GROUP PROJECT
Pasquet Ferdinand
Ravera Quentin
I- Presentation of the company
The company was founded by Walter Chrysler (1875–1940) on June 6, 1925. The headquarters are in Auburn Hills in the Michigan.
The brand was at the beginning a premium luxury position competing with the brands
Cadillac. Chrysler was the top brand in the portfolio of what was then known as Chrysler
Corporation.
Chrysler's positioning changed several times over the years. Indeed, The Chrysler brand was originally a premium luxury position competing with Cadillac and Lincoln through the cars
Imperial.Then, Chrysler's positioning of the Chrysler brand towards a mid-price brand caused
Chrysler to kill the brand DeSoto after 1961.After that, in 1975, Chrysler once again became the top brand, with range of products competing with both luxury and near luxury brands like Buick. During the 1980s the Chrysler brand took on a similar role as Oldsmobile or
Mercury by offering entry-level luxury cars of various types and sizes.
Before the Fiat's acquisition of a 20% stake in Chrysler in 2009, Chrysler had a discontinued line of cars with economy car and sport cars. This strategy ended when the company was threatened with an imminent bankrupt in 2008.
That’s why, in December 2008, U.S. President George W. Bush agreed to a $4 billion bailout for Chrysler and another one of $8.5 billion was also given by the President Barack Obama early in his administration. Chrysler LLC (new company emerged from a bankruptcy reorganization) received loans for three months in order to come up with specific plans to restructure into viable companies.
Nowadays, Fiat decided to make a new turn toward the luxury market.
II- Why we chose Chrysler
For this project, we decided to choose a venture which is an active actor of the current economy on a dynamic market. Indeed, this month of February 2012 was a monster month for car sales: Auto purchases rose 20%. At the beginning of 2012, analysts predicted that
American drivers would buy a million more cars this year, perhaps hitting the 14 million mark after 12.8 million purchases last year.
The two main companies are Ford and Chrysler. But Chrysler showed the biggest growth of all. The automaker is expected to sell about 126,000 cars in February, or 25% more than in January and 27% more than in February 2011.
Moreover, Chrysler was an interesting choice thanks to its old and well known desire to improve its relations with its supplier and to control as efficient as possible the supply chain to face the foreign competition. For example, since the 1970s American automakers have been facing strong global competition both at home and abroad. Japanese carmakers in particular made a significant impact on the industry by introducing smaller, less expensive, and more fuel-efficient in the United States. This coincided with the oil crisis, which resulted in higher gasoline prices and a shift in consumer tastes toward greater fuel efficiency. Other advantages of the Japanese automakers resulted from their use of just-in-time (JIT) inventory controls, modern manufacturing techniques, quality control and management practices. In order to counter to the « land of the rising sun », Chrysler and the others american companies had to find new solutions.
We can quote the time of Recovery and Restructuring in the 1980s when Chrysler decided to develop on a limited budget, the Dodge Aries and the Plymouth Reliant, code-named the "K-cars". Cars were designed to compete with imports and enjoyed early sales success, which Chrysler rode to profitability in 1982. Chrysler had cut its inventories by $1 billion, reduced white-collar staff by 50 percent, and cut its break-even point by 50 percent in drastic efforts to manage its finances.
Moreover, in 1989, through benchmarking of competitors, listening to suppliers, and experimenting with new ideas and programs, Chrysler's management gradually developed a vision of the changes they needed to make. It was apparent that most delays and cost overruns were finally attributable to the traditional sequential design approach that Chrysler followed all these years, with different functional groups responsible for different steps in the product development process. This resulted in poor coordination and conflicts between the functional groups, often requiring redesign and retooling. A great deal of time was wasted arguing and arbitrating, and every fight that couldn't be resolved had to be decided by a senior executive. What was needed was a new approach that combined individuals from the different functional groups into an integrated product development team.
Another critical element of this process was the relationship between Chrysler and its suppliers. Suppliers were at the same time both essential to and largely uninvolved in the product development process. About two-thirds of the components Chrysler uses in manufacturing come from outside sources. In 1996, after greatly reducing its number of suppliers in preceding years, Chrysler was still purchasing 60,000 different items from 1,140 different suppliers and spending about $34 billion in the process. If Chrysler could involve the suppliers earlier in the process, it might realize time and cost savings by avoiding design problems that were often not discovered until suppliers' components had already been ordered.
More currently, we can not forget to talk about the SCORE program of Chrysler. The carmaker, in August 1989, established a plan which, in an effort to work closely with its suppliers and identify opportunities to reduce costs from the vehicle manufacturing process, allowed both Chrysler and its suppliers to share in benefits of SCORE-related cost savings. From the time the program was introduced to about 1996, Chrysler entertained more than 12,000 supplier-submitted proposals and saved an estimated $25 billion. For 1996 alone, Chrysler Corporation and its suppliers achieved $1 billion in cost savings through the SCORE program. As Thomas T Stallkamp, the Executive Vice President of Procurement and Supply Chain and General Manager of Minivan Operations, “Because we involve our suppliers early-on during the vehicle design process, we are able to discover opportunities to drive out costs before we manufacture an automobile”.
Thus, the SCORE program is an excellent example of how Chrysler is committed to working with suppliers as partners in a long-term relationship. Instead of deal every single part with the lowest bidder, Chrysler have established an honest two-way communication that allows them to work with their suppliers and build a cost efficient automobile.
When we cross all these elements, the bankruptcy which seemed imminent few year ago and the recent increase in the sales, we understand why Chrysler is a relevant topic in Operations Management. That’s why we have chosen this mythical venture.
III – Forecasting
Main source used: Forecasting : Methods and application –Third edition (Spyros Makridakis, Steven C.Wheelwright, Rob J.Hyndman)
Forecasting definition relies on two criterias: time, and status. Thus, a company can make a short-range forecast, medium-range forecast or long-range forecast, as well as choose whether this forecast will be economical, technological or dedicated to demand assumptions. In this case, we chose to create a medium-term demand forecast, for it seemed the most important to us for Chrysler.
“Frequently there is a time lag between awareness of an impending event or need and occurrence of that event. This lead time is the main reason for planning and forecasting”.
If this lead time is zero or very small, there is no real need for planning and forecasting, as it would not necessarily benefit the company and may be time-consuming.
However, successful forecasting is not always directly useful to managers and others. Also, one needs to make the distinction between uncontrollable external events (forecasting), and controllable internal events (decision making).
There is a wide variety of forecasting methods available to management. Thus, this leads the company in taking three different steps :
- establish goals and objectives
- predict environmental factors
- select actions that will hopefully result in attainment of these objectives.
Forecasting, nowadays, is primarily used for 3 different purposes : scheduling (production, transportation, cas, personnel, …), acquiring resources (raw materials, hiring personnel, buying machinery, …), and determining resource requirements.
On a units sold chart, there are 4 possible remarquable patterns : horizontal (the data values fluctuate around a constant mean), seasonal (e.g quarter of the year, month, day of the week), cyclical (when the data exhibit rises and falls are not of a fixed period), and trend.
Chrysler and Forecasting
In Chrysler’s case, the pattern is interesting, for it does not really answer any of those patterns (see next diagrams). January is usually a bad month for sales, but otherwise, the rises and falls mainly depend on the launch of a new product, or a sudden rise/fall of the demand.
We chose to elaborate the demand forecast of this wealthy company, to see if the actual crisis and the economic issues that come out from it can endanger Chrysler for the next coming years. In the year 2011, overall turnover for the auto market in USA rose from about 20%, which leads us to think this sector is growing further and further. Thus, we wanted to see if Chrysler, one of the four major American companies of this sector could live up to these expectations for the year 2012 (and, to an extent, 2013).
By following closely the steps needed to create an efficient forecast, we chose to determine which forecast method is the best, and if a simulation of the demand could correspond to those findings. More generally, we wanted to test our forecast skills and our ability to make correct assumptions and useful recommandations.
The four main steps of forecasting are as follows :
1) Set objectives
2) Obtain forecasts (from existing sources)
3) Evaluate the forecast
4) Persuade the others of its accuracy
We will now make steps 2) and 3), in order to make recommandations.
First thing we did was to draw the Moving Average and Weighted Moving Average curves, to see which is the most accurate for the company. The figures we used (units sold over the period we chose) were those given by the company itself (official figures). We drew the Actual Sales curves, and calculated the Moving Average (over 3 months each time, and
Weighted Moving Average (3 months also, but with coefficient : 3 for the closest month, 2 for the second closest, and 1 for the furthest). This is what we obtained.
Actual Sales/ MA/WMA of Chrysler for the period January 2011 to February 2012
Months Actual Sales Moving Average Weighted Moving Average January 2011
70118
February 2011
133521
March 2011
121730
April 2012
117225
108456,33
117058,3333 May 2011
115363
124158,66
121442,6667 June 2011
120394
118106
117044,8333 July 2011
112026
117660,66
118188,8333 August 2011
130119
115927,66
115371,5 September 2001
127334 120 846,33
122467,1667 October 2011
114512
123159,66
125711 November 2011
101172
123988,33
121387,1667 December 2011
138019
115339,33
109979 January 2012
109 149 117 901
121818,8333 February 2012
133 321
116113,33
117442,8333
Curves of Actual Sales, MA and WMA over this period
What we can see is that the two curves (MA and WMA) are very close to each other, but we chose to take MA’s curve over WMA’s for it seems there is little, if not no trend. Overall, these results don’t seem very accurate, and rather far from the truth. This is why we chose to include an exponential smoothing curve. The value of we used here is 0.5 because the average is likely to change (allowing us to be more accurate). The following charts and curves show our results. Months Actual Sales Moving Average Exponential smoothing using =0.5 January 2011
70118
February 2011
133521
March 2011
121730
April 2012
117225
108456,33
112840,665 May 2011
115363
124158,66
119760,83 June 2011
120394
118106
119250 July 2011
112026
117660,66
114843,33 August 2011
130119
115927,66
123023,33 September 2001
127334 120 846,33
124090,165 October 2011
114512
123159,66
118835,83 November 2011
101172
123988,33
112580,165 December 2011
138019
115339,33
126679,165 January 2012
109 149 117 901
113525 February 2012
133 321
116113,33
124717,165
Chart showing Moving Average and Exponential Smoothing over the period chosen
0
20000
40000
60000
80000
100000
120000
140000
160000
Actual Sales
Moving Average
Weighted Moving Average
Curves of Actual Sales, Moving Average and Exponential Smoothed Average over the period
The results now seem way more accurate, even if some forecast errors can easily be detected. We chose to stop our calculations here, and evaluate the forecast errors (Absolute deviations) for each month, as well as the Mean Absolute Deviation which shows us an average of these errors over the period chosen. Months Actual Sales Moving Average Exponential smoothing using =0.5 Absolute Deviation January 2011
70118
February 2011
133521
March 2011
121730
April 2012
117225
108456,33
112840,665
4384,335 May 2011
115363
124158,66
119760,83
4397,83 June 2011
120394
118106
119250
1144 July 2011
112026
117660,66
114843,33
2817,33 August 2011
130119
115927,66
123023,33
7095,67 September 2001
127334 120 846,33
124090,165
3243,835 October 2011
114512
123159,66
118835,83
4323,83 November 2011
101172
123988,33
112580,165
11408,165 December 2011
138019
115339,33
126679,165
11339,835 January 2012
109 149 117 901
113525
4376 February 2012
133 321
116113,33
124717,165
8603,835
MAD
5739,515
0
20000
40000
60000
80000
100000
120000
140000
160000
January 2011
February 2011
March 2011
April 2012
May 2011
June 2011
July 2011
August 2011
September 2001
October 2011
November 2011
December 2011
January 2012
February 2012
Actual Sales
Moving Average
Exponential smoothing using a=0.5 MAPE = 4.8%
Chart of Absolute Deviations over the period
Our MAD value (5739,515) seems quite low, which proves us our forecast is rather accurate. This value represents the average errors of our forecast for each month (less/more units sold than what was expected).
Our calculation of the MAPE value (4,8%) represents the overall percentage of these errors when compared to the average of the actual sales over the period. This value is also quite low (which confirms our findings).
Simulation
In order to further develop our demand forecast, we chose to create a simulation of the demand for the next 14 months, in order to see what production forecast Chrysler should use (by determining these values). Our calculations are stated in the next charts.
Demand for cars Frequency Probability of occurrence Cumulative Probability Interval of random numbers Less than 100 000 1 (out of 14) 0.07 0.07 01 through 07
100 000 to 105 000
1
0.07
0.14
08 trough 14 105 000 to 110 000 1 0.07 0.21 15 through 21
110 000 to 115 000
2
0.14
0.35
22 through 35 115 000 to 120 000 2 0.14 0.49 36 through 49
120 000 to 125 000
2
0.14
0.63
50 through 63 125 000 to 130 000 1 0.07 0.70 64 through 70
More than 130 000
4
0.30
1.00
70 through 100 Total 14 1.00
Months Random Number Simulated monthly demand March 2012 5 Less than 100 000
April 2012
34
110 000 to 115 000 May 2012 77 More than 130 000
June 2012
8
100 000 to 105 000 July 2012 45 115 000 to 120 000
August 2012
84
More than 130 000 September 2012 84 More than 130 000
October 2012
77
More than 130 000 November 2012 52 120 000 to 125 000
December 2012
69
125 000 to 130 000 January 2013 33 110 000 to 115 000
February 2013
32
110 000 to 115 000 March 2013 30 110 000 to 115 000
April 2013
48
115 000 to 120 000
These results allow use to make assumptions concerning the future of the company, if we use the previous forecast we made. We can see that, even if it is just a simulation, most of the monthly sales are actually quite good.
IV- Conclusion and Recommendations
Chrysler is indeed a wealthy company, even in such an unattractive market. Nowadays, as if the crisis seems to be long forgotten, car sales in the USA are growing more and more, which leads us to think Chrysler has a bright future awaiting.
However, we must take into account the instability of this market, the success of which depends a lot on the USA’s overall wealth and the production issues. Chrysler really needs to plan demand, for a significant error in forecast could lead to an important drop of sales. The forecast we made and the simulation we created show that the company could increase exponentially its sales, but there are a lot of factors/criterias to take into account. Thus, our recommendations would be as follows.
 Predict future demand by creating a lot of simulations, in order to predict shifts in the market and possible trends.
 Plan the production of cars accurately, using tools such as forecasts and surveys.
 Develop new strategies (long-term ones) that take into account launches of new products (from Chrysler as well as from its rivals), possible periods of economic crisis and possible changes of the law (at the State’s scale).
 Make production plannings for each country Chrysler uses to sale in (different ones each time), that take into account the wealth of the country, its laws, and the actual demand.
Bibliography
 Forecasting : Methods and application –Third edition, Spyros Makridakis, Steven C.Wheelwright, Rob J.Hyndman, 2006
 Chrysler’s figures and charts
 www.Chrysler.com
 Chrysler 300 Series : Pedigree, Power and Performance since 1955, Robert Ackerson, 2005