Agribusiness Analysis and Forecasting

Agribusiness Analysis and Forecasting

Ranking Risky Alternatives Overview

  • Purpose: To determine the best scenario from multiple simulations.

  • Methods Evaluation: Discussion of various methods for ranking scenarios based on their effectiveness.

Presentation of Simulation Results

  • Decision Making Tools:
      - Tables of Summary Statistics: Provide baseline data for informed choices.
      - Probabilities for Key Output Variables (KOV): Probability distributions help assess risk.
      - Probability Density Functions (PDFs) and Cumulative Distribution Functions (CDFs): Visualize the likelihood of different outcomes.
      - StopLight Charts: Communicate risk and performance visually.
      - Stochastic Efficiency with Respect to a Function (SERF): Compare alternatives by assessing expected utility under different risk preferences.

Ranking Decision Criteria

  • Expected Outcomes vs. Risk: Decision makers (DMs) consider both averages and risk when ranking alternatives.

  • Methods without Utility:
      - Simple and Easy: Quick comparison but may overlook individual risk preferences.

  • Utility-Based Methods:
      - Advantages: Incorporate risk preferences, leading to consistent rankings, but require more information.

Easy-to-Use Ranking Procedures

  • Mean Only: Select the scenario with the highest mean outcome, disregarding risk.

  • Minimize Risk: Choose the scenario with the lowest standard deviation (Std. Dev.), ignoring potential returns.

  • Mean Variance: Choice often favors scenarios in the lower right quadrant but struggles with non-normal distributions.

  • Worst Case: Based on the highest minimum return, albeit observing it with a very low probability (e.g., 1/500).

  • Best Case: Considers the best possible outcome with the same low probability, also ignoring overall risk.

  • Probabilities of Target Values: Reports the likelihood of achieving targets and falling below minimums, promoting transparency.

Risk Ranking Methods Using Utility

  • Concept of Lottery: An example with a 50% probability of winning $10 and $0:
      - Expected Value: E(V)=0.5imes10+0.5imes0=5E(V) = 0.5 imes 10 + 0.5 imes 0 = 5
      - Risk Profiles:
        - Risk-neutral DM values at $5.
        - Risk-averse DM pays $< 5, e.g., $4.     - Risk-loving DM pays $> 5, e.g., $7.
      - Certainty Equivalent (CE): The amount DM is indifferent to the lottery.
      - Risk Premium (RP): Measures risk cost, defined as RP=E(V)−CERP = E(V) - CE, with an example for risk-averse behavior:
        - RP=5−4=1RP = 5 - 4 = 1

Practical Application of Utility in Agriculture

  • Scenario Analysis: Considering uncertain crop returns:
      - Good year: 10,00010,000 profit (50%)
      - Bad year: 00 profit (50%)
      - Expected Value: E(V)=0.5imes10,000+0.5imes0=5,000E(V) = 0.5 imes 10,000 + 0.5 imes 0 = 5,000
      - Risk Profiles similar as before, leading to CE calculations:
        - Example: Risk-averse DM might value at CE=4,000CE = 4,000 leading to RP=1,000RP = 1,000.

Certainty Equivalent Evaluations

  • Everyday decisions use CE implicitly:
      - Example with boxes containing different values: $0, $5, $50,000, and $250,000.
      - Expected Value Calculation:
        - E(X)=rac0+5+50,000+250,0004=75,001.25E(X) = rac{0 + 5 + 50,000 + 250,000}{4} = 75,001.25
      - If offered $65,000 to exit, accept if $CE < 65,000.
      - If indifferent at $40,000, then CE=40,000CE = 40,000, reflecting risk aversion degree as E(X)−CEE(X) - CE.

Utility Function and Risk Aversion

  • Black Utility Function:
      - Normal Risk-Averse Person: Shape indicates risk aversion.
      - Red Line: Represents a risk lover.
      - Yellow Line: Extremely risk-averse individual.
      - Blue Line: Risk neutral individual.

Ranking Risky Alternatives with Utility

  • Key Assumptions:
      - More is preferred to less.
      - Ranking based on CE derived from specified utility functions, considering the DM’s risk preferences and rationality.

Simulation to Ranking Process

  • Simulation Outcomes: Results from different scenarios assessed.

  • Utility Function Types:
      - Power (CRRA)
      - Negative Exponential (CARA)

  • Certainty Equivalent Calculation:
      - CE=U−1(E[U(X)])CE = U^{-1}(E[U(X)])

  • Ranking: Higher CE indicates preferred choice.

Utility-Based Ranking Tools

  • Certainty Equivalents (CE): Measures risk-taking based on utility.

  • Stochastic Efficiency with Respect to a Function (SERF):
      - Requires utility function and risk aversion characterization.
      - Discrepancy in risk aversion logic between CARA and CRRA.

Linking Risk Aversion Types

  • Absolute Risk Aversion Coefficient (ARAC) is a function of wealth and risk profiles.

  • Steady Risk: CARA maintains dollar-value risk constant versus CRRA which scales proportionately with wealth.

  • Choosing Utility Functions: Use CARA for short-term analyses, CRRA for long-term decisions involving wealth changes.

Parameterizing Risk Aversion: RRAC Benchmarks

  • RRAC Value Interpretations:
      - 0.0: Risk neutral, maximizes expected value.
      - 0.5 to 4.0: Increasing risk aversion from slightly to extremely averse focusing on lower tail outcomes.

  • Risk Sensitivity: High RRAC indicates increasing emphasis on downside risk in rankings.

Practical Applications and Sensitivity in Rankings

  • Sensitivity of Rankings: Higher RRAC means greater weight on undesirable outcomes; lower values are closer to average returns.

  • Behavioral Insight: Understand risk selection based on applicable utility function type (CARA vs. CRRA).

Ranking Methods Recap

  • First-Degree Stochastic Dominance (FSD): Dominance without preference specifications.

  • SERF: Applies utility function to compute CEs across various RRACs.

  • StopLight Charts: Visual aids aiding comprehension of risks in rankings.

FSD and Probability Distribution Analysis

  • FSD Conditions:
      - One distribution consistently yields better outcomes across probabilities.
      - No utility function needs to be applied.

  • CDF Intersection: Crossings induce ambiguity in ranking and necessitate clear methodology to handle.

Utility Function Specifications

  • CARA Utility Function: u(w)=1−e−extλwu(w) = 1 - e^{- ext{λ}w} indicating constant dollar risk acceptance.

  • CRRA Utility Function: u(w)=racw1−<br>ho−11−<br>hou(w) = rac{w^{1-<br>ho} - 1}{1 - <br>ho} where risk tolerance grows relative to wealth.

SERF Methodology

  • SERF Workflow: Capture utility across ranges of RRAC values to rank alternatives.

  • FLOSim Application: Computes CEs by assessing utility curves and visualizing outcomes for comparison.

Interpreting SERF Results

  • Graphical Analysis: Highest CE curve preferred among alternatives.

  • Indifference Point: Where CE curves intersect, identifying breakeven RRAC values.

SERF Applications in Decision Making

  • Annual and Multi-Year Decisions: Adapt utility functions based on the timeframe and nature of the financial evaluations.

Scenario Modeling

  • Scenario Definitions: Each alternative management approach serves as a scenario, preserving risk and outcome consistency.

  • Model Consistency: Essential to ensure all parameters align for valid comparisons across scenarios.

FLOSim Simulation Functionality

  • Scenario Simulation: Use the equals function to define alternatives within FLOSim ensuring values maintain constant definitions.

Visualization Tool: StopLight Chart

  • Communicates risks and performance visually, assisting decision-makers in assessing outcomes clearly.