Answered You can buy a ready-made answer or pick a professional tutor to order an original one.

QUESTION

Decision Tree and Sensitivity analytics question

Machine Works is a firm supplying quality  metal gears to manufacturers  of  high-end  clocks.

Machine Works is faced with the problem of extending credit to a new customer, Qualclocks, in connection with a contract for next year's season. Machine Works has three risk categories for credit - "poor'', "average" and "good" according to repayment behavior. Its experience is that 30% of companies similar to Qualclocks are "poor'', 50% "average", and 20% ''good". If credit is extended, the expected net contribution from the work is a loss of 400,000 for "poor" credit behavior, a profit of 200,000 for "average", and a profit of 300,000 for "good". If Machine Works does not extend credit, Qualclocks will go elsewhere. Address the following questions using decision trees and the expected value criterion:

(a)  Based only on the information presented so far, advise Machine Works as to whether or not to offer credit to Qualclocks.

(b)    What is the most that would be worth paying a rating agency to provide information regarding the credit risk category of the fashion clock manufacturer?

(c)   Machine Works is able to consult a credit rating agency, who, for a fee of £20,000 will rate the manufacturer as A, B or C. The reliability of this credit rating agency is reported in the table below. Each row gives the probability of each credit rating given the actual risk state for the company. For example, if the company truly has a "poor" credit risk, the probability of the credit rating agency giving an A rating is 10%. Advise Machine Works as to what it should do.

                Credit Rating

Actual      C        B        A

Poor      .5         .4         .1

Average  .4       .5         .1

Good      .2       .4          .4

(d)  How does the optimal course of action in part (c) change with respect to the fee charged by the credit rating agency?

(e)   Discuss the use of criteria other than expected value for evaluating decision trees.

Show more
  • @
  • 10 orders completed
ANSWER

Tutor has posted answer for $100.00. See answer's preview

$100.00

*************** =(03 * ******** + (05 X200000) ******** **** * ******* + ****** +600000 ********* **** ******** ** TO ****** ****** ******** ** QUAL ********* *** TABLEPOORAVERAGEGOODEXTEND ******************************** ****** ************** ******************* ****** ******* RELIABILITYPOORAVERAGEGOODC050402B040504A010104Prior ************* ** *** ****** of ***************** *** Bayes theorem ** can ****** *** ***** ************* ** ****** the ********* ************* ********** to *** states of ****** ** ********* *** the ****** company *********** ************* ** ****** the ***** ************* for **** ******** / *** * ******** ***** = ******** p(sj) ********** s are the ****** of ****** ** ******* *** the ******** outcomes ** *** ****** information ******** **** * possible ******* ABC for *** ************** probabilities *** *** states ** ***************** of qualclock being * ** **** ****** behavior ***** **** *** rating company ***** * ****** rating ********** = *** **** * 015/ *** * **** *** x05) +(04x05) +(02 x02)Probability ** ********* ***** * ** average ****** ******** ***** **** *** rating ******* ***** * credit ****** cP(average/c) * ******* = *** *** * ***** *** **** ******** **** *************** ** ********* being * ** average credit behavior given that *** ****** ******* gives * credit ****** ********** * (02 x02) * **** 039 * 4/39 *** **** +(04x05) **** *************** of ********* ***** * ** Poor ****** behavior ***** that the ****** ******* gives * ****** rating ********** * (04x03) * ******* =4/15(04x03) +(05 **** **** *************** of qualclock ***** * of average ****** behavior ***** **** the ****** ******* ***** * credit ****** ********** /B) * *** x05) * ******* *********** **** **** **** *************** of ********* ***** * of ******* ****** ******** ***** that the rating company ***** * ****** ****** ********** * *** **** * ******* ************ +(05 **** +(02 *************** of qualclock ***** * ** **** credit behavior ***** that *** ****** ******* ***** * ****** ****** ********** * (03 x01) = ******* =3/16(03x01) **** x01) **** x04)Probability of qualclock ***** * ** poor ****** ******** ***** that *** ****** company ***** * credit rating ************* = *** x01) = ******* ************ +(05 x01) **** *************** of ********* ***** * of **** credit ******** ***** **** the ****** ******* ***** * credit ****** ********** * *** **** * 008/016 *********** +(05 **** **** ******* ***** *** *** ********* ************* ** ******* ***** ****** ** **** * credit ****** ******* (they *** *********** ****************** can ** **** ****** in a ***** as ****** ***** of ** *** ************************************************************ ** ********* ******** PAYOFFS *** ********* ******** ************ **** the ****** company ***** rating *********** ********* * **** ********* * ***** ******** + **** (300000) * -2051282 -20000 = -405122 EMV ( ******* Extend credit/C) * **** *** * 20/39 *** * **** (0) ** ****** * *************** ** *** ****** ******* ***** rating * ****** ****** *** ** *** **** ****** with EMV * ******* **** *** ****** ******* ***** ****** BEMV(Extend ********* = 4/15 (-400000) * *** (200000) * 8/45 ******** = ******* -20000= *********** ******* Extend ********* * **** (0) * 5/9 *** * **** *** ** ******* *************** ** *** rating company ***** ****** * extend ****** *** ** *** **** ***** * **** *** ************* *** rating ******* ***** rating *********** ********* = **** ********* * **** (200000) * 1/2 ******** * ******** -20000 * 149230EMV( DON’T Extend ********* * **** (0) * **** (0) * 1/2 *** ***************** If *** ****** ******* gives ****** * ****** credit *** ** *** **** option **** *** **************** ***** ** EXPERIMENTATION =( EXPECTED VALUE ** *** **** * ********* ***** ** THE ************ **** ****** INFORMATION) ******* SAMPLE INFORMATION) *** * **** – EMV (BEST) *** ** *** *** * = *** **** * *** **** * ** X039) ********* ***** ******** x 016 * * ****** ********* ************* +(200000X05)+(300000X02) ********* * ****** *** ***** * ******** ***** ***** BY ****** A ****** ******* ** ******* ***** ****** *** ****** company **** ***** ******* ***** ***** **** ** *** a value of ********** ***** ** a loss ******* ***** ****** not **** this ****** ****** ********* ******* ***** ****** go ****** ******** ****** * ****** ******** *** **** ******** ***** on *** ******** ******** where *** best decision is ** ****** ****** ******** ** ********* *** ******** has an *** ** ***** ** If *** *** ******* by * rating ******* ** < ***** *** ******* ******** ** to **** * ****** ********* *** *** charged ** * ****** ******* is ********* the ******* ******** ** to ** direct *** **** decision ** whether ** go ****** *** **** decision ** ** ******* ** ****** ****** or not ** extend ***** ** *** ******** **************** ***** ********* **** ** ******** decision trees is *** Expected *********** ********* which takes **** ** the value ** opportunity **** ** choosing *** ******* ******************** we first ********* *** ****** ****** and ***** **** at *** ******* *** ** *** ******** **** *********** = *** ****** ***** where * **** MaxVij – ***

Click here to download attached files: Decision theory.docx
or Buy custom answer
LEARN MORE EFFECTIVELY AND GET BETTER GRADES!
Ask a Question