Week 7 Flashcards
So far (weeks 1-6) we have assumed what?
We have assumed perfect information
In the next 2 weeks we ask what happens when there is what?
Aysmmetric information
What is Aysmmetric information?
When some individuals have information that others do not
What are 2 types of aysmmetric information?
Moral Hazard, hidden action (week 7)
○ Adverse Selection, hidden information (week 8)
In the finanical crisis what was the problem?
Banks gave subprime loans to people who may have difficulty in maintaining the repayment schedule. Due to the drop of the federal fund rate. the behaviour of lenders changed dramatically. Lenders offered more and more loans to high risk borrowers, thus lending standards deteriotated.
How much did subprime loans grow in the financial crisis from 1994 to 2006?
Subprime mortgages grew from 5% of total originations ($35 billion) in 1994 to 20% ($600 billion) in 2006
What were the 2 main problems in the financial crisis?
Mortgage fraud
Predatory pricing
What is Mortage fraud?
occurs when a potential homebuyer, seller, or lender lies or omits key information that leads to a mortgage loan approval or terms that the applicant wouldn’t normally qualify to receive.
Why was Predatory pricing a problem in the financial crisis?
Lenders made loans that they knew borrowers could not afford and that could cause massive losses to investors in mortgage securities
So with Mortgage fraud and Predatory pricing what questions can you ask?
(Mortgage fraud)Why did borrowers lie on their applications ?
(Predatory pricing) Why did bankers give out loans they knew they would default?
What is Moral hazard?
Moral hazard is the opportunism by an individual that arises when another concerned individual cannot observe his/her action
What is Moral hazard also known as?
Hidden action
Someone who is insured against a loss has diminished incentives to prevent that loss, give examples of this?
driver who has collision insurance has less incentive to avoid collision
patients with medical insurance might do less to lead a healthy life, may use “too many” services
What do you usually have to pay in an event of a loss with insurance?
Deductible ( American or Excess)
Give an example of deductible or excess?
if someone has a deductible of £100 for hospitalisation, then when she goes to hospital, she pays £100 and her insurance company the remainder
What does a Deductible or excess do to Moral hazard?
It decreases moral hazard ( because people internalise paying the amount, thus decreasing moral hazard issue)
What was the rand experiment?
The RAND Health Experiment randomly assigned 5809 households to health-insurance policies that differ in their coverage
What was the effect of the RAND experiment?
- More Co-payment, implied less use of services
- Cost sharing reduced “appropriate or needed” medical care as well as “inappropriate or unnecessary” medical care.
What effect did the RAND experiment have on the MC and the poor?
RAND experiment showed little to no effect on health outcomes of free care for healthy, middle-class people
However, poor and sick participants have better health outcomes when paying lower charges
What is a negative thing about patients for example having too much health insurance?
Cost to society of people having health insurance ( negative externalities)
Where are Patnerships promient in?
Human capital intensive professional services e.g. like banking, consulting, John lewis isi a patnershsip, the question is is i t a good thing.
To model the patnership problem what Letters are we going to use to represent people?
N people in a partnership exert individual, costly effort to produce output owned by the partnership
C(ei ) is the total cost to Person i of exerting effort ei ( e.g John lewis employee selling tv, involves cost e.g. speaking, if you assert 0 effort, you have 0 cost)
MC(ei ) is the marginal cost to Person i of exerting effort ei
With the patnership problem what do we assume about the marginal cost?
We assume the MC is an increasing linear function, the more effort exerted, it becomes harder and harder, so the cost of an additional unit of effort is growing. (e.g. if you produce 10 units of effort the MC is growing
In the patnership problem what are we going to assume about the benefit of exerting effort ?
Assume that each unit of effort increases output Y by one unit . ( so if you exert one unit of effert you sell one more TV
In the patnership model what is the total marginal beneift?
the extra output per additional unit of effort, is 1. So this is fixed.
In the patnership model how does the marginal beneift and cost look like on a diagram?
We put total marginal benefit because we prefer to the total output of the whole firm. (e.g. the gain in output for John lewis,when an individual worker has put an additional unit of effort
What would be the efficient thing for the worker to do?
This is where MC = TMB
When TMB > MC the patnership is enjoying more than it costs the individual to produce that. As long as the total marginal benefitt exceeds the MC, there are more benefits that costs.
In the patnership model lets say N =2 but there is moral hazard( both patners cannot observe each others effort) What is the Y( the value of the output? )
Private marginal benefit is 1/2 as you only get half of Y, so if you produce one more unit of effort, you only get half the benefit and the other half goes to the other patner
In the patnership model lets say N =2 but there is moral hazard( both patners cannot observe each others effort) Draw the Diagram which has the the total marginal benefit and the private marginal benefit on it?
So now each person Private marginal beenfit is 1/2 even though the total marginal benefit of effort is 1 ( they enjoy half of what they produce)
The marginal cost is the same, as they share the Y equally.
What is the Equilibrum effort in an unoberserved (moral hazard) patnership that both patners put in?
For each unit of effort i make, i am getting half, so i will produce where my Private marginal benefit = marginal cost,
So as you can see rather do the efficient thing you are doing something lower, this is due to the inefficiency of the patnership
Lets generalise N, so there is patnership with N patners, with unobserved effort, what is the value of output?
So when there is N patners in a patnership, with moral hazard, what is the equlibrium effort?
So with the princple agent problem if we didn’t have moral hazard, ( so princple observes agents effort), what would be the efficient level of effort and how much would they compensate them by?
(Recall ei *=1 is the efficient level)
Optimal contract: Principal asks Agent to supply effort of ei *=1 and compensates Agent for total cost of effort,
They compensate them by paying the triangle. which is there total cost of effort only if he exerts the effort.
So the profit for the princpal is (1-S) X S, First of all explain the graph and what is the best S for the Princpal ?
When S = 0, the CEO not producing anything, when S = 1, you are giving all the share to the CEO, so the CEO works hard, but has all the profits
The highest profit for Princpal is S*
How do you find out S*?
As you are at the highest point of mountain the slope is 0, so if you go a bit to left or right ( ie increase share or decrease share, you are not changing profits)
The slope of this function (1-S) x S is the derative of the profit ( as it is a quadratic function
(1-S) X S
1s - s^2
f’(x) = 1-2s
1-2s = 0
s* =1/2
( for this simple model, not generally)
As the princpal has to pay the CEO in order for them to exert effort the profit is 1/2 X 1/2 = 1/4.
Two partners choose effort levels e1 and e2 that determine output worth Y =2e1 + 2e2 .
Suppose that the marginal cost of partner 1 is 2e1 while the marginal cost of partner 2 is e2 . Which of the following is true? a. The marginal total benefit of Partner 1’s effort is 1.
b. The marginal total benefit of Partner 1’s effort is 2.
c. The marginal total benefit of Partner 2’s effort is 1.
d. The efficient effort level of Partner 2 is the same as that of Partner 1.
e. The efficient effort level of Partner 2 is higher than that of Partner 1.
f. The efficient effort level of Partner 2 is lower than that of Partner 1
First thing to do to with this question is come up with a conjecture, as we can see the MC( how much does it cost to exert one unit of additional effort ) you see patner 1 finds it more costly to exert one unit of effort than patner 2.
So the efficicent thing to do, is for patner 1 to work less than partner 2
Patner 1 will exert less effect than patner 2, now test this conjecture.
Marginal benefit ( if i was to exert one more unit of effort by how much will output go up)
In this case marginal total benefit for patner 1 is the derative of output Y with respect to effort of Partner 1.
Y = 2ei(as this is a linear function, you drop the ei and get + 2e2 (this goes to 0)
2 = MTB for patner 1
The MTB for patner 2 is the derative of Y = 2ei + 2e2 with respect to e2 which is 2.
As we saw in lectures this is computed where marginal total benefit is equal to marginal cost.
So B AND E is correct
What is the alternative way to find the total marginal benefit of the patners without using deratives?
Two partners choose effort levels e1 and e2 that determine output worth Y =2e1 + 2e2 .
Suppose that the marginal cost of partner 1 is 2e1 while the marginal cost of partner 2 is e2 . Which of the following is true? a. The marginal total benefit of Partner 1’s effort is 1.
b. The marginal total benefit of Partner 1’s effort is 2.
c. The marginal total benefit of Partner 2’s effort is 1.
d. The efficient effort level of Partner 2 is the same as that of Partner 1.
e. The efficient effort level of Partner 2 is higher than that of Partner 1.
f. The efficient effort level of Partner 2 is lower than that of Partner 1
So the first bit of the equation would be the Y after she adds one more unit of effort - the output that would exist if she didn’t exert that additional effort for which is 2ei-2e2.
Take a look at this figure which depicts changes in the price of oil between 1977-1994 as well as changes in CEO compensation of Oil Industry CEO’s at the same period. I want to reflect on these graphs while thinking about the insights and lessons we learned from Video 7.3 about the Principal-agent model and optimal compensation. Is the data presented in the figure consistent with optimal compensation for the CEOs or is there something else going on?
Helpful pieces of information as you look at this figure: The sharp decline in crude oil price at the end of 1985 was caused by Saudi Arabia’s decision to reform its petroleum policy and to increase production. The large oil price increase between 1979 and 1981 is usually attributed to an internal policy change by OPEC.
In the princple agent problem we saw that output should be tied to the exertion effort of the CEO etc, here in the Picture we can see that it seems the oil industry CEO’s compensation is tied with something that effects the whole industry and not related to their effort.
You have a great idea for a new revolutionising app. The app will generate a revenue of Y millions of pounds, depending on the effort, e, of coding and design. The relation between Y and e is given by Y=1+3e. The problem is that you do not know how to write code or design apps. A friend of yours is an excellent coder and designer and she can put in the effort e. Effort is costly for your friend, the marginal cost of her effort is equal to 2e.
b) Suppose you can observe the level of effort, e, your friend puts into the project. Can you achieve the efficient level of output for your app? What will you end up paying your friend if you just want to compensate them for their effort cost?
Yes, if I can observe the effort level of my friend, then I can achieve the efficient level of output as I can condition the payment on her exerting the effort 1.5 we computed above.
I will have to pay my friend to compensate for the cost she incurred, the cost is the area of the triangle in the drawing below, as we did in lectures. This is the area under the marginal cost curve which corresponds to the total cost.
You have a great idea for a new revolutionising app. The app will generate a revenue of Y millions of pounds, depending on the effort, e, of coding and design. The relation between Y and e is given by Y=1+3e. The problem is that you do not know how to write code or design apps. A friend of yours is an excellent coder and designer and she can put in the effort e. Effort is costly for your friend, the marginal cost of her effort is equal to 2e.
d. Suppose you cannot observe the level of effort, e, your friend puts into the project. Suppose you offer your friend a 25% share in revenue. How much effort will your friend put into the project? How much money will you make?
Always put definitions
To calculate the effort the friend will exert we equate marginal private benefit to marginal cost.
Marginal private benefit is the additional increase in her benefit by an additional unit of effort
You have a great idea for a new revolutionising app. The app will generate a revenue of Y millions of pounds, depending on the effort, e, of coding and design. The relation between Y and e is given by Y=1+3e. The problem is that you do not know how to write code or design apps. A friend of yours is an excellent coder and designer and she can put in the effort e. Effort is costly for your friend, the marginal cost of her effort is equal to 2e.
e. You are not happy with these calculations. Someone tells you that you can take an online course and learn how to code and design. If you do so, you yourself can start coding and designing at a personal marginal cost of effort of 3e. How much effort would you put into the project if you took the online course? Comparing to what you computed in (d) should you take the online course?
Mc = 3e
Y = 1+3e still the same
Marginal total benefit is the additional increase in revenue by an additional unit of effort. (it is the derivative of Y with respect to e). In this example this is 3.
f’(x) = 3
Increase in Y due to an increase of one unit in effort = 1+3(e+1) - (1+3e) = 3
So equating the two we get: 3=3e i
mplies that e* = 1.( the amount of effort you would put inyo the project given MC is 3e.
This means that the efficient revenue is: Y * = 1 + 3 × 1 = 4
