1 Civil Systems Planning Benefit/Cost Analysis Scott Matthews Courses: / / Lecture 16
and Admin Project 1 - avg 85 (high 100) Mid sem grades today - how done?
and Recall: Choosing a Car Example CarFuel Eff (mpg) Comfort Index Mercedes25 10 Chevrolet283 Toyota356 Volvo309
and “Pricing out” Book uses $ / unit tradeoff Our example has no $ - but same idea “Pricing out” simply means knowing your willingness to make tradeoffs Assume you’ve thought hard about the car tradeoff and would trade 2 units of C for a unit of F (maybe because you’re a student and need to save money)
and With these weights.. U(M) = 0.26* *0 = 0.26 U(V) = 0.26*(6/7) *0.5 = U(T) = 0.26*(3/7) *1 = U(H) = 0.26*(4/7) *0.6 = Note H isnt really an option - just “checking” that we get same U as for Volvo (as expected)
and MCDM - Swing Weights Use hypothetical combinations to determine weights Base option = worst on all attributes Other options - “swings” one of the attributes from worst to best Determine your rank preference, find weights
and Add 1 attribute to car (cost) M = $50,000 V = $40,000 T = $20,000 C=$15,000 Swing weight table: Benchmark 25mpg, $50k, 3 Comf
and Stochastic Dominance “Defined” A is better than B if: Pr(Profit > $z |A) ≥ Pr(Profit > $z |B), for all possible values of $z. Or (complementarity..) Pr(Profit ≤ $z |A) ≤ Pr(Profit ≤ $z |B), for all possible values of $z. A FOSD B iff F A (z) ≤ F B (z) for all z
and Stochastic Dominance: Example #1 CRP below for 2 strategies shows “Accept $2 Billion” is dominated by the other.
and Stochastic Dominance (again) Chapter 4 (Risk Profiles) introduced deterministic and stochastic dominance We looked at discrete, but similar for continuous How do we compare payoff distributions? Two concepts: A is better than B because A provides unambiguously higher returns than B A is better than B because A is unambiguously less risky than B If an option Stochastically dominates another, it must have a higher expected value
and First-Order Stochastic Dominance (FOSD) Case 1: A is better than B because A provides unambiguously higher returns than B Every expected utility maximizer prefers A to B (prefers more to less) For every x, the probability of getting at least x is higher under A than under B. Say A “first order stochastic dominates B” if: Notation: F A (x) is cdf of A, F B (x) is cdf of B. F B (x) ≥ F A (x) for all x, with one strict inequality or.. for any non-decr. U(x), ∫U(x)dF A (x) ≥ ∫U(x)dF B (x) Expected value of A is higher than B
and FOSD Source:
and FOSD Example Option A Option B Profit ($M)Prob. 0 ≤ x < ≤ x < ≤ x < ≤ x < Profit ($M)Prob. 0 ≤ x < 50 5 ≤ x < ≤ x < ≤ x < ≤ x < 250.1
and
and Second-Order Stochastic Dominance (SOSD) How to compare 2 lotteries based on risk Given lotteries/distributions w/ same mean So we’re looking for a rule by which we can say “B is riskier than A because every risk averse person prefers A to B” A ‘SOSD’ B if For every non-decreasing (concave) U(x)..
and SOSD Example Option A Option B Profit ($M)Prob. 0 ≤ x < ≤ x < ≤ x < ≤ x < Profit ($M)Prob. 0 ≤ x < ≤ x < ≤ x < ≤ x < ≤ x < 250.1
and Area 2 Area 1
and SOSD
and SD and MCDM As long as criteria are independent (e.g., fun and salary) then Then if one alternative SD another on each individual attribute, then it will SD the other when weights/attribute scores combined (e.g., marginal and joint prob distributions)
and Reading pdf/cdf graphs What information can we see from just looking at a randomly selected pdf or cdf?