Case Method:

Cases provide a context for application of analytic concepts, and illustrate the issues that arise in the complex decision-making situations that typically face top-management. As in real business problems, there are no “right’’ answer for case studies that we will examine in the course, although there are correct and incorrect ways to analyze or approach them. The challenge for the student is to (1) examine the facts and the data contained in the case, (2) employ the analytical frameworks learned in earlier classes (and concurrently), (3) reach conclusions, and (4) make specific recommendations that will resolve the issues presented by the case.

To prepare for a case discussion, you should read each case and analyze the data that it presents. Texts and readings from earlier courses should be used to the extent that they assist in your preparation. A thorough preparation for discussions includes systematically (1) outlining the major issues presented by the case, (2) identifying the analytical techniques or frameworks appropriate for resolving the problem, and (very important) (3) outlining steps to implement a specific course of action that is supported by the analysis. I would strongly recommend you to review the “Note to the Student: How to Study and Discuss Cases”.

Questions to think about while reading cases:

• What are the basic facts? What are the characteristics of the company and the market?
• Who are the key players? What are their objectives?
• Is there an organization in distress? Is there an undeveloped market opportunity?
• If so, what are the symptoms? What are the measures or evidence? Are they biased?
• Are there underlying problems or trends? What are they? How do we know?
• Is there one transcendent problem or opportunity? What is it? How do we know?
• What decisions need to be made? What are the alternatives for action?
• What are the pros and cons of each alternative? How do we evaluate them?
• Which alternative do you recommend? Why?
• What should we learn from this case?
• How does this case relate to the course topic? To other cases? To the reading?

Suggested Format for Case Reports:

Case reports are individual assignments. The purpose of the case report is to synthesize all the knowledge you gain in the class relevant to the case and channel it to solve an operational problem. Your case report should be less than 5 pages using 1.5 line spacing and include the following sections: Executive Summary (not more than ½ page), Background and Issues, Situation Analysis, Evaluation of Potential Solutions, and Recommendations.

The Executive Summary is a summary of the report that explains the problem and the proposed recommendations. In the Background and Issues section you describe the situation under study (do not rewrite the case) and identify the key issues addressed by the report. In the Analysis section provide the details of your analysis. You should first start with listing the assumptions made, if any. Then explain the approach taken to analyze the situation, and how you have arrived to the recommendations/findings listed in the following section (Detailed reasoning and analysis in support of your recommendations/findings should be given in an Appendix). If appropriate, you can also suggest further issues to be examined or further studies to be done. In the Evaluation section you will propose potential solutions and evaluate each of them. Finally, in the Recommendations section you should propose a set of specific actions along with the key reasons in support of your recommendations. This section will be the conclusion to your report. You may use bullets when appropriate.

The grades on the reports will be based on the logical consistency, precision and analytic structure of the paper. Specifically you should think about the extent to which the report

• Grounds the analysis on the analytical concepts discussed in class;
• Explicitly states the assumptions in the analysis;
• Isolates the fundamental problems for the situation, and remains focused on these;
• States criteria for choosing among alternative action plans;
• Integrates the action plans with the analysis;
• Ensures that the action plans are situation-contingent;
• Is persuasive that the action plans are reasonable, effective and efficient.

  Student+Simple ES

  HBP Product No.: ST5WS
  UST005/WSS/1207
  Simple Exponential Smoothing Model
  MM-YY	Period	Sales	Forecast	Error	Absolute	Alpha =	0.4
  t	St	Ft	St – Ft	Error
  Jul-09	1	3,924	Formula
  Aug-09	2	2,619	3,924	-1,304	1,304	F2 =	S1
  Sep-09	3	6,920	3,402	Ft+1 =	Alpha*St + (1 – Alpha)*Ft
  Oct-09	4	5,676	Percentage Error =	(St – Ft) / St * 100%
  Nov-09	5	8,348
  Dec-09	6	6,044
  Jan-10	7	6,877
  Feb-10	8	6,535
  Mar-10	9	6,395
  Apr-10	10	6,684
  May-10	11	5,414
  Jun-10	12	3,180
  Jul-10	13	4,350
  Aug-10	14	3,175
  Sep-10	15	6,935
  Oct-10	16	6,356
  Nov-10	17	8,919
  Dec-10	18	7,146
  Jan-11	19	7,763
  Feb-11	20	7,397
  Mar-11	21	7,286
  Apr-11	22	7,498
  May-11	23	6,386
  Jun-11	24	4,209
  Jul-11	25	4,825
  Aug-11	26	3,764
  Sep-11	27	7,066
  Oct-11	28	7,015
  Nov-11	29	9,535
  Dec-11	30	8,278
  Jan-12	31	8,773
  Feb-12	32	8,393
  Mar-12	33	8,288
  Apr-12	34	8,432
  May-12	35	7,455
  Jun-12	36	5,346
  MAD =
  This spreadsheet is created by Professor Ronald Lau to accompany the teaching note, Reference No.: UST005/TN/1808 (HBP Product No.: ST5T), of the case : Chinese Pharmaceuticals (HK) Limited: Effective Forecasting for Optimal Inventory Management, Reference No.: UST005/1808 (HBP Product No.: ST5). © 2012 by The Hong Kong University of Science and Technology. This publication may not be digitized, photocopied or otherwise reproduced, posted, or transmitted without the permission of the Hong Kong University of Science and Technology.

  You should exclude the data of the first two cycles (24months) when calculating the average error (Mean Absolute Deviation), as it takes time for exponential forecasting model to establish before providing an accurate demand forecast

  Student+Adaptive ES

  HBP Product No.: ST5WS
  UST005/WSS/1207
  Adaptive Smoothing Model
  MM-YY	Period	Sales	Forecast	Abolute	% Error	Absolute %
  t	St	Ft	Error	(in dec.)	Error (in dec.)	Formula
  Jul-09	1	3,924	F2 =	S1
  Aug-09	2	2,619	3,924	1,304	0.498	0.498	Ft+1 =	Alpha t+1*St + (1-Alpha t+1)*Ft
  Sep-09	3	6,920	3,274	Percent Error: PEt =	(St – Ft) / St * 100%
  Oct-09	4	5,676
  Nov-09	5	8,348	Alpha t+1 =	0.00001, if | PE t | = 0
  Dec-09	6	6,044	Alpha t+1 =	0.99999, if | PE t | > 1
  Jan-10	7	6,877	Alpha t+1 =	| PE t | otherwises
  Feb-10	8	6,535
  Mar-10	9	6,395
  Apr-10	10	6,684
  May-10	11	5,414
  Jun-10	12	3,180
  Jul-10	13	4,350
  Aug-10	14	3,175
  Sep-10	15	6,935
  Oct-10	16	6,356
  Nov-10	17	8,919
  Dec-10	18	7,146
  Jan-11	19	7,763
  Feb-11	20	7,397
  Mar-11	21	7,286
  Apr-11	22	7,498
  May-11	23	6,386
  Jun-11	24	4,209
  Jul-11	25	4,825
  Aug-11	26	3,764
  Sep-11	27	7,066
  Oct-11	28	7,015
  Nov-11	29	9,535
  Dec-11	30	8,278
  Jan-12	31	8,773
  Feb-12	32	8,393
  Mar-12	33	8,288
  Apr-12	34	8,432
  May-12	35	7,455
  Jun-12	36	5,346
  MAD =
  (MAD for last 12 months only)

  You should exclude the data of the first two cycles (24months) when calculating the average error, as it takes time for exponential forecasting model to establish before providing an accurate demand forecast

  Student+Full ES

  HBP Product No.: ST5WS
  UST005/WSS/1207
  Exponential Smoothing with Trend and Seasonality Model
  MM-YY	Period	Sales	Level	Trend	Seasonality	Forecast	Absolute	% Error	Absolute %
  t	St	Lt	Tt	It	Ft	Error	(in dec.)	Error (in dec.)
  Jul-09	1	3,924	5718	0	0.686	3,924	Alpha =	0.1	Beta =	0.2	Gamma =	0.15
  Aug-09	2	2,619	5718	0	0.458	2,619
  Sep-09	3	6,920	5718	0	1.210	6,920	Initialization (1<= t <= 12) Note: This procedure helps determine the initial values of Seasonality for the first year
  Oct-09	4	5,676	5718	0	0.993	5,676	Set	Ft =	St	Tt =	0	Lt =	average of first year sales
  Nov-09	5	8,348	5718	0	1.460	8,348	It =	St / average of first year sales
  Dec-09	6	6,044	5718	0	1.057	6,044
  Jan-10	7	6,877	5718	0	1.203	6,877	Formula (for t>12)
  Feb-10	8	6,535	5718	0	1.143	6,535
  Mar-10	9	6,395	5718	0	1.118	6,395	Lt =	Alpha (St / It-c) + (1 – Alpha) (L t-1 + Tt-1)
  Apr-10	10	6,684	5718	0	1.169	6,684	Tt =	Beta (Lt – Lt-1) + (1 – Beta) Tt-1
  May-10	11	5,414	5718	0	0.947	5,414	It =	Gamma (St / Lt) + (1-Gamma) It-c
  Jun-10	12	3,180	5718	0	0.556	3,180
  Jul-10	13	4,350	5,780	12	0.696	3,924	F t+m =	( Lt + (Tt * m) ) * It-c+m	(for m-step-ahead forecast)
  Aug-10	14	3,175	5,906	35	0.470	2,653
  Sep-10	15	6,935	5,921	31	1.204	7,191
  Oct-10	16	6,356	5,997	40	1.003	5,908
  Nov-10	17	8,919	6,044	41	1.462	8,813	For F13 to F36:
  Dec-10	18	7,146	6,153	55	1.073	6,433	Ft +1= (Lt+Tt)*It+1-c
  Jan-11	19	7,763	6,233	60	1.209	7,466	e.g. F17= (L16+T16)*I5	(seasonal cycle c = 12)
  Feb-11	20	7,397	6,310	63	1.147	7,191
  Mar-11	21	7,286	6,388	66	1.122	7,128
  Apr-11	22	7,498	6,450	66	1.168	7,544
  May-11	23	6,386	6,539	70	0.951	6,169
  Jun-11	24	4,209	6,705	89	0.567	3,675
  Jul-11	25	4,825
  Aug-11	26	3,764
  Sep-11	27	7,066
  Oct-11	28	7,015
  Nov-11	29	9,535
  Dec-11	30	8,278
  Jan-12	31	8,773
  Feb-12	32	8,393
  Mar-12	33	8,288
  Apr-12	34	8,432
  May-12	35	7,455
  Jun-12	36	5,346
  Average Error =

  You should exclude the data of the first two cycles (24months) when calculating the average error, as it takes time for exponential forecasting model to establish before providing an accurate demand forecast

  Student+Inv. Development

  HBP Product No.: ST5WS
  UST005/WSS/1207
  Replenishment Template of Notoginseng Capsules (Lead time = 2 months)
  Warehouse capacity of inventory	25,000	Safety stock =
  MM-YY	Beginning Inventory (Book Record)	Stock to be received in the month	Stock in Transit (to be received in next month)	Inventory Position	Inventory required covering lead time & review period	Order Quantity	Actual Demand	Forecast Demand	Buffer (gift for promotion)	Forecast & Buffer
  Definition	(=Beg. inv. of last month + stock received – actual demand)	(= Order quantity 2 months ago)	(= Order quantity 1 month ago)	(= Beginning inventory + Stock receipt + Stock in transit)	(= Sum of forecast & buffer for the current month and lead time)	(= Forecast demand over vulnerable period + Safety stock – Inventory position)	Given	Results from Forecast Model	(= forecast*0.2)	(= forecast + buffer)
  Mar-11	26,662	0	0	0	Legend	7,286	4,813
  Apr-11	0	0	0	Order	7,498	4,244
  May-11	1000	No need to order	6,386	3,727
  Jun-11	4,209	3,275
  Jul-11	4,825	3,664
  Aug-11	3,764	3,569
  Sep-11	7,066	7,833
  Oct-11	7,015	8,377
  Nov-11	9,535	11,224
  Dec-11	8,278	8,315
  Jan-12	8,773	8,328
  Feb-12	8,393	7,616
  Mar-12	8,288	7,375
  Apr-12	8,432	7,482
  May-12	7,455	6,567
  Jun-12	5,346	5,033
  Jul-12
  Aug-12
  Sep-12
  Oct-12
  Nov-12
  Dec-12
  Jan-13
  Feb-13
  Mar-13
  Apr-13
  May-13
  Jun-13