ITLS5250 Chap.5 Forecasting, Inventory and EOQ Logic
Forecasting, Inventory and EOQ Logic
Demand forecasting supports resource allocation across network, capacity, transport and inventory decisions. This chapter begins with level, trend, seasonal and life-cycle patterns, then compares qualitative judgement, moving averages, exponential smoothing and explanatory methods. Error is separated from directional bias, and aggregation is evaluated at the decision level.
The final section uses the raw EOQ derivation to explain why falling ordering cost and rising carrying cost produce a minimum.
What this chapter covers
- 01
Forecast purpose and horizon
- 02
Demand patterns and life cycle
- 03
Qualitative forecasting
- 04
Moving averages and smoothing
- 05
Forecast error and bias
- 06
Hierarchical aggregation
- 07
Ordering and carrying costs
- 08
EOQ assumptions
- 09
Reorder-point distinction
Worked example · free
Compute and interpret a basic EOQ
- 1Convert the carrying rate to annual cost per unit: 0.20 × $30 = $6.
- 1Substitute into Q* = √(2DS/IC): √(2 × 2,400 × 75 ÷ 6).
- 1Evaluate √60,000 = 244.95, giving a practical lot of about 245 units.
- 1Ordering cost is (2,400 ÷ 245) × $75 = $734.69 per year.
- 1Carrying cost is $6 × 245 ÷ 2 = $735.00 per year.
- 1Interpret the near equality as the basic minimum, subject to stable demand, lead time and cost assumptions.
Key terms
- Forecast Error
- The period difference between observed demand and forecast demand under a declared sign convention.
- Forecast Bias
- A persistent tendency for forecasts to sit above or below observed demand rather than vary randomly around it.
- Moving Average
- A time-series method that forecasts from the average of a chosen number of recent observations.
- Exponential Smoothing
- A time-series method that combines the previous forecast with newly observed demand using a responsiveness parameter.
- Hierarchical Forecast
- A forecast produced and reconciled across related aggregate and disaggregate levels.
- Economic Order Quantity
- The basic order quantity that minimises annual ordering plus cycle-stock carrying cost under its assumptions.
- Reorder Point
- The inventory-position trigger for placing an order, driven by lead-time demand and any protection stock.
Forecasting, Inventory and EOQ Logic FAQ
What is the difference between error and bias?
Error is the period difference between actual and forecast demand, while accuracy measures summarise its magnitude. Bias is a persistent direction in signed errors. Declare the sign convention before calling positive values over- or under-forecasting.
How should demand pattern affect method choice?
Stable level demand can suit averaging, trend needs directional responsiveness, and seasonality requires its recurring cycle to be represented. New products and life-cycle shifts often require judgement because historical observations are limited or structurally outdated.
Why can an aggregate forecast look better?
Positive and negative local errors can cancel when demand is pooled. That can improve a national capacity or stock decision, but local allocation can still fail, so evaluate forecasts at every level where an operating choice is made.
Are forecasting formulas assessed for calculation?
The weekly learning page says students will not need to memorise the forecasting formulas or perform calculations with them. They are used to explain responsiveness, trend and seasonal capability, so revision should prioritise interpretation and method selection.
What assumptions limit the EOQ result?
The basic relationship assumes stable average demand and lead time, replenishment without shortage, and ordering plus carrying as the relevant cost balance. Discounts, capacity limits, variability and stockout consequences can change the policy.
How should a human forecast adjustment be reviewed?
Record the unadjusted model forecast, the override, the reason and the later actual demand. Compare their errors over repeated decisions and across relevant product groups. This reveals whether judgement adds event knowledge or introduces optimism, and it prevents an unexplained final number from erasing what the forecasting process could learn.
Review the override by horizon as well, because useful near-term operational knowledge may not improve a longer-range capacity forecast.
Exam move
Start every forecast problem by sketching the series and labelling level, trend, seasonality, life-cycle position and known influences. Write the error sign convention before interpreting bias. Compare methods by responsiveness and pattern fit rather than by prestige, and evaluate them at the aggregation level used for capacity or inventory decisions.
Treat formulas as explanations of behaviour because the unit says forecasting calculations are not memorised assessment tasks. For EOQ, derive the direction first: ordering cost falls as Q rises, carrying cost rises, and total cost reaches a minimum. Keep EOQ, which answers how much, separate from the reorder point, which answers when. Finish quantitative work with assumptions and an operational implication.
Review short time series by first identifying structure, then marking every signed error and asking whether a model change or an external event explains the pattern. Compare bottom-up and top-down forecasts against the decisions they serve rather than declaring one universally superior.
For inventory, practise direction questions before arithmetic: lower ordering cost reduces the preferred lot, higher carrying cost also reduces it, and higher demand increases it under the basic relationship. Recompute one full example with units, verify ordering and carrying costs are close, and then list the real conditions that could invalidate the result.
Working through Forecasting, Inventory and EOQ Logic in ITLS5250? Sia is AskSia’s AI Management tutor — ask any ITLS5250 Forecasting, Inventory and EOQ Logic question and get a clear, step-by-step explanation grounded in how ITLS5250 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.