Production cost modelling by system dynamics approach: An application in white goods sector

In order to be sustain, companies have to protect their profitability. Today, increasing competition, technological developments, changing market needs from day to day, companies have become more dynamic than ever. Companies are aware that one of the most important ways to maintain or upgrade their profitability is right budgeting and cost control. In this context, this study aimed to provide an alternative budget tool that would allow decision makers to create ideas under different circumstances and to take action quickly. For this, system dynamics approach is used which takes complex systems as a whole, and basically the behavior that these systems show in a certain period, by using the dynamics of the system itself. System dynamics have been used as alternative budget instruments in the state-controlled sectors such as the service sector or the energy sector. But more usage in production planning, inventory control and used for similar studies. Also system dynamics is used for understading of macro economics relations in Turkey. With this study, performance of system dynamics has been observed in a different usage area. The study not only modeled the production costs of an industrial establishment, but it also revealed the relationship of the microeconomic system and macroeconomic system in a simple way. As in the study, white goods sector has been selected which is after China, Turkey became the second manufacturer in the world. The model, which is prepared according to the dynamics of the white goods sector, will enable each white goods manufacturer to produce the results by modeling their own dynamics. It is also possible to use it in sectors which have similar cost structure with white goods sector and whose main input items are sheet, plastic and electronic. In this study, a washing machine factory which is in Turkey's largest white goods manufacturer was selected and its dynamic's handled and production costs are modeled. In the scope of the study, firstly literature research on system dynamics was conducted while sector reports were also examined in order to determine the macro economic factors affecting the production costs of factories in white goods sector. At the same time, managers, budget and controlling experts and projet leaders who are currently working in three major white goods producer in Turkey are selected as focus group for survey and about 60 people participated in the survey. Data for the study were collected from the company, from the Central Bank of the Republic of Turkey and from the World Bank and the model was established with the help of software Vensim PLE. During the modeling phase, regression analyzes were conducted with the aim of supporting the published studies about relations of macroeconomic factors. As a result of these studies, the dollar rate, the interest rate and the GDP were found to explain inflation at 98%, and the H0 hypothesis was accepted because the P values were below 5%. In addition, the H0 hypothesis was accepted because population and inflation explained GDP at 99% and P's were less than 5%. After defining the macroeconomic relations, the relationship between production cost items and the macroeconomic factors were builded with the light of the sector reports and the survey results. The model has been revised until pass as successfully from the direct structure test, structure oriented behavior test, behavior pattern test, according to the published verification flow report in the literature. After the model successfully passed all tests, the model was run between 2007 and 2025. The model was evaluated in the extreme condition test in structure oriented behavior test step and two different extreme conditions were applied. The first extreme condition is defined and investigated as 40% increase in dollar exchange rate each year. In the second condition, the annual increase rate of 40% per year and only one year 50% decrease in dollar exchange rate. In both cases, the behavior of inflation and production costs was examined and similar behavior was observed with the literature and empirical studies. In the behavior pattern section, the relationship between model and actual data between 2007 and 2017 is examined. The first pair tested was the dollar rate and inflation, and the behavior observed as identical with the actual data. In the second pair, the dollar exchange rate and the production cost of company were examined. Again, the model is reliable because the actual results match the model results individually. In order to observe the performance after making sure the model is reliable, the data for 2007-2017 are compared with the actual data. As a result, items were observed as follows; general expenses (energy, maintenance, scrap, auxiliary materials, depreciation and the like) with 4% standart deviation, production costs with 6,6% standard deviation, labor costs with 6,7% standard deviation, material costs with 7,6% standard deviation, production unit with 3.8% standard deviation, inflation with a standard deviation of 3.8% and a GDP with 6,8% standard deviation. With the this results, dollar exchange rate and Turkey's population were selected as the main scenario makers for four different scenario which are provide projection until 2025. In the first scenario, the dollar rate linearly increased by 3% every year from 2018 to 2025, while other parameters, such as population, interest rates, global commodity prices and oil prices, were considered constant. In the second scenario, the dollar rate increased by 20% and after increased year, exchange rate was decreased 2% in the each following years. Other parameters are fixed as in the first scenario. The third scenario was produced as an alternative to the first scenario. While the dollar was accepted as the same as the first scenario, the population was accepted according to TUIK's 2025 population projections. According to TUIK' projection, Turkey' population is estimated to increase by 9% in between the years 2017-2025. In the third scenario, where the dollar and the population are scenario determinants, other parameters are fixed. The fourth scenario, the last scenario, was established as an alternative to the second scenario. While the dollar exchange rate fluctuated as in the second scenario, the population parameter was accepted according to TUIK's data as in the third scenario. Other parameters are fixed. As a result of projections, unit energy costs showed only the same increases in the second and fourth scenarios. This is the result that the population factor is not directly related to energy costs. Another parameter, unit material costs, is clearly seen with the results that there is no effect from the population factor. Apart from this, unit material cost is very sensitive to exchange rate change. The average wage and inflation rate is affected by the population factor by exhibiting different behavior from energy and material costs. They have shown a more limited increase in population growth scenarios than in the first and second scenario. This has also been proven by previous empirical studies on the relationship between economic growth and population. Restricted economic growth, which prevented population growth from growing economically, also reduced the rate of increase in total domestic prices. As a result, a less inflation rate incerese is seen than in the scenario which population is constant. Finally, the unit production costs are a representing the total of parameters, which is a mix of the effects of the other parameters. For this reason, in the last two scenarios, there is a slight decrease in the rate of increase of unit production costs. It would not be wrong to comment on the fact that it is caused by the cost items of the inflation effect included in the production costs. It also emphasized that factories should always keep their labor costs under control over the coming years and will need to closely monitor technological developments, automation investments and industry 4.0 applications. In conclusion, this study is planned to develop and promote the structure of production costs of companies in white goods sector with the system dynamics model. The aim of the model is to prove that system dynamics can be used as an alternative budget tool in an industrial enterprise. At the same time, it is aimed to model the structures that are related in both economic systems, not only macro or micro economic systems, both of them like real systems. Because no microeconomic system can be considered independent of the macroeconomic system, it is aimed to reflect the effects of macroeconomic structure on the model. System dynamics is a methodology for generating very fast output for managers, decision makers and policy makers in order to take a action in complex systems with many variables. In this context, this study has been used as an alternative planning tool for the producers in the white goods sector. Any business in the white goods sector will be able to produce output by reflecting its own dynamics. Relations and equations related to macroeconomic factors are available for all white goods producers in their current state. The model will help create a quick forecast for decision makers by comparing many different scenarios at the same time. The results are hopeful, but there is no optimum points for system dynamics models. The presented model can be further improved or redirected for different purposes with new revisions to be made. Also, although 2025 projections have been made in this study, short-term planning will be more appropriate rather than long-term planning for system dynamics approach. The reason for this, long duration of the planned period will cause to an increase in assumptions, which may not provide the decision makers with the basis for action on the real system.