfolio Management and Decision Making is …
• To optimise on discovered volumes, use GCoS
• To optimise on economic volumes, use ECoS / CoD
• To optimise on value, use ECoS / CoD
• When using ECoS / CoD, take the mean of the distribution, not the most likely, or perform the calculation probabilistically, using the whole distribution
Thank you … questions?
Back-up Slides
Maximise value of a portfolio of assets by:
Optimising, relative to goals and constraints, in terms of:
– Asset content and timing
– Identification, by product type, portfolio performance, including resource maturation
Hence identify options and opportunities for improving resource management performance
Objective of Resource Management is to …Objective of Resource Management is to …
Portfolio Management Schematic
Robust, coeval,Robust, coeval,consistent,consistent,technical & technical & commercialcommercialanalysis, analysis, on asset by on asset by asset basis:asset basis:
VolumetricsVolumetricsDev PlngDev PlngEconomicsEconomicsRisk & UncertaintyRisk & UncertaintyOptionsOptionsStandalone valueStandalone value
Full underFull under--standing and standing and clear expression clear expression of organisation’sof organisation’sgoals, objectives, goals, objectives, targets and targets and constraints:constraints:
MetricsMetricsRisk measuresRisk measures
Generation Generation and analysisand analysisof possibleof possibleportfolios:portfolios:
OptimisationOptimisationIncrementalIncrementalvalue/riskvalue/risk
Incorporation of Incorporation of insights gained intoinsights gained intoportfolio managementportfolio managementstrategy formulation strategy formulation and implementation:and implementation:
Qualitative inputsQualitative inputsStrategic thinkingStrategic thinking
1 2 3 4
Standalone asset valuation
PortfolioDecision making/
Strategy formulation
Analysis
GCoS calculation …
Reservoir
Min ML Max
0.7 0.8 0.9
Source 0.4 0.5 0.6
Timing 0.3 0.4 0.5
Trap and seal 0.7 0.8 0.9
Multiply, to get ….
GCoS distribution …
Best fit is an asymmetric Beta distribution
Recommended