Lecture 4
UNCONSTRAINED OPTIMIZATION: THE ONE-VARIABLE CASE
Key Concepts
- Extremum of a function
- First-order conditions
- Second-order conditions
- Concavity and convexity
1. Optimization
- To find a set of set of values of the choice variables that will yield the
desired extremum of the objective function
1.1 Objective function
- The function to be maximized or minimized; y = f(x)
- The dependent variable, y, represents the object of maximization or minimization
1.2 Choice variables
- The independent variables that can be chosen to achieve the desired objective
2. Extremum
- An extreme value - a maximum or a minimum
2.1 Local (relative) vs Global (absolute) extremum
- Local extremum:
The maximum or minimum value of a function in the neighbourhood of a point
- Global extremum:
The maximum or minimum value of a function in the range of the function
3. First-order condition (FOC)
- Consider the function y = f(x).
- The necessary condition for a relative extremum (maximum or minimum) is that the first-order derivative be zero, i.e.
f'(x) = 0.
3.1 Interpretation of the FOC
- At the highest and lowest points of a curve, the tangent to the curve at such points is horizontal.
- The slope of the curve is zero.
3.2 Critical value
- Let f'(x) = 0 at x = x0.
- x0: Critical value of x
- f(x0): Stationary value of the function
- (x,f(x0)): Stationary point
3.3 Question
- Why is f'(x) = 0 not a sufficient condition for a local maximum or minimum?
- Answer: Because f'(x) = 0 at some inflexion points.
Note: The first-order condition does not distinguish between a maximum
and a minimum.
4. Second-order condition (SOC)
If the first-order condition is satisfied at x = x0,
- f(x0) is a local maximum if f''(x0) < 0
- f(x0) is a local minimum if f''(x0) > 0
4.1 Interpretation of the SOC
Maximum:
- As you move up a curve from the left, leading to a maximum, the curve
gets increasingly flatter, i.e. the slope gets smaller and smaller.
- This means that f'' < 0.
- For example, if f' goes from 6 to 2, it means that f'' < 0.
Minimum:
- As you move down a curve from the left, leading to a minimum, the curve
gets increasingly flatter.
- However, since the slope is negative, a flattening of the curve implies
that f'' > 0.
- For example, if f' goes from -3 to -2, it means that f'' > 0.
5. Obtaining a local extremum
Consider the function y = f(x).
- Step 1: Obtain the first-order derivative of f(x).
- Step 2: Set f'(x) = 0.
- Solve for x.
- These are the critical values of x.
- But, at this point, you do not know if they yield a maximum or a minimum.
- Step 3: Obtain the second-order derivative of f(x).
- Step 4: Determine the sign of f''(x) at the critical values of x.
- If f'' < 0, the critical value corresponds to a maximum.
- If f'' > 0, the critical value corresponds to a minimum.
5.1 Example
Consider the function y = f(x) = x2 - 5x + 8.
- Obtain f'(x).
- Set f'(x) = 0. Solve for x.
- Obtain f''(x).
- What is the sign of f''(x) at the critical value? Does the critical value
yield a maximum or a minimum?
- Compute the stationary value of the function.
6. Concavity and convexity
- Refers to the curvature of a graph
- Linked to the second-order derivative
6.1 Concave curve
- A line segment joining any two points on the curve lies on or below the curve
6.2 Concave and strictly concave functions
- If f''(x) < 0 for all x, f(x) is strictly concave
6.3 Example
- f(x) = -3x2 + 5x - 2
- The function is strictly concave. Why?
6.4 Convex curve
- A line segment joining any two points on the curve lies on or above the curve
6.5 Convex and strictly convex functions
- If f''(x) > 0 for all x, f(x) is strictly convex
6.6 Example
- f(x) = 3x2 + 5x - 2
- The function is strictly convex. Why?
7. Profit maximization by firms
- Maximization problem:
Firm seeks to maximize profits by choosing optimal Q.
- Objective function:
The profit function
- Choice variable:
Output level, Q
- First-order condition:
- R'(Q) - C'(Q) = 0
- MR = MC (ah, yes, the familiar condition)
- The firm maximizes profits by producing the output level at which marginal
revenue equals marginal cost.
But how do we know that profits are maximized, not minimized, at
this point?
- Second-order condition:
Sufficent condition for maximum profits
8. Application: Profit maximization
Data for a monopolist:
- Demand curve: P(Q) = a - bQ
- Cost function: C(Q) = cQ2 + eQ + F
Questions
- What is the firm's fixed cost? Variable cost?
- Obtain the firm's revenue R(Q) and marginal revenue (MR) functions.
- Obtain the firm's marginal cost (MC) and average cost (AC) functions.
- Write down the firm's profit function in terms of Q.
- Obtain the first-order derivative of the profit function.
- Obtain the second-order derivative of the profit function.
- Under what conditions is the profit function strictly concave? Assume these conditions hold.
- Using the first-order condition, obtain the critical value of Q.
Call this Q*.
- Using the second-order condition, establish whether the critical value corresponds to a maximum or minimum.
- Obtain the firm's optimal price. Call this P*.
- What is the relationship between MR and MC at Q*?
- Sketch the demand, MR, MC and AC curves.
- Indicate the firm's optimal output and price on the graph.
- Use derivatives and graphs to show how the firm's price and output combination is affected by:
- A decrease in c.
- An increase in a.
- A decrease in F.
EC 309 home