1、ce fixes the problem that outliers pose to the expected shortfall by squaring the distance from each observation below the benchmark return. Hence, observations that are farther below the benchmark return are penalized more by this system. Using the semi-variance in lieu of the standard deviation re
2、sults in more appealing asset allocation decisions since only downside risk is considered 3. DISCUSS tracking error and how its computation is related to other measures of risk ( and ) Tracking Error is defined as the standard deviation of the difference in return between the investment and a specif
3、ied benchmark or target position. It measures how closely the investment performs relative to the benchmark. However, tacking error relies on a normal distribution. As a result, tracking error does not distinguish between deviations above and below the benchmark. Assuming that the deviations on the
4、downside are more serious, tracking error does not give a complete picture of the risk involved in a particular investment. The variance of the total return is a special case of the tracking error when the benchmark return and the expected return of the investment are equal. The formula for the trac
5、king error for stocks looks much like the measurement of total risk. If the is close to 1, the tracking error is primarily a of Unsystematic Risk. Similarly, if the duration of the bond is similar to that of the market index and the yield beta is equal to 1, then the tracking error will equal the du
6、ration-adjusted residual yield volatility risk management CFA Level III Page 3 of 6 Gillsie VAR for the Asset Manager flashcard concepts The Variance/Covariance VAR method uses a process where complex portfolio risks are MAPPED to individual risk factors and then the portfolio is computed using the
7、individual standard deviations of these risk factors and their relative portfolio weights. Once the standard deviation is computed, VAR can be computed by finding the Maximum dollar (or return) loss for a given probability level and time period The Strengths of Variance/Covariance VAR are that: ? It
8、 is Relatively Straight-forward ? Market Data Used to Compute VAR is Readily Available ? the RiskMetrics system can even model some non-linear positions Weaknesses of Variance/Covariance VAR: ? Assume Stable Variance/Covariance matrix through time (may not be true) ? Assume Normally Distributed Retu
9、rns (may not be true) ? Variance/Covariance Matrix becomes larger and more complex as the number of risk factors grows With HISTORICAL VAR, you simply line up returns from low to high and pick the return level that corresponds to the probability level you are interested in. Strengths of Historical V
10、AR are: ? Easy to Use and Understand ? Makes no assumptions re: return normality or linearity ? Variance/Covariance matrices are not necessary Weaknesses of Historical VAR are: ? Large Data Requirements ? Assuming that the future will be like the past ? Valuation models may be necessary for assets w
11、ithout frequent return statistics MONTE CARLO Simulation utilizes asset pricing models to generate many future portfolio price paths. The distribution of asset returns is then generated from the results of this path determination process. Once the distribution of asset returns is determined, VAR is
12、easily derived by looking at the appropriate probability level. The strengths of Monte Carlo VAR are that ? it is more flexible relative to parametric or historical VAR ? no assumptions re: linearity or normality of returns are required Weaknesses of Monte Carlo VAR include ? Portfolio complexity an
13、d the number of scenarios needed ? the path generation process requires asset pricing models which may or may not be reflective or reality STRESS TESTING is “what if” analysis applied to a particular VAR computation. Stress testing allows you to determine the validity and reliability of a VAR estima
14、te The practical applications of VAR for portfolio managers are ? Comparable risk evaluation across asset classes ? Identification of Key Portfolio risks ? Portfolio construction ? Setting portfolio risk limits risk management CFA Level III Page 4 of 6 Gillsie Absolute performance evaluation using V
15、AR should be performed with the problems of VAR in mind Peer group risk evaluation using VAR is not a good idea due to the fact that VAR differs significantly depending on the methodology and assumptions employed Relative risk evaluation v. a benchmark portfolio is more effective than relative peer
16、evaluation because the same methodologies and assumptions can be applied to both the portfolio under evaluation and the benchmark. Hence, comparability problems are minimized. Problem Set: VAR for the Asset Manager by Stocks & Ito 1. DISCUSS the strengths and weaknesses of using variance/covariance
17、VAR computation Strengths: - Using RiskMetrics, the computation of VAR is relatively easy - In most cases, the market data necessary to compute VAR is available - The RiskMetrics system can even model the risk of some non-linear positions Weaknesses: - System assumes that the variance/covariance mat
18、rix is Stable through time. We know now from the Clarke article that the and Correlations may not be stable through time - Some assets may have non-normal return distributions. Variance/Covariance VAR has limited success in handling non-normal distributions - The VAR computation becomes more difficu
19、lt as the complexity of the portfolio increases. As portfolio complexity increases, the number of required risk factors grows, and the variance/covariance matrix rises proportionately 2. CONTRAST the use of VAR for relative risk evaluation v. a competitors portfolio relative to a similar evaluation
20、relative to a benchmark portfolio. Relative evaluation v. a Market Index is more Viable than Peer Group Evaluation because a particular individual will be making the VAR computation for both the underlying portfolio and the market index. Hence, you dont have the large comparability problems that are
21、 inherent in peer comparisons. Also, the characteristics of the market index are better known than the characteristics of the competitors portfolio. 3. Briefly OUTLINE the uses of VAR for portfolio managers Uses of VAR for portfolio managers include: - Comparability - VAR measures risk comparably ac
22、ross asset classes. The result is that with VAR, the risk of a bond portfolio can be compared against the risk of an equity portfolio - Risk Identification - Recall that the key to the parametric VAR computation is the dis- aggregation of complex risks into isolated risk factors. A portfolio manager
23、 can use this process to identify the macroeconomic factors that have the mot impact on the overall portfolio manager. - Portfolio Construction - The parametric method relies on the determination of the variance/covariance matrix of these individual risk factors. The portfolio manager can use this v
24、ariance/covariance matrix to develop optimal portfolio allocations - similar to standard modern portfolio theory risk management CFA Level III Page 5 of 6 Gillsie - Risk Limits - Portfolio manager can use VAR to set risk limits for a portfolio - either on an absolute level or on a relative basis v.
25、a benchmark. Global Risk Management: Are We Missing the Point flashcard concepts The point of this article is to show that the main problem facing a risk manager is to uncover and analyze risks that are not readily apparent in the current market environment. e.g., if you use VAR to measure risk, you
26、 should look at what VAR is NOT telling you. You should look at what can happen the other 5% of the time A Hedge that appears quantifiably riskless based on a pricing model or other hedging framework may still contain some risk. Risks that may remain include (a) Model Risk, and (b) Inventory Risks -
27、 such as operational, legal, and liquidity risks. The implication of fat-tailed distributions is that there are actual market events that cannot be readily explained or predicted by using a standard normal distribution to measure risk. The risk manager should carefully evaluate the tails of an asset
28、s distribution. You are attempting to guard against the portfolio damage that would be caused by severe market moves. Asset correlations tend to rise during market events. As correlations increase, the benefits of global diversification falls. Also, VAR risk estimates become unreliable due to a shif
29、t in the variance/covariance matrix. As correlations rise, you can lose more for a given probability level and time due to the loss of diversification benefits. Hence, diversification is reduced at the time when you need it most. Non-linear risks are called Negative Risks and are represented by loss
30、es that rise at an increasing rate as a particular risk factor moves against you Most catastrophic risks in global asset management are of an organizational nature, rather than of risk management structure An effective organizations structure is contingent upon appropriately structured incentives, a
31、ccess to information with respect to strategy and revenue, open interaction within the organization, and the knowledge to seize issues directly Problem Set: Global Risk Mangement: Are we missing the point by bookstaber 1. INTRODUCE Two reasons why a hedged position may still contain some risk There
32、are TWO main factors that can have an impact on the risk of a Hedge: - Model Risk - Suppose a hedge was created relative to a model like VAR. This hedge is only as good as the model that was used to construct it. If there are problems with the model or the assumptions underlying the model, the hedge
33、 will suffer - Inventory Risk - Most trading desks will have net long or short inventory positions that will need to be hedged on a daily basis. What is meant by Inventory Risk are all of those non-quantifiable risk factors that are not considered in VAR or other risk models. Examples include legal
34、risks, operational risks, liquidity risks, etc. 2. COMMENT on the normality of return distributions and the implication of that normality for asset managers Many academic studies have shown that asset returns are NOT normally distributed. Many asset distributions exhibit fat tails. The key here is a
35、 discussion of OUTLIERS. As risk managers we are interested in what may consider the other 5% of the time in a VAR computation. Bookstaber is telling us that we should throw out the 95% of the distribution that everyone else is looking at risk management CFA Level III Page 6 of 6 Gillsie and concent
36、rate on the improbable, yet possible, lower 5% of the distribution. If the西方经济学（本）行考练习题2分类整理答案一、单项选择题第二章 供求理论1需求曲线是一条倾斜的曲线，其倾斜的方向为 （A右下方） 2下列体现了需求规律的是 （D照相机价格下降，导致销售量增加）3其他因素保持不变，只是某种商品的价格下降，将产生什么样的结果（C需求量增加）4下列变化中，哪种变化不会导致需求曲线的位移 (B产品的价格) 5当汽油的价格上升时，在其他条件不变的情况下，对小汽车的需求量将 （A减少）6当咖啡价格急剧升高时，在其他条件不变的情况
37、下，对茶叶的需求量将（C增加）7消费者预期某种物品将来价格要上升，则对该物品当前的需求会 （C增加）8需求的变动与需求量的变动（B需求的变动由价格以外的其他因素的变动所引起的,而需求量的变动由价格的变动引起的）9整个需求曲线向右上方移动，表明 （A需求增加）10对化妆品的需求减少是指（A收入减少引起的减少）11在同一条曲线上，价格与需求量的组合从A点移动到B点是（D需求量的变动）12导致需求曲线发生位移的原因是（C因影响需求量的非价格因素发生变动，而引起需求关系发生了变动） 13下列因素中哪一种因素不会使需求曲线移动（B商品价格下降）14供给曲线是一条倾斜的曲线，其倾斜的方向为（B右上方）15
38、鸡蛋的供给量增加是指供给量由于（C鸡蛋的价格提高而引起的增加）16如果某种商品供给曲线的斜率为正，保持其他条件不变的情况下，该商品价格上升，导致（C供给量增加）17建筑工人工资提高将使（A新房子供给曲线左移并使房子价格上升）18假如生产某种商品所需原材料的价格上升，则这种商品（B供给曲线向左方移动）19供给规律中可以反映在（D某商品价格上升将导致对该商品的供给量增加）20当供求原理发生作用时，粮食减产在市场上的作用是（B粮食价格上升） 21关于均衡价格的正确说法是（C供给曲线与需求曲线交点上的价格）22均衡价格随着（D需求的增加和供给的减少而上升）23当某种商品的需求和供给出现同时减少的情况时
39、，那么（C均衡价格无法确定，均衡产量减少） 24已知某商品的需求函数和供给函数分别为：QD=14-3P，QS=2+6P,该商品的均衡价格是（A4/3）25假设某商品需求曲线为Q=3-2P,市场上该商品的均衡价格为4,那么,当需求曲线变为Q=5-2P后，均衡价格将（A大于4）26需求的变动引起（A均衡价格和均衡数量同方向变动）27供给的变动引起（B均衡价格反方向变动，均衡数量同方向变动）28政府为了扶植农产品，规定了高于均衡价格的支持价格。为此政府应采取的措施是（C收购过剩的农产品） 29政府把价格限制在均衡价格以下，可能导致（A黑市交易）30假定A为自变量，B为因变量，弹性的数学表达式为（A
40、E=（B/A）（A/B） 31在下列价格弹性的表达中，正确的是（A需求量变动对价格变动的反应程度）32需求完全无弹性可以用（C一条与纵轴平行的线表示）33哪一种商品的价格弹性最大（D化妆品）34若某商品价格上升2%，其需求量下降10%，则该商品的需求的价格弹性是（B富有弹性）35.如果某商品是富有需求的价格弹性，则该商品价格上升（C销售收益下降） 36.一般来说，某种商品的需求价格弹性与购买该种商品的支出占全部收入的比例关系是（A购买该种商品的支出占全部收入的比例越大，其需求价格弹性就越大）37如果价格下降20%能使买者总支出增加2%，则该商品的需求量对价格（A富有弹性）38如果人们收入水平提
41、高，则食物支出在总支出中的比重将会（C下降）39假定某商品的价格从5元降到4元，需求量从9个上升到11个，则该商品的总收益将（C减少）40两种商品中如果其中的一种商品价格发生下降或上升时，这两种商品的购买量同时增加或同时减少，请问二者的交叉弹性系数是（A负）41某类电影现行平均票价为4元，对该类电影需求的价格弹性为1.5，经常出现许多观众买不到票的现象，这些观众大约占可买到票的观众的15%，要想使所有想看电影而又能买得起票的观众都买得到票，可以采取的办法是（D电影票提价10%）42什么情况下应采取薄利多销政策（B价格弹性大于1时） 43所有产品的收入弹性（D不一定）44下列产品中，哪种产品的交
42、叉弹性为负值（A汽车和轮胎） 45对于一个不吃猪肉的人来说，猪肉的需求量和牛肉价格之间的交叉弹性是（A0）46如果Y的价格上升，对替代品X的影响是（A需求量增加） 47已知需求方程为：Q=50-2P，在P=10处的点价格弹性是（B0.67）48假如Q=200+0.1M，M=2000元，其点收入弹性为（D0.5）49如果说两种商品A和B的交叉弹性是-3，则（DA和B是互补品）50劣质品需求的收入弹性为（B负） 51蛛网理论主要是针对（B周期性商品）52已知某商品的需求弹性与供给弹性均等于1.5，则蛛网的形状是（C封闭型）53已知某商品的需求弹性等于0.5，供给弹性等于1.8，则蛛网的形状是（B发
43、散型）54已知某商品的需求弹性等于1.5，供给弹性等于0.8，则蛛网的形状是（A收敛型）55一般来说，香烟的蛛网的形状是（A收敛型）56一般认为，农产品的蛛网的形状是（B发散型）57.目前我国私家车的蛛网形状是（A收敛型）第三章 效用理论1某消费者偏好A商品甚于B商品，原因是（D对其而言，商品A的效用最大）2下列对物品边际效用的不正确理解是（C消费一定数量的某种物品所获得总效用）3关于基数效用论，不正确的是（C基数效用论和序数效用论使用的分析工具完全相同）4某消费者逐渐增加某种商品的消费量，直到达到了效用最大化，在这个过程中，该商品的（C总效用不断增加，边际效用不断下降）5总效用曲线达到顶点时
44、（B边际效用为零）6边际效用随着消费量的增加而（A递减）序数效用论认为，商品效用的大小（D可以比较）8如果消费者消费15个面包获得的总效用是100个效用单位，消费16个面包获得的总效用是106个效用单位，则第16个面包的边际效用是（D6个）9已知商品X的价格为8元，Y的价格为3元，若某消费者买了5个单位X和3个单位Y，此时X、Y的边际效用分别为20、14，那么为获得效用最大化，该消费者应该（C增加Y的购买，减少X的购买）10消费者剩余是指消费者购买某种商品时，从消费中得到的（B满足程度超过他实际支付的价格部分） 11消费者剩余是消费者的（B主观感受）12实现了消费者均衡的是（C.MUA/PA
45、= MUB/PB）13已知某家庭的总效用方程为TU=14Q-Q2，Q为消费商品数量，该家庭获得最大效用时的商品数量为（B7）14已知商品的价格为元，商品的价格为元，如果消费者在获得最大满足时，商品的边际效用是元，那么，商品的边际效用是（D60）15.一般来说，无差异曲线的形状是（B向右下方倾斜的曲线）16无差异曲线为斜率不变的直线时，表示相结合的两种商品是（B完全替代的） 17在同一个平面图上有（B无数条无差异曲线） 18在一条无差异曲线上（D曲线上任两点对应的消费品组合所能带来的总效用相等）XY19无差异曲线上任一点上商品X和Y的边际替代率等于它们的（C边际效用之比） 20如图所示的无差异曲线表明X和Y的关系为（A完全互补品）21同一条无差异曲线上的不同点表示（B效用水平相同，但所消费的两种商品的组合比例不同）22预算线上的表明（C预算支出条件下的最大可能数量组合）23预算线的位置和斜率取决于（C消费者的收入和商品的价格）24商品X和Y的价格按相同的比率上升，而收入不变，预算线（A向左下方平行移动B向右上方平行移动）25M=PxX
个人主页
