114 年 國立臺灣大學經濟學系碩士班《個體經濟學》
Notes on preference:
Weak preference relation :
means "the consumer wants at least as much as " or "the consumer weakly prefers to "
Indifference preference relation :
, which reads "the consumer is indifferent between and "
Strong preference relation :
where means not .
Notes on utility function:
A utility function can represent/characterize preference means:
第 1 題10 分
Suppose you could vary prices and income at any positive level and you could observe individuals' corresponding consumption choices . In each following option, you need to distinguish between two preferences. Please choose the option(s) in which two preferences are distinguishable according to the observational data mentioned above.
(A) Preference A1 is described by a utility function:
Preference A2 is described by a utility function:
(B) Preference B1 is described by a utility function:
Preference B2 is described as:
(C) Preference C1 is described by a utility function:
Preference C2 is described by a utility function:
(D) Preference D1 is described by a utility function:
Preference D2 is described by a utility function:
(E) Preference E1 is described by a utility function:
Preference E2 is described by a utility function:
登入後即可作答並保存紀錄。
核心觀念
本題考察「偏好是否能由消費選擇資料辨識」。即使兩個效用函數形式不同,只要在所有正的價格與所得下,所產生的最適消費選擇都相同,便無法僅靠觀察需求加以區分。
消費者在預算限制
下選擇效用最大的消費組合。判斷兩種偏好是否可辨識,重點是比較兩者在所有可能的 下,是否存在不同的最適選擇。
(A) 與
的最適選擇
其中 。因為效用對兩種商品皆嚴格遞增,消費者會花完全部所得:
利用邊際替代率等於價格比:
因此
只要選擇價格使
則 的最適點就不在 的對角線上。
的影響
在所有 的消費組合上額外增加 7:
選擇所得 足夠小時,對角線上的額外 7 可能大於偏離對角線所能增加的 Cobb-Douglas 效用,因此 會選擇預算線上的對角線點。
例如,令價格滿足 的最適點不在對角線上。預算線與 的交點為
當 足夠小時,該點的效用為
其中的 7 會使其超過所有非對角線消費組合的效用。
因此存在一組價格與所得,使 選擇非對角線點,而 選擇對角線點。
結論: 與 可辨識。
(B) 與
的偏好由
決定。
的規則是:
- Cobb-Douglas 乘積較大者較受偏好;
- 若乘積相同,且一個組合位於 、另一個不在 ,則偏好對角線上的組合;
- 若兩者乘積相同且皆不在對角線上,則無差異。
然而,在標準正價格、正所得的預算問題中, 的最適點是唯一的。其需求為
由於最適點唯一,預算集合中不存在另一個具有相同 Cobb-Douglas 乘積的不同最適點。因此, 的「同效用時偏好對角線」規則沒有機會改變實際選擇。
換言之, 雖然改變了某些無差異曲線上的排序,但不會改變每一組價格與所得下的最適消費選擇。
結論: 與 不可辨識。
(C) 與
考慮一般形式:
由於 ,效用函數嚴格凹,最適解為內部解。
邊際效用為
一階條件為
第 2 題10 分
Consider potential firms could produce the same good (; is an extremely large positive integer). Each firm uses labor and capital to produce the good with the same production function:
In the labor market and capital market, all firms are price takers. The price of labor is , and the price of capital is . However, the fixed cost varies across firms. Firm 's fixed cost is:
Finally, we assume that the market of the output is competitive without restrictions on entry and exit. Please consider only the case in the long-run.
(A) If stays the same but increases, firms will use relatively more labor in production ( is higher).
(B) Every firm has the same marginal cost function which is linear.
(C) Because of competition, firms will earn 0 profit.
(D) Suppose . If the price of the output is larger than 4 and less than 6, there will be 2 firms staying in the market.
(E) Suppose . If in the equilibrium, aggregate quantity of the good is 9, there are 3 firms staying in the market.
登入後即可作答並保存紀錄。
核心觀念
- 里昂提夫生產函數(Leontief Production Function)之成本極小化:
對於固定比例型生產函數 ,廠商最適要素投射比例滿足 ,即 且 。投射比例完全由技術決定,不受要素相對價格 影響。 - 異質固定成本與完全競爭市場之長期均衡:
廠商 的總成本函數包含可變成本與異質固定成本:
在價格接受的市場中,廠商依據 決定產量。由於各廠商的固定成本 隨 遞增,邊際成本 相同,但平均成本 隨 增加而上升。 - 廠商進退場決策與均衡廠商數:
在長期,廠商維持在市場上的條件為利潤 ,即價格 。均衡時,在場內營運的最高編號廠商 其利潤必須非負(),而下一家廠商 進場利潤為負()。
解題方法
1. 成本函數推導
由生產函數 ,可得最適要素需求為:
因此,相對要素投入比例為:
廠商 的長期總成本函數(TC)、邊際成本函數(MC)與平均成本函數(AC)分別為:
2. 個別廠商的最適產量與利潤
當市場價格為 時,廠商由 決定產量:
注意:最適產量 僅取決於價格與要素價格,與廠商編號 無關,因此所有在場內的廠商生產相同的產量 。
廠商 的最大利潤為:
3. 當 時的具體數值代入
代入 :
選項分析
- (A) 錯誤
里昂提夫生產函數的要素投入比例完全由技術決定(固定為 ),滿足 。即便 上升而 不變,兩要素之間完全不可替代,要素比例保持為 ,不會使用相對較多的勞動。
第 3 題10 分
Suppose there are smokers and non-smokers in a room ( is a positive integer), and each of them has dollars. A smoker decides his consumption of cigarettes ( can be non-negative real numbers). We denote as the price of one cigarette. Smoker only cares about his consumption of cigarettes and his wealth left. His utility would be:
A non-smoker never consumes cigarettes, and he hates the second-hand smoke in the room. His utility would be:
(A) If every individual is selfish and there's no transfer or bargaining, the equilibrium would be efficient when .
(B) For some , it's possible to find efficient outcome in which , , and for some .
(C) Given symmetric efficient outcome (), the individual-level consumption of cigarettes () should decrease as increases. (Please only consider interior solution.)
(D) Given symmetric efficient outcome (), the aggregate consumption of cigarettes () should decrease as increases. (Please only consider interior solution.)
(E) Given symmetric efficient outcome (). If is large enough (still finite), it's possible to find . (You need to consider the possibility of corner solution)
登入後即可作答並保存紀錄。
本題考查外部性(Externality)與柏拉圖效率(Pareto Efficiency)。題目設定 個吸菸者與 個非吸菸者,吸菸會對非吸菸者產生負外部性(二手菸)。
核心觀念
-
競爭均衡/極私利結果(Selfish / Market Outcome):
在沒有交易或談判(no transfer or bargaining)的私利條件下,個體 僅極大化自身效用:
一階條件(FOC)為:
私利下的個體消費量為 ,忽略了對非吸菸者造成的外部成本。 -
柏拉圖效率條件(Social Planner's Problem / Samuelson Condition):
極大化社會福利總和 :
對特定吸菸者 的消費量 求導(若為內角解 ):
即邊際效益等於社會邊際成本:
解題方法與選項分析
(A) 錯誤
- 推導與說明:
當 時(1 個吸菸者,1 個非吸菸者):- 私利均衡條件:。
- 效率條件:。
由於 (只要 ),且 ,只要效率解 ,私利均衡消費量 會高於效率消費量 。因此私利結果存在過度消費(Over-consumption),競爭均衡不具效率。
(B) 正確
- 推導與說明:
- 效用函數對財富 為準線性(Quasi-linear),轉移支付(Lump-sum Transfer)不影響香菸的效率消費量 。
- 若存在某種財富轉移方案,吸菸者 與 均消費效率數量 ,且獲得足夠多的財富補償轉移 ,則其最終效用為:
(因為 在財富 足夠大時顯然成立)。 - 此外,效率解並不要求對稱的香菸消費(例如總效率消費量 可在吸菸者間任意配置,只要邊際效用相等),故存在 且 的柏拉圖有效結果。
(C) 錯誤
- 推導與說明:
在對稱內角解下,,總消費量 。效率一階條件為:
對 進行隱函數微分(Implicit Differentiation):
對兩邊關於 求導:
整理可得:
第 4 題10 分
In this question, consider an exchange economy with two individuals () and two goods (). Suppose is individual 's consumption and is individual 's initial endowment. Individuals' utility functions are:
Initial endowment follows:
(A) Contract curve is a smooth curve.
(B) It's possible to have with as general equilibrium (Walrasian equilibrium).
(C) It's possible to have with as general equilibrium (Walrasian equilibrium).
(D) The result of the First Welfare Theorem always holds in this economy.
(E) The result of the Second Welfare Theorem always holds in this economy.
登入後即可作答並保存紀錄。
這是一道關於**一般均衡理論(General Equilibrium Theory)與福利經濟學基本定理(Welfare Theorems)**的經典多選題。以下為詳細解析:
核心觀念
-
契約曲線(Contract Curve)與帕雷托效率(Pareto Efficiency):
- 帕雷托效率條件要求兩人的邊際替代率(MRS)相等,即 (若均為內生解/可導)。
- 個人 1 的效用函數為 Cobb-Douglas 形式:(其中 ),其邊際替代率為:
- 個人 2 的效用函數為線性形式:,對商品 的邊際效用為 0()。因此個人 2 的邊際替代率 。
- 由於個人 2 完全不在意商品 ( 無法帶來任何效用),只要將經濟體中所有的商品 全部分配給個人 1(即 ,),就能在不損害個人 2 的情況下,使個人 1 的效用達到極大。因此,所有的帕雷托效率配置均滿足 。
-
瓦爾拉斯均衡(Walrasian Equilibrium / General Equilibrium):
- 均衡價格 與配置 須滿足:
- 個人效用極大化(在預算限制下)。
- 市場清算條件(Market Clearing): 且 。
- 若 且 ,由於個人 1 對商品 有嚴格正偏好(Cobb-Douglas),面對免費的商品 ,個人 1 對 的需求量將會趨近於無窮大(),導致商品 市場無法清算。
- 若 且 ,個人 2 面對免費的商品 (且 ),對 的需求量將會趨近於無窮大(),導致商品 市場無法清算。
- 均衡價格 與配置 須滿足:
-
福利經濟學定理(Welfare Theorems):
- 福利第一定理(First Welfare Theorem):若偏好滿足**局部非飽和性(Local Non-Satiation, LNS)**且無外部性、完全競爭,則競爭均衡必為帕雷托效率配置。
- 福利第二定理(Second Welfare Theorem):若偏好為凸性(Convex)、滿足局部非飽和性,且存在適當的重分配,則任何帕雷托效率配置均可透過競爭均衡實現。
- 本題中,個人 1 的偏好為嚴格凸且嚴格單調(滿足 LNS);個人 2 的偏好 雖然對 不敏感,但只要 增加效用就增加,故亦滿足局部非飽和性與凸性。
解題方法與推導
1. 契約曲線(Contract Curve)的推導
在艾吉沃斯盒(Edgeworth Box)中,總資源為 。
- 個人 2 的無差異曲線在 平面上是一組垂直線(因為 ,效用僅取決於 )。
- 個人 1 的無差異曲線為標準的凸向原點曲線。
- 若 (即 ),由於個人 2 拿著 無法產生任何效用,若將 全部無償轉移給個人 1,個人 1 的效用嚴格上升( 增加),而個人 2 的效用保持不變( 不變)。這代表任何 的配置都存在帕雷托改善(Pareto Improvement)。
- 因此,所有的帕雷托效率配置(Pareto Efficient Allocations)必須滿足:
第 5 題10 分
Consider a parent with a child. The probability of the child being sick is . Given that the child is sick, the probability that the child passes away is also . Suppose that the parent has as initial wealth. If the child gets sick, the parent will definitely spend saving the child because the parent cares about the child. The parent's utility function is when the child is alive, and the parent's utility function is ( means natural log). means that losing the child hurts the parent. The parent is facing either one of the following two cases of insurance provision.
Case 1-insurance on child's being sick:
The parent can purchase units of the insurance. When the child gets sick, the insurance company will pay dollars to the parent. The unit price of insurance is which makes the insurance company earn 0 profit in expectation.
Case 2-insurance on child's death:
The parent can purchase units of the insurance. When the child passes away, the insurance company will pay dollars to the parent. The unit price of insurance is which makes the insurance company earn 0 profit in expectation.
(A) is and is
(B) If losing the child hurts the parent more (smaller ), the parent will purchase more units of insurance under both cases.
(C) Under Case 1, the parent will purchase some positive units of insurance to cover the medical cost ().
(D) Under Case 2, the parent will purchase some positive units of insurance to cover the medical cost ().
(E) The parent will purchase more units of insurance under Case 2 than under Case 1.
登入後即可作答並保存紀錄。
核心觀念
本題考查不確定性下的抉擇(Choice under Uncertainty)、期望效用極大化(Expected Utility Maximization)以及公平保險(Actuarially Fair Insurance)。
-
事件機率與狀態空間(State Space):
- 狀態 1 ():孩子健康(Healthy),機率 。
- 狀態 2 ():孩子生病但存活(Sick & Alive),機率 。
- 狀態 3 ():孩子生病且離世(Sick & Dead),機率 。
-
公平保險定價(Zero-Profit Insurance Premium):
- 保險公司的期望利潤為 0,意味著保險單位費率 等於損害發生之客觀機率 。
-
期望效用函數:
- 孩子存活時(),效用函數為 ;孩子離世時(),效用函數為 ,其中 。
解題方法與完整推導
一、零利潤(公平保險)單位價格 與 之推導
保險公司精算公平(Actuarially Fair)條件下,單位保費等於給付機率:
- Case 1(針對生病投保):孩子生病的機率為 。保險公司期望給付為 ,保費收入為 。零利潤條件:
- Case 2(針對死亡投保):孩子死亡的機率為 。保險公司期望給付為 ,保費收入為 。零利潤條件:
二、Case 1(生病保險)的最佳保險購買量
在 Case 1 下,購買 單位保費支出為 。各狀態下父母的最終財富與效用為:
- 狀態 1(健康):機率 ,財富 ,孩子存活,效用 。
- 狀態 2(生病且存活):機率 ,財富 ,孩子存活,效用 。
- 狀態 3(生病且離世):機率 ,財富 ,孩子離世,效用 。
父母的期望效用函數:
化簡得:
求解極大化一階條件(FOC):
解得:
Table 1 shows the respective demand schedules and the combined demand schedule for enthusiasts and normies:
| price | enthusiasts' quantity demanded | normies' quantity demanded | total quantity demanded |
|---|---|---|---|
| 7 | 1 | 0 | 1 |
| 6 | 2 | 0 | 2 |
| 5 | 3 | 0 | 3 |
| 4 | 4 | 1 | 5 |
| 3 | 5 | 2 | 7 |
| 2 | 6 | 3 | 9 |
| 1 | 7 | 4 | 11 |
| 0 | 8 | 5 | 13 |
第 6 題10 分
There are two types of consumers for a monopolist's product: enthusiasts and normies. The product is indivisible and can only be sold in integer units. The population consists of equal number of enthusiasts and normies. Table 1 shows their respective demand schedules and the combined demand schedule. There is no fixed cost of production. The monopolist faces a constant marginal cost of $1.5.
(A) Suppose the monopolist cannot tell enthusiasts from normies. So the monopolist can only charge the same price to both types of consumers. The profit maximizing price is 2.5.
(C) The outcome where the monopolist produces 9 units, the enthusiasts consumes 6 units and the normies consume 3 units is Pareto efficient.
(D) Suppose the monopolist can distinguish enthusiasts from normies and can charge these two different types of consumers different prices. Total consumer surplus is strictly lower when the monopolist can price discriminate and charge different prices than when the monopolist can only charge a uniform price.
(E) Suppose the monopolist can identify enthusiasts and normies and charge these two different types of consumers different prices. Suppose the monopolist changed the product design that doesn't affect the normies' demand curve but shifts enthusiasts' demand curve upward vertically by 1.5. Under the new product design, the monopolist faces an increasing marginal cost curve
$$MC(Q) = \begin{cases} 0.3, & \text{if } Q \le 3 \\ 0.3 + 0.6(Q - 3), & \text{if } Q > 3 \end{cases}$$
where Q$ stands for total quantity produced. Since the new design does not affect the normies' demand curve, the profit maximizing price the monopolist charges the normies will remain the same.
登入後即可作答並保存紀錄。
核心觀念
本題為典型的**獨占市場(Monopoly)與價格歧視(Price Discrimination)**題型,測驗考生以下核心觀念:
- 單一價格獨占(Uniform Pricing Monopoly):獨占廠商在無法實施價格歧視時,透過比較總收益與總成本(或邊際收益與邊際成本)尋求利潤極大化的單一價格與產量。注意題目設定產品具不可分割性(Discrete/Integer Units)。
- 邊際收益(Marginal Revenue, MR):在離散單位下,增加一單位產品所帶來的總收益變動量,即 。
- 帕累托效率(Pareto Efficiency):市場總剩餘達到極大化,且邊際效益等於邊際成本()的配置狀態。
- 第三級價格歧視(Third-Degree Price Discrimination):獨占廠商能識別不同消費者族群,並對各族群單獨設定利潤極大化價格。
- 聯合生產成本與定價(Joint Cost Function):當廠商面臨非常數邊際成本(如邊際成本遞增)時,對某一市場的銷售量會透過總產量改變邊際成本,進而影響對另一市場的定價。
解題方法與基礎數據計算
根據題目給定之表格,需求函數係以離散數量呈現。廠商邊際成本為常數 ,無固定成本()。
1. 單一價格下(均一定價)之總收益與邊際收益表格
當廠商只能收取單一價格時,面對的是「總需求數量」(Total Quantity Demanded, ):
| 價格 () | 總需求量 () | 總收益 () | 邊際收益 () | 總成本 () | 利潤 () |
|---|---|---|---|---|---|
| 7 | 1 | 7 | 7 | 1.5 | 5.5 |
| 6 | 2 | 12 | 5 | 3.0 | 9.0 |
| 5 | 3 | 15 | 3 | 4.5 | 10.5 |
| 4 | 5 | 20 | - | 7.5 | 12.5 |
| 3 | 7 | 21 | - | 10.5 | 10.5 |
| 2 | 9 | 18 | - | 13.5 | 4.5 |
| 1 | 11 | 11 | - | 16.5 | -5.5 |
註:在 時,總需求量跳至 。
2. 第三級價格歧視下之個別市場分析(當 時)
若廠商能區分「熱狂者」(Enthusiasts, E)與「一般人」(Normies, N):
-
狂熱者市場(Enthusiasts, E):
需求點為 。
由此可知,對狂熱者定價 (或 時利潤 9,定價 5 時極大),最佳銷售量 ,極大利潤為 $10.5。
-
一般人市場(Normies, N):
需求點為 。
由此可知,對一般人定價 ,最佳銷售量 ,極大利潤為 $3.0。
選項分析
(A) 錯誤
- 理由:
在均一定價下,比較各價格對應的總利潤:- 當 時,,總收益 ,總成本 ,利潤 。
第 7 題10 分
The village of Danan has two coffee shops. Each produces coffee at zero marginal cost. There are two coffee shops in Danan: Ace Coffee and Better Coffee. The two coffee shops simultaneously choose their prices. Prices have to be non-negative. Let denote the price at Ace Coffee and denote the price at Better Coffee. Given price profile , the market share of Ace Coffee is
and the market share of Better Coffee is
(A) The strategies of Ace Coffee and Better Coffee are strategic complements: one shop's best response is increasing in the other shop's strategy.
(B) Because strategies are strategic complements, there are multiple Nash equilibria in this game.
(C) There is a Nash equilibrium in which Ace Coffee charges price of 4.
(E) Suppose Ace Coffee chooses its price first. After observing Ace Coffee's price, Better Coffee then chooses its price. Better Coffee's profits are lower in this situation than in the situation where both choose prices simultaneously without knowing the other firm's price.
登入後即可作答並保存紀錄。
核心觀念
本題考查異質產品/價格競爭之賽局理論(Price Competition with Differentiated Products / Hotelling-like Bertrand Model)以及靜態賽局(Simultaneous Game)與動態賽局(Sequential Game / Stackelberg Model)的求解與比較。
- 最適反應函數(Best Response Function)與策略互補(Strategic Complements):
- 廠商的利潤函數為 。
- 透過一階條件(First-Order Condition, FOC)可求出最適反應函數 。
- 若 ,代表對手的價格提高時,本廠商的最適價格亦隨之提高,稱為策略互補(Strategic Complements)。
- 納許均衡(Nash Equilibrium, NE):
- 同時選擇價格時,納許均衡點為兩廠商最適反應函數之交點 。
- 子賽局完美納許均衡(Subgame Perfect Nash Equilibrium, SPNE):
- 先動者(Ace Coffee)預測後動者(Better Coffee)的反應函數 ,將其代入自身的利潤函數中極大化,採用逆向歸納法(Backward Induction)求解。
解題方法與推導過程
1. 市場需求與利潤函數定義
兩家咖啡店邊際成本 。
已知當 時:
- Ace Coffee 的市場佔有率(需求量):
- Better Coffee 的市場佔有率(需求量):
兩廠商的利潤函數分別為:
2. 最適反應函數(Best Response Functions)推導
-
Ace Coffee 的極大化問題:
一階條件(FOC):
-
Better Coffee 的極大化問題:
一階條件(FOC):
由上述反應函數可知:
兩廠商的策略彼此為策略互補(Strategic Complements)。
3. 同時賽局之納許均衡(Simultaneous Move NE)
聯立解反應函數:
將 代入 :
驗證價格差:
符合內設解條件,且聯立方程有唯一解。
在此均衡下,兩廠商的利潤分別為:
第 8 題10 分
The government of Wonderland gives alcohol license to only one firm. Firm A and B are both interested in obtaining the license. Having the alcohol license for is worth 2 million dollars in total for firm B. The government official in charge of this decision is known to be corrupt and the firms can influence his decision by offering bribes. If firm A and B offer different amount of bribes, the firm offering a strictly higher amount of bribe will be granted the license. If firm A and B offer equal amount of bribe, then firm A gets the license with probability 1/3 whereas firm B gets the license with probability 2/3. A firm cannot take the bribe back even if it does not obtain the license. Each firm's payoff is equal to the value of the license times the probability of getting the license minus the bribe. For example, if firm A pays bribe of million dollars and obtains the license with probability 1/3, then firm A's payoff is equal to . Firm A and B approach the official in secret without knowing what the other firm has offered. For , let denote firm 's pure strategy that gives million in bribery. Assume that has to be a non-negative integer. That is, .
(A) is strictly dominated for firm B.
(B) It is a Nash equilibrium for firm A to offer 1 million dollars, and offering x_B = 0, 1, 21 million dollar (so each with probability 1/2).
登入後即可作答並保存紀錄。
核心觀念
本題考查:
- 納許均衡:每位玩家的策略都是對對方策略的最佳反應。
- 嚴格劣勢策略:若某策略對所有對手策略的報酬,都嚴格低於另一策略,該策略即被嚴格支配。
- 混合策略納許均衡:玩家在混合策略中給予正機率的純策略,必須具有相同且最高的期望報酬。
- 競標賄賂的得標規則與沉沒成本:即使未取得執照,賄賂仍須支付。
以下金額均以「百萬美元」為單位。
若一家公司得標,其報酬為執照價值減賄賂;未得標時,報酬為賄賂的負值。平手時:
- 公司 A 的得標機率為 ,故報酬為 。
- 公司 B 的得標機率為 ,故報酬為 。
解題方法
先剔除明顯劣勢的高額賄賂,再列出低額賄賂的最佳反應,最後檢查純策略與混合策略均衡。
公司 A 的主要報酬如下:
公司 B 的主要報酬如下:
選項 A:正確
考慮公司 B 的任何賄賂 。比較公司 B 出價 :
- 若 ,出價 時平手,報酬為
而出價 時,即使得標,報酬至多為
- 若 ,出價 時公司 B 落標,報酬為 。
- 出價 時:
- 若落標,報酬為 ;
- 若平手,報酬為 ;
- 若得標,報酬為 。
因此,無論公司 A 出價多少,公司 B 出價 的報酬都嚴格高於任何 。
所以 對公司 B 是嚴格劣勢策略。選項 A 正確。
選項 B:錯誤
若公司 A 出價 、公司 B 出價 :
- 公司 A 得標,報酬為
- 公司 B 落標,報酬為 。
但在公司 A 出價 1,可以平手,報酬為
因此,公司 B 出價 並非對公司 A 出價 $1 的最佳反應,這不是納許均衡。
選項 B 錯誤。
選項 C:錯誤
高額賄賂不會出現在均衡中,因此只需考慮:
- 公司 A:
- 公司 B:
各方的最佳反應如下。
| 對手出價 | 公司 A 的最佳反應 |
|---|---|
| 或 |
公司 B 的最佳反應為:
| 對手出價 | 公司 B 的最佳反應 |
|---|---|
逐一比較可知:
- :公司 B 想改為 。
- :公司 B 想改為 。
- :公司 B 想改為 。
- :公司 B 想改為 $0
第 9 題10 分
Ben has a brilliant business idea and wants to start a company based on that idea. To make a prototype, he needs an initial investment of . However, he has no capital of his own and thus goes to Venture Capitalist (VC) for investment. If VC does not invest, then both VC and Ben get 0. If VC invests , then Ben will choose whether to work or to shirk. If Ben works hard, he will turn his idea into reality and the value of his startup will become . If Ben shirks and diverts the investment, he can pocket the entirety of the investment and the value of his startup will become 0. Suppose VC gets a share of the company if he invests in the company. Therefore, if VC invests and Ben works, then VC's payoff is and Ben's payoff is . If VC invests and Ben shirks, then VC's payoff is and Ben's payoff is .
(A) In the unique backward induction solution, VC will invest and Ben will work hard, because Ben gets by working hard after VC invests whereas Ben gets 0 if VC does not invest.
(B) There is no backward induction solutions in which VC invests.
(C) Suppose VC demands only of the company from his investment so that VC just breaks even with the investment, then VC invests in the unique backward induction solution.
(D) Suppose VC demands 80% of the company from his investment. Suppose regulation is much more stringent and Ben can divert only 50% of the investment if Ben shirks. That is, if Ben shirks, then VC's payoff is whereas Ben's payoff is . Then VC invests in the unique backward induction solution.
(E) Suppose VC demands 80% of the company from his investment and he can get the entirety of the investment if he shirks. Suppose it is commonly known that Ben will have a second business idea and want to do a second startup. Suppose the relevant parameters for his second startup is exactly the same as described up. So, the exact same situation will repeat twice. Then, future concern will discipline Ben. There is a SPNE in this twice-repeated game in which VC invests in Ben's first startup.
登入後即可作答並保存紀錄。
核心觀念
本題屬於賽局理論(Game Theory)中的動態賽局(Dynamic Games)與重複賽局(Repeated Games)應用,探討創投(VC)與創業家(Ben)之間的委託代理問題(Principal-Agent Problem)與道德風險(Moral Hazard)。
涉及的核心定理與概念如下:
- 逆向歸納法(Backward Induction):用於求解有限期完美資訊動態賽局的子賽局精明納什均衡(Subgame Perfect Nash Equilibrium, SPNE)。分析時須從賽局的最末階段(後動者)開始推導其最佳決策,再倒推回前動者的最佳決策。
- 道德風險與激勵相容條件(Incentive Compatibility Constraint, IC):當創業家偷懶(shirk)能獲得個人私利(Private Benefit / Diversion)時,其努力工作的報酬必須大於偷懶報酬,否則創業家必定選擇偷懶。
- 逆向歸納法的有限次重複賽局連鎖剝離定理(Chain-Store Paradox / Selten's Theorem):在有限次重複賽局中,若階段賽局(Stage Game)存在唯一的 SPNE,則由逆向歸納法可知,在最後一期 participants 必定選擇階段賽局的均衡策略;倒推回第一期,每一期都會重複階段賽局的均衡結果,無法利用「未來聲譽/關切」來懲罰或激勵當期行為。
解題方法
本題為兩階段動態賽局(動態樹狀賽局):
- 第一階段:創投 VC 決定是否投資 (Invest or Not Invest)。若不投資,兩位報酬均為 。
- 第二階段:若 VC 投資,創業家 Ben 決定要「努力(Work)」還是「偷懶/挪用資金(Shirk)」。
1. 原始參數下的階段賽局推導
- 若 Ben 努力工作(Work): startup 價值 。VC 獲得 ;Ben 獲得 。
- 若 Ben 偷懶挪用(Shirk):挪用資金 ,startup 價值變為 。VC 獲得 ;Ben 獲得私利 。
逆向歸納法分析:
- 第二階段(Ben 的決策):
Ben 比較努力與偷懶的報酬:
因此,一旦 VC 選擇投資,Ben 的最佳回應永遠是偷懶(Shirk)。 - 第一階段(VC 的決策):
VC 預期到若自己投資,Ben 必定偷懶,導致自己的報酬為 。
比較投資與不投資:
因此,VC 的最佳回應是不投資(Not Invest)。
結論:唯一的子賽局精明納什均衡(SPNE)為:VC 選擇「不投資」,Ben 選擇「若獲投資則偷懶」。
選項分析
(A) 錯誤
- 原因:選項宣稱 Ben 在 VC 投資後會努力工作,比較對象誤用「偷懶時報酬(應為 )」與「不投資時報酬()」。Ben 在獲投資後的行為選擇取決於 與 的比較。由於 ,Ben 必定偷懶。VC 預測此結果後不會投資。
第 10 題10 分
Quality of used cars in the town of Nowhere is distributed according to the probability density function where for . So, for , the expected value of conditional on is equal to . In particular, the expected value of conditional on is equal to . If the owner of a quality used car keeps the car, then his payoff is equal to . Because the used car would be new to people other than the original owner and add spice to life, the payoff of a quality used car to a potential buyer is . So, if the owner of a used car with quality sells the car at price , then his payoff is while the payoff to the buyer is .
Suppose the owner of the used car knows the quality of his car but the buyer does not. In equilibrium, what is the maximum possible quality of a car that is offered for sale in the market? If owners of all used car qualities offer their car for sale in some equilibrium, then . If owners of all used car qualities keep their car in every equilibrium, then . If owners of cars of some quality levels offer their cars for sale while others don't in some equilibrium, then is such that owners of used cars with quality will keep their cars in every equilibrium while owners of used cars with quality will sell their car in some equilibrium.
Just write down your answer of .
登入後即可作答並保存紀錄。
核心觀念
本題考查**逆向選擇(Adverse Selection)與檸檬市場(Lemons Market / Market for Lemons)**的模型推導。
在二手車市場中存在資訊不對稱(Asymmetric Information):賣方完全知道自己二手車的品質 ,而買方只知道品質的先驗機率分布 。
-
賣方供給決策(Reservation Price / Participation Constraint):
品質為 的車主若留著自用,效用為 ;若以價格 售出,效用為 。因此賣方願以價格 賣車的保留條件為:
意即在市場價格為 時,只有品質 (若上限超過 則為 )的車主願意賣車。 -
買方願付價格(Competitive Market / Zero Profit Condition):
假設買方市場為完全競爭,買方的期望支付價格等於市場上所有出售車輛的品質在買方心中的期望效用(保留效用):
-
市場均衡(Market Equilibrium):
在均衡時,市場價格 與市場上最高二手車品質 必須滿足自我一致性(Self-Consistency)。
解題方法
步驟一:分析賣方的保留供給區間
假設在某一價格 下,市場上出售車輛的最高品質為 。
由賣方效用函數可知,品質為 的臨界車主(Marginal Seller)售出與自用的效用相等:
因此,若最高品質為 ,市場對應的價格必定為 。品質 的車主都會選擇賣車。
步驟二:計算買方對該品質區間的期望效用
給定車主只會在 時賣車,買方得知出售車輛品質落在 。
題目給定條件:當 時, 在 上的條件期望值為:
因此,當 時,買方對市場上出售車輛品質的條件期望值為:
買方願意支付的價格等於其條件期望效用:
步驟三:結合買賣雙方條件,求解市場最高品質
將買方願付價格與賣方供給價格等量比較:
- 賣方要求價格:
- 買方願付價格:
由於對任意 ,買方的願付價格 恆大於賣方臨界車主的保留價格 :
這意味著對於市場上任何品質上限 ,買方都願意付出高於賣方臨界保留價的金額。