Skip to content

Measurement of Risk

Analysis of Risk & Uncertainty


Description & Measurement of Risk

Risk

  • Variability in actual returns wrt \(E(r)\), in terms of cashflows1.

Sensitivity Analysis (Absolute measure)

  • How sensitive estimated project parameters are to estimation errors
    • Parameters: cash flows, cost of capital, economic life
    • Estimates made under three assumptions: pessimistic, most likely, optimistic
  • Example 1: Calculate NPV of Project \(X,Y\) for each possible cash flow (worst, most-likely, best) using sensitivity analysis
    • NPV determination
    • conservative and risk-taking attitude towards risk

Assign probabilities

  • Drawback of sensitivity analysis: Doesn't disclose the chances of occurrence of these variations
  • Example 2: Find expected return on the project. PV of expected monetary values.

Scenario Analysis (Absolute measure)

  • Evaluate the impact on the project's profitability of simultaneous changes in more than one variable at a time (inflows, outflows, cost of capital)
    • Ask operating manager (production, sales, personnel)
      • worst-case (high fixed costs, high VC, low Selling price, low sales volume, higher cost of capital) etc.
      • best-case - vice versa
    • NPV guides the decision and helps assess the risk
    • Highly unlikely (both the scenarios), but useful \(\implies\) If NPV positive in worst case, then project is worth accepting.
    • Limited Usefulness: Considers few discrete outcomes, infinite number of possibilities exist

Simulation

  • get a feel of the risk (statistics based)
  • Apply predetermined probability distributions + random numbers \(\implies\) estimate risky outcomes

Precise Measures of Risk: Standard Deviation and CoV

Standard Deviation (Absolute measure)
\[\sigma = \sqrt{\sum_{i}^n P_{i} (CF_{i}- \bar{CF})^2 }\]
CoV (Relative measure)
\[V = \dfrac{\sigma}{\bar{CF}}\]
  • Example 4: Compare the risk of projects on the basis of SD and CoV

Risk Evaluation Approaches

1. Risk-adjusted Discount Rate Approach (RAD)

  • relatively risky \(\implies\) relatively high discount rates
  • Accept-reject Decision
    • NPV and IRR approach
    • -ve NPV \(\implies\) Reject
    • internal rate of return, \(r \gt\) risk-adjusted rate \(\implies\) Accept
  • Example 5: Evaluate project (accept/reject)
  • Evaluation (+1 -3)
    • + Simple to calculate
    • - How to determine risk-adjusted discount rate?
    • - Doesn't use direct info available distribution of expected future CF, it adjust \(r\) instead of CF
    • - Risks compounds overtime, not theoretically desirable into practice. Discounting process should only consider time value and not risk.

2. Certainty-Equivalent Approach

  • We adjust the expected cash flow directly.
  • Risk-adjustment factor is expressed in terms of a certainty-equivalent coefficient \(a_{t}\)
    • Based on firm's utility preference: \((12000) \sim (20000,0.6)\)2
    • This coefficient \(a = 0.6\) when multiplied to a risky CF gives us a riskless CF
    • \(a\propto \dfrac{1}{\text{Risk}}\)

Accept-Reject Rule

\[NPV = \sum_{i=1}^n \dfrac{a_{i}CFAT_{i}}{(1+i)^i} - CO\]
  • Illustration using Example 5
  • Evaluation (+2 -1)
    • + simple to calculate
    • + modifies CF that is subject to risk

3. Probability Distribution Approach

  • Dependent CF
  • Independent CF
\[NPV = \sum_{t=1}^n \dfrac{\bar{CF}_{t}}{(1+i)^t} - CO\]
\[\sigma(NPV) = \sqrt{ \sum_{t=1}^n \dfrac{\sigma_{t}^2}{(1+i)^{2t}}}\]
  • You are essentially finding \(Var\left(\dfrac{CF}{(1+i)^t}\right)\)

  • Where \(\sigma_{t} = \sqrt{ \sum (CF_{jt} - \bar{CF_{t}})^2 \cdot P_{jt}}\)

Problem Checklist

  • Example 6: Expected Cashflow and standard deviation
  • Normal distribution: (\(\leq 0, \gt 0, a \lt CF \lt b\))
  • Example 7: Find probabilities (i) \(\leq 0\); (ii) \(\gt 0\); (iii) \((25,45)\); (iv) \((15,30)\)
  • Example 8:
    • (i) \(\mu = E(NPV)\);
    • (ii) \(\sigma_{NPV}\) ;
    • (iii) \(P(NPV\dots)\) :
      • (a) \(\leq 0\); (b) \(\gt 0\); (c) \(\geq \mu\)
    • (iv) Profitability Index of \(\mu\);
    • (v) \(P(PI < 1)\)

4. Decision-tree Approach

Problem Checklist
  • Example 9: Investment outlay with 9 distinct possibilities (shown by decision tree)
Notation
  • Decision is squared, chance is circled
  • Index like so \(D_{11}, D_{22}\) for Decision \(D_{1}\)

At \(D_{1}\), choose \(D_{11}\) and wait for outcome at \(C_{1}\)


Risk & Real Options

Types of Option

Growth
  • expand (demand > expectations)
  • open new doors, if successful \(\implies\) invest cash
  • embedded in capital budgeting projects
Abandonment
  • abandon/terminate/shutdown prior to its expected economic useful life
  • minimize firm's losses
  • projects with abandonment value, lower the project's risk by limiting downside losses and improve NPV
  • sell off some capacity and put to other use
  • variant: suspend (temporarily)
    • mineral extractions
    • extraction costs > selling price
Timing
  • optimal time postpone
  • accelerate or slow process of implementation
  • negative NPV today doesn't mean forever
  • traditional contrast, now or never
Flexibility
  • accept multiple inputs and…

Other Topics

PC

  • Sources and Perspectives on Risk
  • Break-even Analysis
  • Hillier Model
    • Uncorrelated Cash Flows
    • Perfectly Correlated Cash Flows
  • Simulation Analysis

Limitations Of Sensitivity, Scenario, Simulation & DT Analysis

  • Corporate Risk Analysis
  • Managing Risk
  • Project Selection under Risk3

Ross

  • Monte Carlo Simulation

  1. In hard currency coins. Not in terms of goodwill and all those abstract concepts. 

  2. Here \(0.6 = \dfrac{12000}{20000}\) 

  3. More than it was discussed in the [[Risk Analysis & Cap Budgeting K&J]]