Foreign currency traders who use data for decision-making are likely to gain competitive edge in foreign exchange than those who rely on gut feelings.
The project area
Data for this study will be obtained from Exchange-Rates.org. Historical exchange rates between the British Pound (GBP) and the US Dollar (USD) will be used for forecasting future exchange rates. Data for analysis will be based on the preferred periods.
The technical approach (TA) will be used in forecasting the exchanges for a given period. TA relies on a subset of historical data, particularly price data (exchange rates). This approach is technical because it does not account for fundamental economic determinants that could influence exchange rates. Instead, it relies on data extrapolation obtained from historical observed trends.
Technical analysis will yield certain patterns that will be used for forecasting exchange rates between the GBP and the USD. In time-series analysis, it is imperative to determine an appropriate length of intervals because they determine the effectiveness of forecasting (Tayal, Sonawani, Ansari and Gupta 132-135). Research shows that the length of the intervals influences the accuracy of the forecast. Thus, effective selection of the length of intervals could significantly enhance the accuracy of forecasting outcomes (Tayal et al. 132-135).
SPSS or any other effective analytical tool will be used for data analysis to identify significant trends that can support decision-making. Foreign currency traders will use these significant trends to buy or sell their currencies.
A research question
Can data-driven decision-making create competitive edge for foreign currency traders?
Significance
In most cases, international transactions may take time be settled. In this regard, traders should assess exchange rate forecasts based on foreign currencies. In addition, foreign exchange traders must also understand future trends in their markets. Therefore, exchange rate forecasting is critical and can help traders to assess both benefits, risks and other challenges in trade.
Project Proposal
Forecasting has long been recognized as a fundamental tool across various industries. In the past, however, traders used various approaches to forecast and inform their decisions. One primary goal of understanding trends of exchange rate is to be able to forecast them. Given the fluctuation in exchange rate trends, forecasting exchange rates could be difficult without reliable data. The research question, therefore, would provide intellectual foundation to encourage foreign exchange traders to adopt data-driven decision-making rather than use gut feelings to make decisions.
Thomas H. Davenport noted that forecasting could be more difficult in retail industries because of several variables and constantly changing trends and influences on demands, alternative channels among others (Davenport 11). In most cases, there could be large amount of data that analysts can leverage to create accurate prediction of the exchange rate. Foreign exchange forecasting is critical for key decision-making processes, particularly in finance management, operations and budgeting among others. The use of gut feelings or manual forecasts cannot support effective decision-making. Such approaches consume management time and they could be difficult to understand due to lack of any meaningful data. In this regard, data-driven decision-making can assist managers to make effective decisions for competitive advantage. Davenport observed that there is little doubt, however, that the aggressive adoption and exploitation of analytics has led to competitive advantage among some of the worlds most successful retailers (Davenport 11). Therefore, data-driven decision-making can solve many challenges international traders face.
Works Cited
Davenport, Thomas H. Realizing the Potential of Retail Analytics: Plenty of Food for Those with the Appetite. Babson Park, MA: Babson Executive Education, 2009. Print.
Tayal, Devendra, Shilpa Sonawani, Gunjan Ansari, and Charu Gupta. Fuzzy Time Series Forecasting of Low Dimensional Numerical Data. International Journal of Engineering Research and Applications (IJERA) 2.1 (n.d): 132-135. Print.
In the hospitality sector, demand for accommodation rooms fluctuates daily, seasonally and annually. Analysis of these fluctuations in demand produces unreliable forecasts. Effective revenue management realization requires the development of an appropriate forecasting mechanism. Attainment of reliable forecasts is never easy even though the process of forecasting is simple.
Hence, Cross, R. G. (1997) suggested some rules applicable in forecasting. Cross suggested that the prediction must stay at comprehensive level. Detailed predictions contribute significantly to current Hotel Revenue Management Systems.
The second suggestion is that a large amount of information has to be used in the investigation. Additionally, the predictions must be adjusted regularly to potential changes in the business environment.
Precise predictions are vital in improvement of price and availability suggestions for hotel rooms. In addition, precise forecasting improves decisions made on recruitment of workers, purchase of goods and budget preparations.
In contrast, incorrect forecasting results into adoption of inefficient decisions on price and availability suggestions that the revenue management systems produce. Such inefficient decisions affect the revenue of a hotel negatively.
Talluri (2004) identified two forms of revenue management predictions. The first kind is quantify-dependent revenue management prediction. This is mainly used in aviation and hospitality industries. The other type is price-dependent revenue management prediction.
Apart from demand information, quantify-dependent revenue management prediction needs data on the arrival of reservation requests from different types of clients. This data is obtained at the time when clients make orders. This means that the information utilized in lodge demand prediction depends on present reservation activities, past data that relates to every day arrivals or the number of sold rooms.
Furthermore, this kind of prediction requires guess on cancelled reservations and the number of clients who do not show up. A reservation is grouped as No Show if the client who made the request does not terminate it before the deadline of the hotel reaches and does not come to claim it.
Prediction used in hotel revenue management should consider two variables that relate to time. These two variables are reservations and consumption times.
There exist other vital issues considered apart from the prediction method chosen. Some of the key issues are the forecast period, aggregation levels and measurement of the predictions precision. Weatherford et al (2001) proved that completely disaggregated predictions provide dependable outcomes compared to partially aggregated strategies.
Therefore, it is appropriate for hotels to examine arrivals by the duration customers spend in the facilities and the category of price. One key concern to take into consideration is the method applicable in time division to produce an appropriate basis for prediction. A common practice is to consider a week as a forecast. In this method, each day is considered differently.
For example, predictions for Mondays rely on information collected on other Mondays and so on. Another vital point to consider is the periodic phenomenon, which has significant influence in the hotel industry. Consideration of limited data set limits managers ability to capture seasonality. Conversely, the use of many periods may make predictions made unresponsive and rigid.
Managers physically developed predictions through analysis of historical data before the introduction of revenue management systems. They analyzed the length of stay in hotels and the price categories. The process lacked sophistication and consumed a lot of time. Consequently, the manual system is inappropriate in the current business environment due to stiff competition and fast-paced market activities.
Prediction strategies can be grouped into three categories namely historical, advance and combined forecast models. Historical reservation models only take into consideration total rooms booked on earlier nights. It considers the number of rooms booked and arrivals at a hotel. Advance reservation models take into consideration the booking behavior of customers over a given period.
Finally, joint reservation methods consider features of both chronological and advance methods to establish predictions. Weatherford and Kimes, (2003) showed that exponential smoothing, pickup and moving standard forms are the most dependable prediction strategies.
However, the performance of these strategies depends on the data available. Hence, hotel revenue managers must have adequate prediction models. Additionally, they should use unlike strategies at different times and in diverse markets.
Forecast Accuracy
Accuracy is a vital component in prediction since it determines the method selected. Numerous measures of the performance of the predictions made exist. The Mean Absolute Deviation (MAD) is the easiest and popularly used method. It involves determination of the averages of the absolute values of prediction errors made. The Mean Percent Error (MPE) is the mean of the percentage errors made.
The Mean Absolute Percent Error (MAPE) and MPE are similar. However, MAPE averages the total figures of the prediction errors. The Root Mean Square Error (RMSE) takes into consideration the square root of the averages of the squared prediction errors.
MPE and MAPE are appropriate since they provide limitless figures. Nonetheless, they stay unclassified when the total number of reservations is zero. Hence, Thiels Inequality Coefficient is used to overcome the disadvantage. The coefficient is denoted by U in which F* and F are exact and predicted reservations.
Pickup Model
Pickup model is a well-known advance reservation strategy that takes advantage of the unique features of booking information. It relies on reservation information instead of reliance on arrival histories to make dependable predictions.
However, it provides variations in the predictions. The model can be further divided into preservative or multiplicative, standard or superior and straightforward or subjective average prediction strategies.
Judgmental Forecasting
Human judgment plays a significant role in revenue management practice. This is despite the ability of a revenue management system to perform multiple tasks in the process. Conventionally, predictions include human opinions. Judgmental prediction involves the inclusion of human opinions. A different research exhibited that opinions can be used to make adjustments in past predictions.
A revenue manager can then supervise the prediction process using the adjustments. Moreover, hotels can make appropriate use of revenue management systems in educating managers. A well-trained revenue manager can know how to analyze data and make appropriate plans. This can result into efficient operations in a hotel. Notably, business people agree that human judgment play a pivotal role in predictions.
Human judgment plays a vital role in generation of accurate predictions. However, human judgment is inclined towards biasness. Individuals may make decisions or predictions that can result into achievement of their personal goals. In addition, the differences in skills and abilities affect peoples judgment. Hence, well-trained people can make appropriate judgment compared to undertrained persons.
In early prediction studies, Hogarth and Makridakis, (1981) examined the effectiveness of human opinions in prediction generation. They concluded that statistical predictions were dependable compared to judgmental forecasts. Their study found that human predictions had biasness and errors. Human judgment involved much control and humans had overconfidence.
Fischhoff, (1988) noted that appropriate forecasts required human judgment in the selection of the model used, determination of parameters and investigation of study outcomes. There are two reasons for need of accuracy in judgmental prediction. Experts have adequate data and are able to acquire information in time.
There exists some features that can assist revenue managers improve their judgment. Revenue management systems are acceptable in prediction generation. The perception of forecasters on revenue management systems affect decisions or forecasts made. Secondly, revenue management systems are easy to manage, and they enable comparison of statistical predictions made.
Hence, they assist revenue managers to determine errors easily. Additionally, revenue management systems provide flexibility in methods used in generation of forecasts. This makes forecasters feel responsible and involved in revenue management. Revenue management systems also make managers employ correct techniques.
Comprehensible assistance and clear methods eliminate confusion and mistakes in generation of predictions. Revenue management systems also support the integration of human judgment and statistical methods in generation of demand forecasts. Reliable revenue management systems produce dependable statistical predictions. They also support judgmental adjustments in generation of demand forecasts.
An income manager can examine the effectiveness of a revenue management system. The manager can examine the usefulness of a database and the appropriateness of statistical techniques. In addition, the manager can compare the error measurement methods used in generation of demand predictions.
Examination of these factors can assist a manager to establish whether a revenue management system can enable development of appropriate forecasts. Integration of the mentioned features assists managers to design revenue management systems that enable development of precise predictions.
Economic, Technological and Operations Management Demand Forecast
Forecasting is very important in management. It is majorly concerned with the use of past and present trends in the market to determine what may happen in the future (Clements & Hendry, 1998). It helps firms, organizations and countries prepare in advance on how to deal with future uncertainties. Despite having many similarities, economic forecasting, technological forecasting and operations management demand forecasting also have significant differences. The main difference between the three types of forecasting is the element that is foreseen (Clements & Hendry, 1998). Precisely, technological forecasting deals with the anticipation of possible changes in technology and its effects on other operations (Henry, 1991).
Therefore, it solely deals with technological inventions and their impacts (Henry, 1991). On the other hand, economic forecasting deals with future economic possibilities (Marquez, 2002). This is usually done by countries, individual firms and other organizations.
Unlike technological and economic forecasting, operations management demand forecast deals with the resources that organizations may need for efficient production in the future and possible changes in consumer behaviour (Kumar & Suresh, 2009). It is mainly concerned with the type and amounts of good and services their customers may demand in the future (Kumar & Suresh, 2009).
Strategic Importance of Forecasting to Human Resources, Production Capacity and Supply Chain Management
Forecasting helps ensure that the process of production is efficient. Managers who carry out analyses of market trends always ensure that the use of resources in the production process is controlled. The raw materials are used according to the number of products likely to be demanded. In the long-run, there are very minimal wastages.
Forecasting also ensures that the supply chain does not break at some point due to sudden changes (Molnar, A2010). A firm that carries out its forecasting process properly always prepares in advance for changes even before they occur. As a result, the production and supply processes are continuous since they are not affected by the changes. Such readiness also ensures that the production capacity of the firm is not affected negatively. In fact, good predictions always help firms maintain or increase their production capacities depending on the projected demand.
Qualitative and Quantitative Methods of Forecasting
Qualitative forecasting
The Delphi Method involves interrogating a panel of experts about their opinions on probable events (Naik, 2004). However, the interrogation is not done in a boardroom. Instead, each of the experts is questioned independently. The questioning is then followed by a compilation of all their responses by an outsider, who then brings them more questions. This process repeats itself until a common stand is reached. Its main purpose is to prevent the experts from arriving at a consensus prior to the interview.
Advantage
This method completely eliminates the possibility of mob psychology and external pressure as the experts do not meet.
Disadvantage
The level of reliability is usually very low
The experts rarely reach a consensus.
Quantitative forecasting
The Simple Moving Method involves the analysis of trends in an adjustable period (Nikolopoulos, 2010). The length of the period does not change. Its main purpose is to quantify the performance of the business in given periods in order to use them in predicting future trends. This method is commonly used by mobile phone manufacturers in determining the number of devices people buy each year.
Advantage
It is more reliable since it compares trends over a uniform period. Its predictions are usually very accurate.
Disadvantage
It is not good at predicting sudden changes.
References
Clements, M., & Hendry, D. (1998). Forecasting economic time series. Cambridge: Cambridge University Press. Web.
Henry, B. (1991). Forecasting technological innovation. Dordrecht: Kluwer Academic. Web.
Kumar, S., & Suresh, N. (2009). Operations management. New Delhi: New Age International. Web.
Marquez, J. (2002). Understanding Economic Forecasts. International Journal of Forecasting, 18(3), 464-466. Web.
Molnar, A. (2010). Economic forecasting. New York: Nova Science Publishers. Web.
Naik, G. (2004). The structural qualitative method: a promising forecasting tool for developing country markets. International Journal of Forecasting, 20(3), 475- 485. Web.
Nikolopoulos, K. (2010). Forecasting with quantitative methods: the impact of special events in time series. Applied Economics, 42(8), 947-955. Web.
The second part of the Business Report for Balogne Pty. Ltd addresses the issues of financial forecasting and the sphere of business and technology. Microsoft Excel has been applied to the creation of four graphs and one table demonstrating the companys predicted performance based on such key indicators as earnings, earnings by product, variable costs, and sell prices. It has been explained how certain market influences, such as inflation/deflation, can affect the calculations of predictions. Further, contemporary business issues have been addressed, including knowledge management, business intelligence, Five Forces, and security.
Introduction
Upon reflecting on the role of business modeling and process modeling in decision-making, the use of Microsoft Excel for calculations, and the adoption of Web 2.0 applications for internal and external communications, the business analyst of Balogne Pty. Ltd, a company that produces several types of sugar products, turns to forecasting and strategic business planning.
For the newly appointed General Manager, it is important to receive predictions as per the companys expected performance and factors that may affect the operation in the nearest future. Three areas will be addressed as part of these business analytics efforts: forecasting financial performance (with the use of graphs demonstrating key indicators of financial dynamics), accumulating predicted financial data (with the use of a pivot table), and investigating contemporary issues related to technology and business, including knowledge management, business intelligence, market forces, and security.
Forecast Graphs
In financial planning, strategic decisions should be based on future projections; even though such projections accuracy may be questionable (due to the constantly changing market trends and influences), there is still a need for Balogne Pty. Ltd to build forecasts as per such crucial financial indicators as revenues, costs, and earnings and such considerations as the dynamics of these indicators among the companys products. The first prediction to be designed is the prediction of earnings; it should be based on the current indicators, inflation/deflation rates, and volume sold. The result of the prediction is presented below (Figure 1).
Earnings are calculated as the difference between revenue and cost: revenue, in turn, is the product of multiplying volume sold by the selling price, and the cost is the product of volume sold and variable cost plus fixed cost (overhead).
In 2016 and 2017, costs and prices are affected by the inflation/deflation rate, and there is growth/loss dynamics for volume sold, too. By considering all these factors and inputting data correctly into relevant software, such as a Microsoft Excel spreadsheet, a business analyst will be able to predict earnings and provide the General Manager with the information needed for strategic decision making. Further, looking into the dynamics of earnings distribution among different products is needed to improve product policies of Balogne Pty. Ltd (Figure 2).
As can be seen from the graph above, golden syrup is accountable for the smallest portion of earnings. White sugar sales generate the largest portion of earnings, and these earnings are expected to grow in 2017 (Figure 1). The product accountable for the second-largest portion of earnings is brown sugar (Figure 2), but it has been demonstrating a consistent decline in earnings generation within the three years (Figure 1).
However, as it has been mentioned, this information is based on certain predicted financial rates, and it should not be overlooked that these rates dynamics may be significantly different from the predictions. The use of such software as Microsoft Excel provides convenient opportunities for incorporating these considerations into a business analysis spreadsheet.
By modifying the rate in one cell (which contains a fixed inflation/deflation rate), a user can see the changes in numbers in all the cells that refer to inflation/deflation, i.e. the numbers that depend on the rate; also, the graph immediately changes and displays new numbers based on the newly determined rate. For example, variable cost per unit dynamics can be explored for the situation in which the inflation/deflation rate for 2016 and 2017 is not -5 percent and 20 percent respectively but 15 percent and 25 percent respectively (Figure 3).
When modifying the variable cost per unit dynamics, a user immediately modifies the total annual cost indicator and the total earnings level. However, it should be considered that the growth of variable cost per unit cannot pass unnoticed for the market, as it will affect the selling price per unit indicator. According to a new prediction, the latter rate in 2016 and 2017 is 25 percent and 35 percent respectively instead of 5 percent and 10 percent (Figure 4).
Along with the fixed cost, both indicators presented in the two graphs above will affect total earnings. Therefore, their dynamics should be monitored by a business analyst. Particular attention should be paid to the role of external market influences on the variable cost and sell price.
Pivot Table
Pivot tables can be successfully used for analysis in various spheres, and Microsoft Excel provides proper technologies for that. It has been claimed by researchers that a pivot table is a highly flexible contingency table (Dierenfeld & Merceron 2012, p. 116), which means that the format of a table allows modifying previously inputted data using formulas and connections established among different cells. A pivot table for Balogne Pty. Ltd is presented below (Table 1).
Table 1: Pivot Table of Balogne Pty. Ltds Earnings from 2015 to 2017 by Product.
Earnings Forecast by Product
Brown Sugar
2015
$6 447 500
2016
$5 673 800
2017
$4 198 612
Product earnings
$16 319 912
Caster Sugar
2015
$1 465 800
2016
$1 377 852
2017
$1 240 067
Product earnings
$4 083 719
Golden Syrup
2015
$407 000
2016
$431 420
2017
$457 305
Product earnings
$1 295 725
Icing Sugar
2015
$598 000
2016
$663 780
2017
$1 008 946
Product earnings
$2 270 726
Raw Sugar
2015
$3 553 000
2016
$4 050 420
2017
$5 063 025
Product earnings
$12 666 445
Sugar Cubes
2015
$329 400
2016
$467 748
2017
$725 009
Product earnings
$1 522 157
Treacle
2015
$713 000
2016
$470 580
2017
$414 110
Product earnings
$1 597 690
White Sugar
2015
$2 042 500
2016
$2 042 500
2017
$2 369 300
Product earnings
$6 454 300
Total 2015
$15 556 200
Total 2016
$15 178 100
Total 2017
$15 476 374
Total Product Earnings
$46 210 674
The table allows accumulating the earnings indicators within three years for each of the eight products. Also, it can be used for the creation of graphs, such as graphs presented in the previous section (see Forecast Graphs) to visualize the predictions and dynamics of the key indicators of the companys financial performance. In case the predictions change (e.g. by the changing inflation/deflation), the data in the table can be modified for each product, and the final indicators, i.e. total earnings, will be recalculated by the software.
Investigation of Contemporary Issues: Technology and Business
Further analysis of Balogne Pty Ltds current and expected performance will encompass the issues of knowledge management, business intelligence, Five Forces, and security. Concerning knowledge management, scholars pay much attention to the way the work of knowledge workers can be improved, optimized, and made more efficient, and the definition of a knowledge work that is widely adopted by scholars today includes both the utilization and creating of abstract/theoretical knowledge (Hislop 2013, p. 71). Also, an important aspect of knowledge work is that it is intellectual and mostly non-routine.
Therefore, it may present challenges for managing, monitoring, and evaluating because the product and value of intellectual work are hard to measure. Many business management strategies have been proposed, and a particular aspect of a companys performance addressed by them is a situation with senior workers in Balogne Pty. Ltd. Senior workers are used to certain practices and procedures that may be unfamiliar or poorly understood by new workers, and the former do not always document information related to these practices and procedures because it is a routine for them. As a result, business analysts are not supplied with complete information on the companys internal operation, and the effectiveness of analytics is undermined.
A way to address this complication is to establish a teaching and learning environment. Senior workers should be encouraged to share various aspects of their experience with new workers. As a result, the process of new workers improvement will be accelerated, and all the necessary documentation of processes in which senior workers are involved will be maintained because they will be asked to properly reflect on the practices they teach.
Also, additional training provided by senior employees will contribute to the knowledge management efforts of the company because relevant knowledge will be accumulated and exchanged. This practice will help the company pursue the purpose of business intelligence, i.e. to support better business decision making (Business intelligence n.d., para. 1) using collecting, storing, retrieving, exchanging, and analyzing information.
In analyzing Balogne Pty. Ltds supplier power, the Five Forces model designed by Porter can be applied (Dobbs 2014, p. 32). The model encompasses four kinds of external influencestwo bargaining powers (of suppliers and buyers) and two threats (of new entrants and substitutes)and one internal, ongoing influence: industry rivalry, i.e. the competition itself. For a company, it is important to recognize all five influences and adjust its operation to the changing conditions of the market. It is particularly relevant for Balogne Pty. Ltd because the company is currently growing, which means that its enlarging operation is more likely to face the impact of industry trends and possible external changes. Future sales can be particularly affected by the two threats recognized in Porters model.
Recent research confirms that sugar is a much more harmful substance for humans than it was previously believed (Groopman 2017); therefore, it can be expected that, in the nearest future, many new players will appear in the industry that will offer sugar substitutes, and many people may be willing to buy those substitutes instead of sugar and sugar products produced by Balogne Pty. Ltd.
Also, the company should not overlook the bargaining power of its suppliers, i.e. farmers, on future sales. In case suppliers face difficulties due to the decreasing demand for sugar, they may renegotiate the existing conditions of collaborating with the company. Depending on the extent of amending the current agreements, the company may face the need to reorganize its operation and its production facilities. Finally, there is a factor within the industry.
The companys competitors can manage to adapt to the arising conditions of decreasing demand for sugar, i.e. by offering substitutes, and this will mean that these competitors will address the same needs of the same customers as those they addressed before but in a new way. In the case of Balogne Pty. Ltd fails to make this shift in the nearest future, its indicators of successful financial performance will be damaged.
Finally, in terms of security, the new General Manager needs to take into consideration different threats, both external and internal. In this context, security is understood as the companys ability to collect, store, and retrieve information regarding its suppliers, customers, and internal operation safely. Two internal threats to proper security are poorly designed information systems and poorly trained employees. To address the former threat, the General Manager should ensure that the software used by the company for any operations with information is of good quality, regularly updated, and adjusted to the companys needs.
The processes of selecting and operating information systems should be coordinated by the management so that the risk of compromising important (and often confidential information) is minimized. Concerning the latter threatpoorly trained employeesit should be recognized that important data can be compromised due to the incorrect use of information systems or data collection instruments (McBride, Carter & Warkentin 2012, p. 1).
However, employees should be trained not only on the technical aspects of working with information (i.e. on how to use particular software) but also on corporate ethics, e.g. it should be generally understood and respected that certain types of information should remain confidential so that the company does not let down its suppliers, customers, or employees.
Conclusions and Recommendations
Balogne Pty. Ltd is expected to demonstrate moderately successful financial performance, as its earnings growth for most products, although the earnings for some popular products decrease. However, an important threat is recognized: the demand for sugar can decline significantly in the nearest future, which should prompt the company to look for alternative ways of addressing its customers needs.
Improved knowledge management strategies will allow pursuing the goals of business intelligence, and increased attention to the selection and operation of information systems and technology- and ethics-related training for employees will allow improving the companys security. Overall, it is assessed that the company, with its current performance, can overcome these challenges, and a properly designed business strategy will help address the issues outlined above.
Reference List
Business intelligence n.d. Web.
Dierenfeld, H & Merceron, A 2012, Learning analytics with Excel pivot tables, in Moodle research conference proceedings, Beuth University of Applied Sciences, Berlin, pp. 115-121.
Dobbs, ME 2014, Guidelines for applying Porters five forces framework: a set of industry analysis templates, Competitiveness Review, vol. 24, no.1, pp. 32-45.
Hislop, D 2013, Knowledge management in organizations: a critical introduction, 10th edn, Oxford University Press, Oxford, UK.
McBride, M, Carter, L & Warkentin, M 2012, Exploring the role of individual employee characteristics and personality on employee compliance with cybersecurity policies. Web.
Globalization has limited distances to the extent of words only. No more do companies or businesses think twice before reaching foreign markets to increase profits and customer base. Globalization isnt the name of the game. The concept involves interlinked processes such as logistics, supply chain, licensing, franchising, offshoring, outsourcing, etc. These activities have made life for multinational corporations easier than ever before.
Offshoring is when some internal functions of an organization are transferred to another country within the same firm (Schroeder, 2008). The idea is to cut costs. A firm can have several reasons for off-shoring such as cheap labor, abundant supply of skills, cheap technology, etc. Companies argue that if something can be produced cheaply abroad then why not import it rather than producing it at a high cost within ones own country. Off-shoring is controversial though. While reducing costs, results in increased unemployment and wage erosion, etc (Farrell, 2006). The critics against off-shoring say that in developed countries offshoring has more negative impacts than under developed countries. The reason is that people who lose jobs get a job at a lower level than the one they were employed at. This results in lowering down their buying power even though goods have reduced in prices due to off-shoring.
Outsourcing is another phenomenon that goes along with off-shoring. The basic motive behind both is cost reduction which is not possible without eradicating unnecessary expenditures. Outsourcing however is transferring an area of operations to another firm which can be native or foreign (Schroeder, 2008). In reality, in recent years purchase of off-the-shelf products has become as easier as it is stated in theory. Today, one almost can purchase just any function that is needed to run a company. Procter and Gamble for example have outsourced its IT infrastructure, human resource management, and even its products in the last three years (Engardio, 2006). Outsourcing though also results in loss of jobs and has been a very debatable topic. The issue however remains the same; cost cuts are no good if there is no demand for products that fall down in the parent country due to wage erosion and unemployment. In 2009, CEO of General Electric, Jeff Immelt, for example, urged the government to reduce outsourcing and increase the amount of workforce hired by manufacturing units as the United States could no longer depend on consumer spending to create demand (Bailey, 2009).
The concept that is gaining more popularity than off-shoring and outsourcing these days is offshore outsourcing. Offshore outsourcing is the transfer of activities from a domestic facility to another firm situated in another country (Schroeder, 2008). As already discussed the threats remain the same. However what the point that needs to be emphasized here is that when to use offshore outsourcing? To put it simple, if outsourcing or offshore outsourcing does result in work being done more effectively and less costly without jeopardizing other functions, then only it should be done (Kehal, 2006). A company shouldnt move its call center to India only because call center representatives can be hired at a lower cost. The company might lose its customer base for the simple reason that they are unable to communicate because of accent differences. This will result in cost savings at the expense of losing customers.
Despite disadvantages, it is seen that many of the firms around the globe are opting for outsourcing and off-shoring activities. Companies are relying heavily upon such activities to juice their performances. The near future holds better and enhanced prospects for such endeavors.
References
Farrell, D. 2006. Offshoring: understanding the emerging global labor market. Illustrated edition. Published by Harvard Business Press.
Kehal, H. and Singh, V. 2006. Outsourcing and off shoring in the 21st century. Published by Idea group Inc.
Schroeder, R. 2008. Operations Management: Contemporary Concepts and Cases. 4th edition. McGraw Hill publishers.
Bailey, D. and Kim, S. 2009. GEs Immelt says U.S. economy needs industrial renewal. Web.
Engardio , P and Arndt, M. 2006. The future of outsourcing. Web.
The oil industry has kept a prominent place in the worlds economics for over a century. Sales forecasting in this field has remained somewhat stable and has developed a number of reliable methods. Nevertheless, the changing trends and permanent conditions of economic mechanisms significantly influence oil sales forecasting. This paper will observe how the law of supply and demand and financial market trends impact sales forecasting for oil in the world.
The first factor affecting sales forecasting for the oil industry is based on the fundamental rule of supply and demand that defines the prices on the market and interactions between sellers and buyers. The paper by Arezki et al. (2017) has confirmed that the depletion of oil sources results in supply shortages and an increase in prices. The paper also comments on the possibility of new oil sources being discovered, immediately affecting the demand and supply ratio (Arezki et al., 2017). With this in mind, it is possible to alter sales forecasts based on the information about newly found or depleted oil fields when considering them as suppliers.
The other key element that influences oil sales is its relationship with financial market trends. According to Miao, Ramchander, Wang, and Yang (2017), three elements correlate with oil prices: U.S. interest rates, exchange rates, and the stock market. Since sales forecasting is directly dependent on global oil prices, all of those factors are widely used to obtain short-term prospects. For example, Miao et al. (2017) explain that due to producers trying to maintain the same purchasing power of the export revenue, the weakening of the U.S. dollar would lead to increased oil prices. Also, one could predict reduced demand for oil in countries whose currencies are dependent on the U.S. dollar, illustrating how financial market trends can be used in oil sales forecasting.
To summarize, several global factors have an impact on oil sales. This paper has observed two of them: the depletion of oil sources which activates supply shortage and higher demand, and the changes in the financial market, which forced producers to raise export prices. With the information on just these two elements of sales forecasting, one is able to develop a strategy for understanding and predicting oil price variations.
References
Arezki, R., Jakab, Z., Laxton, D., Matsumoto, A., Nurbekyan, A., Wang, H., & Yao, Z. (2017). Oil Prices and the Global Economy. International Monetary Fund. Web.
Miao, H., Ramchander, S., Wang, T., & Yang, D. (2017). Influential Factors in Crude Oil Price Forecasting. Energy Economics, 68. Web.
A companys planning activities are centered around finances, because the language and plans are states in terms of financial states, and measures to evaluate plans are financially focused as well. Furthermore, money remains the most critical resource for a commercial organization. Since practically every corporate action has financial consequences, the viability of any plan is seen through the lens of whether it is attainable in the context of limited financial resources (Higgins, 2019). Therefore, the process of financial forecasting can be described as using historic data to predict the financial future in a wide variety of contexts (Investopedia). Business managers may use forecasting to predict sale and plan operations, as many other practices such as cash flow, hiring, or projects depend on having the revenue to ultimately fulfill them. One of the most common tools used are pro forma financial statements, which is a prediction of a companys financials at the end of the specific forecast period. These can vary from detailed plans and budgets to simply rough estimates (Higgins, 2019). Other applications of financial forecasting can include investors predicting market trends and government economic agencies forecasting economic trajectories to guide policy.
It is important to distinguish forecasting from budgeting. Budgeting is a regular required practice where management plans expenses, quantifies expectation of revenues, and ultimately sets the financial direction for a company in the short-term (usually no more than a year), it is a baseline to which actual results are compared. Financial forecasts are long-term, used to determine how companies should allocated their budgets in the future, helping to generate long-term strategy in determining the viability of major projects, and unlike budgeting, it does not analyze the difference between forecasts and actual performance. Therefore, it is a difference of quantification and estimating. Financial forecasting remains important for companies, providing insight to business performance in the past and how it compares to the future. In turn, it allows executives to establish business goals that are both feasible and will help guide the long-term direction of a company to achieve the planned objectives and performance.
References
Higgins, R. (2019). Financial management (12th ed.). McGraw-Hill.
Formalized forecasting has massive benefits which involves the ability of the management of an organization to plan inventory levels, shift or add labor, decrease or increase production and determine whether or not to order raw material. It is also vital to mention that forecasting is a fundamental factor which affects the cost of raw materials and shipping cost such as petroleum increase, tax fluctuations and other overheads incurred by an organization. The margin or profitability of a firm is determined by the latter. On the same note, the overall productivity of a firm is also dictated by the ability to carry out thorough business forecasting.
The posterior demand of a product is a major consideration that is shaped by various types forecasting methods. This can equally enhance the process of evaluating the anticipated value of a product. Therefore, effective planning and shrewd utilization of the available resources can be put in place. In this case, forecasting limits inconveniences caused by delay of delivery of products to customers.
This is through the number of products required and the arte of producing them at the right time. It also factors in number of units to be produced, the inventory, cash flow, and amount of money that will be used for the process. Hence, augmenting distribution process and minimizing challenges associate with inventory are among the major attributes of business forecasting. The latter is also a vital business undertaking that makes it possible for firms to organizations to execute proper planning. The strategic supply chain management also requires effective business forecasting.
Prepare a weekly forecast for the next four weeks for each product. Briefly explain why you chose the methods you used.
In order to prepare a weekly forecast for roses, it is crucial to employ a simple method of time series. Using the latter approach, it is indeed possible to determine the forecast for the next four weeks. From the 14th week data, the sales result is 3.457. The latter is then multiplied by the number of weeks and then the value +48.28 is added. These forecasts will appear as demonstrated below. It will adopt the formula used for the 4 weeks within a period of four weeks. Hence, from the 15th to 18th week, the forecast is demonstrated below based on the data obtained from previous weeks.
Following falling profits and pressure from rising competition, I would adopt the market based sales forecast for Gold Coast blossoms. This will be based on market research to determine how much to produce and to sell. The figures will be above value based sales forecast.
This method helps find breakeven sales figures by dividing annual overheads by profit gross margin (selling price-direct cost *100). I will be able to determine the forecast for the next four weeks. From the 14 week data, the sales result will appear as demonstrated below.
Works Cited
William, Stevenson. Operations Management with Student DVD and Powerweb, 11th Edition, Boston: McGraw-Hill Irwin Publishing, 2005.Print.
This paper is devoted to the study of judgmental forecasting. It contains general data about this type of forecasting, cases of its application, examples of its most effective use, and its advantages and disadvantages. As an example of judgmental forecasting, the Delphi method is proposed, which involves identifying a consistent assessment of an expert group through an independent, anonymous survey of experts in several rounds, which involves reporting the previous rounds results to the experts. Resources such as the book Forecasting: Principles and practice by Hyndman and Athanasopoulos, and Sniezeks article A comparison of techniques for judgmental forecasting by groups with common information are used to support the presented data.
What is Judgmental Forecasting
In cases of the extreme complexity of the problem, its novelty, insufficiency of available information, or the impossibility of mathematical formalization of the solution process, one has to turn to competent experts. The solution of problems by experts, their commenting, the formation of quantitative estimates, and the processing of the latter by formal methods is called the method of judgmental forecasting. Hyndman and Athanasopoulos (2018) assert that this method involves a set of logical procedures aimed at obtaining from experts the data necessary for the preparation and selection of rational decisions. In the theoretical aspect, the legitimacy of using judgmental forecasting is confirmed because methodologically correctly obtained expert judgments meet two criteria for the reliability of any new knowledge: accuracy and reproducibility of the result. Therefore, judgmental forecasting is created by experts based on their experience and assessments.
Judgmental forecasting helps to formalize the procedures for collecting, summarizing, and analyzing expert opinions to transform them into the most convenient form for making an informed decision. Moreover, this method of forecasting is continuously developed and improved. Currently, judgmental forecasting, in combination with other mathematical and statistical methods, can minimize the risk of making the wrong decision.
Where is Judgmental Forecasting used Most Effectively?
According to Hyndman and Athanasopoulos (2018), judgmental forecasting is applied in certain cases. Firstly, when an object or some phenomenon cannot be described mathematically. Secondly, it is used when there is no reliable, accurate, statistical information about the object that allows the use of objective methods. Thirdly, when there are no means for mathematical processing. For example, if there is no computer equipment, software, or qualified personnel. Fourth, it is a suitable method for an emergency situation when a quick decision is needed. Thus, there are four main cases in which judgmental forecasting is used.
Examples of how Judgmental Forecasting is Used Most Effectively
One example of the most effective use of judgmental forecasting may be the case of central banks. Hyndman and Athanasopoulos (2018) note that they involve judgment in forecasting the current level of economic activity, a procedure known as nowcasting because GDP data is only available quarterly. Moreover, this type of forecasting model is especially useful in such fields as research and development. Focus groups and expert groups can provide information that no computer model can provide. For example, by asking a group of people about what they are looking for in a product, companies can gauge their direction in developing specific product features.
Advantages and Disadvantages of Judgmental Forecasting
Like any other phenomenon, judgmental forecasting has both advantages and disadvantages.
Advantages
The undoubted advantage of judgmental forecasting is that it is universal and applicable to various forecasting objects. Moreover, it is relatively simple from a methodological point of view and does not place high demands on the quality of the initial basic information. Sniezek (1990) notes that the scope of statistical forecasting methods is limited to short- and medium-term forecasts. On the contrary, judgmental forecasting will allow to make long-term forecasts.
Disadvantages
Judgmental forecasting has several disadvantages. First, Hyndman and Athanasopoulos (2018) affirm that it can be inconsistent. Unlike statistical forecasting, which can be generated using the same mathematical formulas, judgmental forecasting relies heavily on human cognition and is subject to its limitations. For example, a limited attention span can lead to important information being missed, or a misunderstanding of cause and effect relationships can lead to erroneous conclusions. Moreover, human judgment may vary due to psychological factors. Second, judgment can be clouded by personal or political agendas when goals and projections are not separated. Even when goals and projections are separated, judgment can be clouded by optimism or wishful thinking. Third, there can be anchoring effects in judgmental forecasting where subsequent forecasts tend to converge or be close to the original familiar reference point. Anchoring can lead to conservatism and underestimation of new and more relevant information and thus create a systematic bias. Therefore, whenever possible, the judgmental results obtained by the methods of expert assessments should be supported by other forecasting methods.
The Delphi Method of Judgmental Forecasting
One of the methods for obtaining an individual opinion of experts is the Delphi method. Its essential features are anonymity, multi-level, and correspondence. According to Sniezek (1990), the Delphi method is effective because it allows to consider the opinions of all people who are relevant to any issue through a consistent combination of considerations, suggestions, and conclusions, and then come to a specific agreement. The advantage of the Delphi method is the use of feedback during the survey, which significantly increases the objectivity of expert assessments. Thus, using the Delphi method, expert opinions are determined by periodically weighing their views, taking into account the answers and arguments of colleagues.
References
Hyndman, R.J., & Athanasopoulos, G. (2018). Forecasting: Principles and practice (2nd ed.). OText.
Sniezek, J.A. (1990). A comparison of techniques for judgmental forecasting by groups with common information. Group & Organization Studies, 15(1), 5-19.
The analysis of sales data of five products (1-5) offers important information regarding their structure, appropriate models of the forecast, and prediction of demand. The attributes of product 1 indicated that it has a cyclical structure without seasonality and trend. Based on these attributes, SMA and SES were suitable models of forecasting. Product 2 exhibited sales with an additive trend but without seasonality, making the Naïve and Holts methods as appropriate forecasting models. In the analysis of product 3, both seasonality and trend were evident. In this view, TSEM and regression analysis fitted in the prediction of sales. Product 4 exhibited seasonality without trend, making SEM and linear regression analysis appropriate models of forecasting them. In products 1-4, MAE and MAPE were suitable in the measurement of errors for accurate forecasting. Product 5 had an intermittent structure of sales, which suited the use of SBA and Crostons methods. The scaled MSE fitted the measurement of errors in product 5 sales data owing to the sporadic nature of trends.
Introduction
Analysis of sales data is critical in business because it generates information that informs operations and trends of consumer demand. The purpose of this report is to analyze sales data of five products (P1 to P5) and generate forecasts based on time and temperature. To highlight trends in sales data, the analysis decomposed the time series of each product. Further, the analysis employed a feasible candidate method in examining the structures of different sales data. After separating out-of-sample and in-sample datasets, the forecast was done by employing the in-sample one in predicting the trend of sales and creating a reliable model. Subsequently, error measures were selected based on key performance attributes and then employed in the evaluation of the created model. The evaluation of forecast methods and respective error metrics was done to achieve comprehensive findings. Critical examination of findings led to the generation of practical recommendations and conclusions for the company to implement.
Description and Analysis
Time series graphs were plotted for products 1 to 5, depending on frequencies of 100 observations of sales recorded daily and weekly. By examining trends and seasonality, the structure of sales of each product was established. Decomposition was performed to highlight the presence of trends, seasons, and noise attributes of sales.
Product 1
Since the sales of product 1 were collected weekly 100 times, a frequency of 13 was considered appropriate to provide a quarterly times series. The plot also shows a 13-weeks centered moving average (13-CMA) without a clear trend. Figure 1 shows that product 1 exhibits some seasonality because sales appear low in every even quarter and high in every odd quarter.
The plot of seasonality (Figure 2) shows that sales of product 1 do not have seasonal variation since the trend lines appear random. The WO-test confirms that the trend of product 1 sales does not have seasonality.
The decomposition of data (Figure3) indicates an irregular structure of product 1 sales and a considerable degree of noise.
Product 2
also shows weekly sales recorded over 100 weeks and a frequency of 13 was used for the quarterly analysis of its trend. The sales data shows an increasing trend as shown in figure 4.
The de-trended seasonality plot (Figure 5) shows that sales of product 2 do not exhibit a seasonality trend because it has a lot of noise in the data.
The multiplicative decomposition (Figure 6) depicts that the product has an increasing trend without seasonality. The observed trend seems to be stable up to the fourth quarter where it starts to exhibit significant variability and decline after the seventh quarter. The WO-test confirms that sales of product 2 do not exhibit seasonality in its multiplicative trend.
Product 3
As product 3 sales were recorded daily for 100 days, the frequency of seven was used to analyze the weekly trend of data. Figure 7 shows that sales of product 3 have both additive trend and seasonality with a weekly centered moving average (7-CMA). The de-trended plot shows that sales of product 3 follow the weekly season with a minimal level of noise (Figure 8). The additive decomposition plot shows that product 3 has an increasing trend with a clear pattern of seasonality (Figure 9). The WO-test confirms that the apparent weak seasonality is statistically significant (p<0.001).
Product 4
As sales of product 4 were measured daily and weekly trend exhibited a lot of noise, the frequency of 14 days was used (14-CMA). The plot (Figure 10) shows that product 4 sales data do not have a trend since the variation is constant along the trend line. However, the seasonality plot (Figure 11) reveals that sales of product 4 exhibit a fortnightly season with a limited degree of noise. The additive time series (Figure 12) indicates seasonality and an increasing trend in product 4 sales. The WO-test reveals that product 4 sales data has statistically significant seasonality (p<0.001).
Product 5
The examination of the plot of product 5 sales shows that it has neither season nor trend, which defines the structure of time series (Figure 13). The existence of zero implies that product 5 has an erratic demand that is hard to predict using standard decomposition methods. Based on the average demand of 0.95 and the squared coefficient of variation, the SBA method indicated that the demand interval was greater than 1.33 (Figure 14). In this view, the SBA model became an appropriate one in forecasting sales of product 5.
Candidate Forecasting Methods
Product 1
Given that the time series plot of product 1 did not have a trend or season with a lot of noise, a linear forecast method is suitable. In this view, the simple moving average (SMA) applies because it is not sensitive to a high degree of noise. Additionally, an extended period of SMA is necessary to minimize the effects of a high degree of noise in product 1. Single exponential smoothing (SES) is also another feasible method that could be used in forecasting sales of product 1 as it places more weight on recent data than old ones. To achieve accurate prediction, a small smoothing parameter (alpha value) is essential to diminish noise and set the appropriate degree of weight and length of moving average.
Product 2
The time series plot of product 2 shows that it has an additive trend, but it does not exhibit seasonal variation in sales data. The Naïve forecast method of the forecast is appropriate because it assumes an increasing trend is consistent over time. Since the Naïve forecast method does not entail smoothing, Holts exponential smoothing is essential in instances where variation and noise are high to generate a consistent moving average and weight of sales. As de-trended seasonality was determined, the use of the damped method of exponential smoothing was employed.
Product 3
The time series plot for product 3 shows that it has both additive trend and seasonality. The sensible forecast method for this kind of data is trend-seasonal exponential smoothing (TSEM). This method has alpha, gamma, beta, and parameters, which smoothen the level, seasonality, and the trend of sales data, respectively. These parameters are critical in forecasting because they mask outliers and noise in time series data. Depending on the relationship of product 3 and product 4, as well as the explanatory variables of price and temperature, the forecast method could be a linear regression model.
Product 4
Product 4 has a time series that has a constant trend without seasonality, but it has a lot of noise. SEM is an appropriate method because it has an alpha parameter that would minimize noise by smoothing trends and enhancing the accuracy of forecasts. If predictor variables of temperature and price have a significant influence on the sales of product 4, the linear regression method would be appropriate (Rosen, 2018).
Product 5
The analysis of the time series plot reveals that product 5 has no trend and seasonality because of erratic demand. In this view, the Poisson distribution is a sensible method of forecast because it requires the mean demand. Moreover, calculation of the squared coefficient of non-zero demand and average demand interval would determine if the best method would be Croston or SBA. The product 5 sales data shows that it has a high degree of the demand interval, which is suitable for the SBA method of the forecast.
In and Out-of-Sample
The sales data was partitioned into out-of-sample and in-sample datasets to allow evaluation of the accuracy of the prediction models. The out-of-sample dataset would be used to check the accuracy of the model, while the in-sample dataset would be utilized to create the model. A comparison of the errors of forecasts and the holdout data would indicate enable the evaluation of the model.
Product 1 and 2
As the sales data for products 1 and 2 comprised of weekly observations, the frequency of 13 weeks was used in setting in-sample data of 78 observations (6 quarters) and the remaining 22 as out-of-sample observations. The SEM and Holts exponential smoothing methods for forecasting products 1 and 2, respectively, do not need a higher proportion of holdout data.
Product 3 and 4
Since sales data for products 3 and 4 were collected daily for 100 days, a higher proportion of in-sample data was selected at 79 observations and an out-of-sample data of 21. These proportions are close to the standard one of 80%/20% (Quirk & Rhiney, 2017), and it allows the division of 21 days into three weeks for out-of-sample data.
Product 5
The presence of an erratic demand in the sales of product 5 does not provide a specific principle of partitioning data into in-sample and out-of-sample datasets. In this case, the standard criteria of 80% in-sample and 20% out-of-sample method were employed.
Error Measures
Products 1, 2, 3, and 4
In the measurement of errors, the mean absolute error (MAE) would be used in forecasting sales data for products 1, 2, 3, and 4 because it is not sensitive to outliers in SEM and TSEM forecasts. Mean squared error (MSE) would also be used although it is sensitive to the existence of outliers. In the assessment of products 3 and 4, mean absolute percent error (MAPE) would be used because it allows comparisons of errors between products.
Product 5
The presence of the problem of intermittent demand makes MAE and MAPE as inappropriate methods of evaluating errors. In this case, MSE is a suitable method because it is independent of the scale used in the measurement of sales data for product 5.
Analysis of Forecasting Methods
Product 1
According to Figure 15, SEM offers an accurate method of predicting sales of product 1. Using SEM, the forecast of product 1 indicated an optimal value of alpha of 0.99 gave an MAE value of 13.01 and MAPE of 8.85.
Product 2
Forecast of product 2 shows that sales increases with both an increasing trend and seasonal trend variation.
Product 3
Regression analysis indicates that price and temperature accounts for 55.48% of the variation in the sales of product 3 (Figure 17 and Table 2)
Product 4
Table 3 shows that price and temperature do not account for significant influence on the sales of product 4 (Table 3)
Table 3: Correlation analysis.
P4
Price
Temperature
P4
1
Price
-0.17837
1
Temperature
-0.06704
0.15822
1
Product 5
Time series established that Croston and SBA methods predict variation in sales of product 5 in time series (Figure 18).
Evaluation of Results and Recommendations
Product 1
The calculations of forecasts of product 1 using SES and SMA generated accurate values despite the presence of noise and outliers. As the SMA method achieved lower values for MAE than the SES method, it is a relatively better method. However, the analysis recommends both SMA and SES as a suitable forecasting method.
Product 2
Forecast of sales data for product 2 using the Naïve and Holts methods generated accurate values. Holts method provided more accurate forecasts than the Naïve method owing to the low MAE values obtained. Additionally, as Holts method generates extended periods of averages, which minimize the effects of noise and outliers. In forecasting the sales data of product 2, Holts method is fit.
Product 3
The presence of both additive seasonality and trend fits the use of TSEM in forecasting sales data for product 3. The use of the linear regression analysis is appropriate because price and temperature account for 54.48% (R2 = 0.5548) of the variation in the sales level of product 3 (Rosen, 2018).
Product 4
The application of correlation shows that price had a higher level of correlation than temperature. The multiple regression analysis indicated that both temperature and price are not statistically significant predictors of the sales of product 4 because they account for 3.3% of the variance (R2 = 0.033). The analysis recommends the use of price and other explanatory variables in predicting the sales of products 4.
Product 5
The erratic demand for product 5 sales fitted the use of both SBA and Croston forecast methods generated the same outcomes. Although SBA provides accurate findings, the report recommends the use of both methods to allow comparisons of trends, seasons, and levels in sales.
Conclusion
The examination of plots and decomposition of time series of data sales (1-5) indicated different structures. As the structure of sales data determined their forecast methods, SEM, SMA, Naïve methods, TSEM, Holts method, regression, and intermittent methods were employed to forecast sales data. SEM and SMA were utilized in the forecast of product 1, the Naïve and Holts methods in product 2, TSEM in product 3, regression model in product 4, and SBA/Croston in product 5. In error analysis, MAE and MSE were relevant to products 1-4 because they are not sensitive to outliers. In addition, MAPE was employed with the consideration of the effects of noise and outliers. Scaled type of MSE derived from MSE and the mean of squared overage of the demand.
References
Quirk, T. J., & Rhiney, E. (2017). Excel 2016 for advertising statistics: A guide to solving practical problems. Springer.
Rosen, D. (2018). Bilinear regression analysis: An introduction. Springer.