• Post:By Admin
• January 13, 2024

Inventory Management

1. Introduction

This report provides information about data analysis of the given data series. It also provides information on different forecasting methods which help an organisation to manage its inventory properly. This report uses the information which is provided in the case study to prepare demand forecasting for the future period. Make a comparison of the different methods of forecasting to understand better forecasting methods. This report also discusses the advantage and disadvantages of excess stock maintenance. Advantages of excess stock include low risk of shortage, fast order fulfilment and cost savings.

2. Data analysis 

The given actual annual demand table represents that the annual demand of a company increases every year. In 2018, the actual demand for a product was 9821. But in 2019 this demand increased to 11886. It means a total of 2065 product demand increases in 2019 compared to previous years. In 2019 demand increased by 21 per cent. In 2019, actual demand increases every month but in July and October month product demand decreases as compared to previous years. In 2020 product demand slightly increases as compared to previous years. In 2020 product demand increased only by 499. Product demand increases only by 4 per cent as compared to previous years.  

Demand dropped every month in 2020 except the month of April, May, July, August and October. In this selected month actual demand increases by 23, 50, 15 and 35 per cent respectively. In 2021 actual demand for the product increases by 14 per cent as compared to previous years. In 2021 demand increased by 1790 as compared to the previous years 2021. By analysing the given table it is concluded that demand for the product increases in the months of May, June, July and August for 2019 and 2020. But in 2021 actual demand increased every month except the month of August and September.                 

Diagram 1: Actual Demand

(Source: Author)

3. Forecasting 

The given case study uses the seasonality forecast method to forecast the actual demand for a product with help of the given data. Seasonality forecasting is a feature of time series in which predictable variation data experience repeat every financial year. Seasonable forecasting also refers to predictable fluctuation and it occurs every period. The trend forecasting method is created on the basis of the linear regression technique of time series forecasting (Lebotsa et al. 2018). Through demand forecasting technique help to determine future demand for a product by utilising a certain period of information and data. The given case study uses three years of data from 2019 to 2021 and determined the 2022 forecasting table which is provided in appendix 1 (Cai et al. 2021).  With help of the future demand forecasting organisation can change their perspective as per their current markets. It will also help an organisation to make informed pricing, market potential decisions and business growth plans. As per the current demand trend, it is calculated that in 2022 actual demand of the organisation will increase by 584 (Faraji et al. 2020) With the use of this forecasting, an organisation can reduce their inventory cost by way of ordering cost and warehousing costs. By analysing the given table it is concluded that demand for the product increases in the months of May, June, July and August for 2019 and 2020. But in 2021 actual demand increased every month except the month of August and September.             

Diagram 2: Seasonality Forecast with the trend

(Source: Author)

4. Different Methodology for the forecast 

Two accurate methods of demand forecasting are the average method and the exponential method.  

Average method

This method is to make forecasts accurately. In this method, the next twelve-month forecast is prepared by averaging the monthly data. for example, in case January month forecasting the last four years' January month data is averaged. Not only that every month same calculation is used to determine future demand for the product. The bar chart represents the demand for the future period and the line chart represent the four-year actual demand and forecast for the 2022 period. In the line chat series, 1 represented actual demand for 2018. Series 2 represented 2019 actual demand, series 3 represented actual demand for 2019 and series 5 represented forecasting demand for 2022 (Fattah et al. 2018) The average method demand forecasting method is a projection demand customer demand for services and goods. This is a continual process and the organisational manager uses previous years' historical data to calculate future years' expectations of demand for the goods and services (Debnath and Mourshed, 2018). Organisation information is used and adds this information with the economic data to determine demand growth and decline. From the line chart, it is clear that demand not increases as much from the 2021 actual demand but demand fluctuates compared to 2018, 2019 and 2020 actual demand (Yao et al. 2018) The organisation can use the prediction and order inventory to make fulfil future demand. This demand helps the organisation to set its goal and keep it in line with the organisational goal.

The main advantage of the short-term average method is that it allows traders and analysts to focus on the demands of action instead of reacting to sudden changes in the market. However, it is slower to respond to demand movements that occur at market reversal points. This is why it is often favoured by traders and analysts who are operating on longer time frames (Debnath and Mourshed, 2018). Also One of the advantages of using an average method is that it can provide a more accurate view of the overall trend. It can also help identify areas of resistance and support. Another advantage of using an average method is that it can separate out the random variations in the market. In this approach, establishing a historical database is a fundamental requirement for the model since a forecast cannot be made until there are sufficient data points on demand. Once the historical era has passed, every future projection is simply defined as the most recent one produced based on previous demand. This indicates that this model's projection of future demand will be flat, which means that one of its main limitations will be its inability to extrapolate any trend (Debnath and Mourshed, 2018). The average model, the most fundamental demand forecasting technique, is predicated on the notion that anticipated future demand would be comparable to actual observed recent demand. It is simply assumed in this model that the prediction is the average of the demand over the previous n periods.

Diagram 4: Average Method

(Source: Author)

Diagram 4: Average Method

(Source: Author)

Exponential Method

It is a time series forecasting technique  and it is used for univariate data is exponential smoothing. Time fore casting series method is also called Box Jenkins ARIMA family method. And it is used for the recent weighted liners of sum recent or past lags and data. Although the model apparently utilise as prior observation weight can be decrease by the exponential method.  Weighted sum of past observation are similarly lined with the exponential smoothing approach. Past observations are specifically weighted using a geometrically diminishing ratio (Sinaga and Irawati, 2020). weighted average of earlier year past data is used in exponential forcasting technique, with the weights degrading exponentially with time. Aditionally it can be describe that, when the observation are most recent then it related weight is larger.

In the given case exponential smoothing is used to accurately forecast demand for 2022 by taking the total demand of the last four years' actual demand. Exponential smoothing is simple to learn and put into practice.

in the given case the forecast forecast for the most recent time period and the actual data for that time period, and the smoothing fixed are the only components required for this approach. organizational demand plan is perfect the exponential smoothing approach through trial and error and apply it to future estimates.

The exponential smoothing approach gives the most weight to the most recent data (Rendon-Sanchez and Menezes, 2019).

This approach takes longer time horizon to considered during making the projections. it is because this approach provides more eight to recent data to determine current consumer data into account.   

 It is because this approach considered various factors have changed over the course of those three years. For example, January 2018 would not be weighted as highly as January 2021 (Rendon-Sanchez and Menezes, 2019). This is because the approach prefers to predict a forecast that is similar to that of the preceding year. Not only that, this expound that data spikes don't affect forecasts as much as they do  the averages. Forecasts are précised using exponential smoothing. This strategy forecasts that anticipate the following period will generates trustworthy and accurately. In the forecast it is shown demand projections and actual demand. This makes it possible to efficiently prepare for demand. This leads to an appropriate inventory level (Rendon-Sanchez and Menezes, 2019). But there is a certain drawback of the exceptional forecasting methods This method's "smoothing" step covers both low and high variants. It's vital because while the prediction graph represent a successive line of data. The data spikes may not constantly be reflected. Trends are not handled effectively by exponential smoothing. This approach works well for anticipating events that will happen soon. But it is less successful at long-term forecasting because it constantly anticipates net year patterns that substantially resemble existing ones (Sinaga and Irawati, 2020).

Diagram 5: Exponential Smoothing

(Source: Author)

5. Comparison of different methodologies 

Exponential smoothing is a technique that may be used to smooth time series data. It may also be viewed as a rough manner of accomplishing something, or as a thumb rule methodology. In order to assign weights that decrease exponentially over time using this strategy, we employ an exponential window function. Many decisions that are based on presumptions may be made using this way of smoothing time series. However, organisations often use it to smooth out time series data. This filter operates by averaging a predetermined portion of a set of numerical values. Consider a time series that can be divided into months. In this case, we have a set of values and the first moving average of the smoothing process may be thought of as the average of the values from the first month of the time series. The collection of numbers will vary as the series progresses in time, and as a result, the average value will also change with time.

One of the key difference between the above  two approaches may be guessd from their names.

In order to minimise the data's noise. on thesecond side the average window is applied. But the exponential smoothing method is applies an exponential window on the first.

It has techniques that give protection against time series elements like trend and seasonality under exponential smoothing. Average method refers to that the methods employed for this smoothing process place greater emphasis on the values and timings of the data. Similar to how the exponential moving average emphasises recent values.

The fact that the average method must be started with a window size greater than three in order to do genuine smoothing, whereas exponential smoothing may be done with as little as two data points, further distinguishes these two approaches. Here, it can be stated that, in accordance with the smoothing factor, the exponential eliminates more noise from the data, whereas the moving average does so by utilising the mean value of the data points. But there is a certain drawback of the exceptional forecasting methods This method's "smoothing" step covers both high and low variants. It's vital to remember that while the prediction graph displays a continuous line of data, data spikes may not always be reflected.

The simple exponential smoothing (SES) forecast is marginally better than the simple average forecast for a given average demand because it gives relatively more weight to the most recent observation, making it marginally more "responsive" to changes that have occurred recently.

6. Advantages and disadvantages of too much stock 

6.1 Advantage

There are a few benefits of excessive stock which are as follows

Low risk of shortage

An organisation can reduce their machinery and worker ideal time by purchasing the bulk of products. An organisation do not need to order inventory when the inventory level will reduce. Every organisation maintain a certain level of stock to mitigate production shortage but every time organisation have to reorder their stock if the stock level reduces to a dangerous level (Beheshti et al. 2020). If the stock department fails to reorder their stock in a reasonable time then production will be stopped. But organisations can mitigate these problems by ordering bulk stock. When demand for a product will increase then the organisation can maintain demand and supply chain by purchasing excessive stock.

Fast order fulfil

The customer easily gets their required product when a product is available in the market in huge amounts. Also, consumers have to incur lower costs to purchase this product because the product supply is higher rather than demand (Jilcott Pitts et al. 2020).

Order saving   

An organisation can reduce their inventory ordering cost by purchasing a bulk inventory order. For example, an organisation reorder inventory 4 times annually. Therefore organisations have to incur transportation costs every time purchasing inventory. In the market inventory cost also fluctuates every time. If the inventory cost increases then the organisation have to incur extra cost to purchase inventory (Yahaya et al. 2018). If an organisation purchase inventory in bulk when the price of inventory is lower then the organisation have to incur one-time transportation cost and also get inventory at a lower price. Not only that, when organisation purchase inventory in bulk they get special discounts from suppliers which help to get earn more profit (Jilcott Pitts et al. 2020).

6.2 Disadvantage

Storage Cost

If an organisation purchase excessive stock then the organisation have to incur extract cost to maintain its inventory. The organisation have to manage extra space to manage their inventory. It means organisations have to add this extra cost to the product price. For this reason, the product price of the organisation will increase and the sale of products will decrease because customers will get a similar product from their competitors at a lower price. Additionally, the maintenance risk of excessive inventory will be high. Moreover, the organisation have to incur additional insurance and interest cost (Jilcott Pitts et al. 2020).

Obsolete inventory

Organisations may face various quality problems such as potential obsolescence and degradation. Organisations may be anticipated a high level of demand from long-standing customers for these reasons organisations maintain an excessive level of stock. But if consumers change their specifications for different materials then organisation stock me be obsolescence. It is because organisations have to purchase inventory as per their customer specification  (Yahaya et al. 2018).  

Higher carrying and opportunity cost

If a company have excessive inventory company can use inventory for research and development which helps the organisation to bring additional business but not hold an additional level of inventory.  

7. Conclusion   

 This report analyses four years of actual demand data which is provided in the case study. Data analysis represented that the given actual annual demand table represents the annual demand of a company increasing every year. The given case study uses the seasonality forecast method to forecast the actual demand for a product with help of the given data. With help of the future demand forecasting organisation can change their perspective as per their current markets. It will also help an organisation to make informed pricing, market potential decisions and business growth plans. The simple exponential smoothing (SES) forecast is marginally better than the simple average forecast for a given average demand because it gives relatively more weight to the most recent observation, making it marginally more "responsive" to changes that have occurred recently. Also, there are a few benefits and drawbacks of excessive stock which are discussed in this report.

Reference

Beheshti, H.M., Clelland, I.J. and Harrington, K.V., 2020. Competitive advantage with the vendor-managed inventory. Journal of Promotion Management, 26(6), pp.836-854. https://www.tandfonline.com/doi/pdf/10.1080/10496491.2020.1794507

Cai, W., Song, Y. and Wei, Z., 2021. Multimodal data guided spatial feature fusion and grouping strategy for E-commerce commodity demand forecasting. Mobile Information Systems, 2021. https://www.hindawi.com/journals/misy/2021/5568208/

Debnath, K.B. and Mourshed, M., 2018. Forecasting methods in energy planning models. Renewable and Sustainable Energy Reviews, 88, pp.297-325. https://orca.cardiff.ac.uk/109350/1/Manuscript.pdf

Faraji, J., Hashemi-Dezaki, H. and Ketabi, A., 2020. Multi-year load growth-based optimal planning of grid-connected microgrid considering long-term load demand forecasting: A case study of Tehran, Iran. Sustainable Energy Technologies and Assessments, 42, p.100827. https://www.researchgate.net/profile/Jamal-Faraji/publication/344308075_Multi-year_load_growth-based_optimal_planning_of_grid-connected_microgrid_considering_long_term_load_demand_forecasting_A_case_study_of_Tehran_Iran/links/5f65fcd7299bf1b53ee11ef4/Multi-year-load-growth-based-optimal-planning-of-grid-connected-microgrid-considering-long-term-load-demand-forecasting-A-case-study-of-Tehran-Iran.pdf

Fattah, J., Ezzine, L., Aman, Z., El Moussami, H. and Lachhab, A., 2018. Forecasting of demand using the ARIMA model. International Journal of Engineering Business Management, 10, p.1847979018808673. https://journals.sagepub.com/doi/pdf/10.1177/1847979018808673

Jilcott Pitts, S.B., Ng, S.W., Blitstein, J.L., Gustafson, A., Kelley, C.J., Pandya, S. and Weismiller, H., 2020. Perceived advantages and disadvantages of online grocery shopping among Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) participants in Eastern North Carolina. Current developments in nutrition, 4(5), p.nzaa076. https://academic.oup.com/cdn/article/4/5/nzaa076/5820813

Lebotsa, M.E., Sigauke, C., Bere, A., Fildes, R. and Boylan, J.E., 2018. Short-term electricity demand forecasting using partially linear additive quantile regression with an application to the unit commitment problem. Applied Energy, 222, pp.104-118. https://eprints.lancs.ac.uk/id/eprint/124464/4/Applied_Energy_sigauke_APEN_D_17_10255R2_edit.pdf

Rendon-Sanchez, J.F. and de Menezes, L.M., 2019. Structural combination of seasonal exponential smoothing forecasts applied to load forecasting. European Journal of Operational Research, 275(3), pp.916-924. https://openaccess.city.ac.uk/id/eprint/21136/1/

Sinaga, H. and Irawati, N., 2020, February. A medical disposable supply-demand forecasting by moving average and exponential smoothing method. In Proceedings of the 2nd Workshop on Multidisciplinary and Applications (WMA) 2018, 24-25 January 2018, Padang, Indonesia. https://eudl.eu/pdf/10.4108/eai.24-1-2018.2292378

Yahaya, W.M.A.W., Dandan, M.A., Samion, S. and Musa, M.N., 2018. A comprehensive review on palm oil and the challenges using vegetable oil as lubricant base-stock. Journal of Advanced Research in Fluid Mechanics and Thermal Sciences, 52(2), pp.182-197. https://www.akademiabaru.com/submit/index.php/arfmts/article/download/2389/1350

Yao, H., Wu, F., Ke, J., Tang, X., Jia, Y., Lu, S., Gong, P., Ye, J. and Li, Z., 2018, April. Deep multi-view spatial-temporal network for taxi demand prediction. In Proceedings of the AAAI conference on artificial intelligence (Vol. 32, No. 1). https://ojs.aaai.org/index.php/AAAI/article/view/11836/11695

Appendix

Appendix 1: Demand forecasting