The useful ranking is per constrained unit of capacity
When one production resource constrains sales, compare throughput earned per hour of that bottleneck rather than per finished unit alone. Under the stated short-run throughput model, sales less direct material cost is throughput and other factory costs are treated as operating expense. ACCA’s primary guidance on throughput accounting explains this model and the demand-limited allocation process. This management analysis is checked on 24 October 2026; it is not a statutory inventory valuation policy.
Establish the cost behaviour and the actual constraint
The material-only calculation depends on the short-run assumption that other costs are largely fixed. Do not ignore genuinely incremental overtime, outsourced conversion, dispatch or other avoidable costs merely because a model calls them operating expense. Show those additional cash costs separately in the decision. Distinguish a normal overhead absorption rate from a cost that changes when the plan changes.
Identify the resource whose available usable time actually limits the feasible output and sales. Use supported cycle times, yield, maintenance availability and setup requirements. If several constraints bind, minimum batches apply or product sequences consume material setup time, a simple single-resource ranking may be insufficient. A high local machine utilisation percentage is not proof that it is the system’s bottleneck.
Allocate capacity only up to real demand
For a single constraint, calculate unit throughput, divide by bottleneck hours per unit, rank products and allocate time in that order subject to each product’s saleable demand and feasibility. The next product can use the remaining time. Do not manufacture extra stock solely to show a full bottleneck when customers will not buy it; production without sales does not generate the model’s sales throughput.
| Illustrative base input | Product A | Product B |
|---|---|---|
| Selling price per unit | ₹500 | ₹450 |
| Direct material per unit | ₹200 | ₹150 |
| Bottleneck hours per unit | 0.50 | 0.25 |
| Maximum saleable units | 200 | 200 |
The linked base illustrations assume 100 usable bottleneck hours, ₹60,000 period operating expense excluding direct material, no inventory change, all planned output sold in the period, and all other capacity sufficient. Rates, demand and costs are fictional management inputs. No GST, tax, discounting, credit-risk or working-capital estimate is implied.
Use the ratio as a cost-coverage lens, not an automatic shutdown rule
Return per factory hour is throughput per unit divided by bottleneck time. Cost per factory hour is total period factory operating expense divided by available bottleneck time. The throughput accounting ratio divides the former by the latter. For products in the same model and factory, its common denominator gives the same ranking as throughput per bottleneck hour. A ratio below or at one should prompt review, but does not itself establish that discontinuing that product saves the allocated operating cost.
Reconcile total sales less direct material to total throughput and then deduct operating expense once. This management result must be reconciled to the financial accounts, especially where inventory changes or statutory cost recognition differ. Do not replace AS or Ind AS inventory requirements with the management model’s cost labels.
Elevating one machine can leave another constraint
For an improvement proposal, calculate additional feasible units sold after all constraints and demand caps, the associated throughput, and genuine incremental costs. A bottleneck upgrade can expose another machine or demand as the new limiting factor. A non-bottleneck upgrade may improve quality or reduce avoidable cost, but extra nominal capacity alone does not establish extra system throughput. Report those distinct benefits with evidence rather than converting all saved machine hours into sales.
Worked illustrations
These illustrations use fictional facts and expressly stated assumptions. Amounts are Indian rupees unless a different currency or unit is specified. They are independent examples, not findings about any actual client.
Illustrative example 1: Same unit throughput, different priority
A earns ₹500 − ₹200 = ₹300 per unit; B earns ₹450 − ₹150 = ₹300. A uses 0.50 bottleneck hours, giving ₹600 per hour. B uses 0.25 hours, giving ₹1,200 per hour. B therefore takes priority under the base model, despite identical ₹300 unit throughput.
Illustrative example 2: Demand-limited production plan
Produce B’s saleable demand of 200 units, using 50 bottleneck hours and generating ₹60,000 throughput. The remaining 50 hours produce 100 A units, generating ₹30,000. Total units are 300, time 100 hours and throughput ₹90,000; do not produce a further 200 B units after its demand cap is reached.
Illustrative example 3: Plan-to-profit bridge
The base plan sells A for ₹50,000 and B for ₹90,000: total ₹1,40,000. Direct material totals ₹20,000 for A and ₹30,000 for B, or ₹50,000. Throughput is ₹90,000; less ₹60,000 operating expense gives ₹30,000 management profit. Producing 200 A units alone would use all 100 hours and generate only ₹60,000 throughput, reducing that profit to ₹0 under unchanged expense assumptions.
Illustrative example 4: Throughput accounting ratios
Cost per bottleneck hour is ₹60,000 ÷ 100 = ₹600. A’s ratio is ₹600 ÷ ₹600 = 1.00; B’s is ₹1,200 ÷ ₹600 = 2.00. These ratios reproduce the ranking; they do not prove A’s allocated factory expense disappears if its residual-capacity production is stopped.
Illustrative example 5: Additional bottleneck hours with real demand
Assume a supported improvement adds twenty usable bottleneck hours, other capacity remains sufficient and A’s demand remains 200. After B’s 50 hours, A can use 70 hours for 140 units, forty more than the base plan. Additional throughput is ₹12,000. An additional period cost of ₹8,000 leaves a ₹4,000 improvement, raising the base ₹30,000 management profit to ₹34,000. This is a period comparison, not a full capital-project NPV.
Illustrative example 6: Non-bottleneck capacity is not automatic throughput
Assume another machine receives extra capacity, but the original bottleneck stays at 100 hours and there is no quality, cost or saleable-output benefit. Base throughput remains ₹90,000. If a genuine additional current-period operating cost of ₹5,000 is incurred, expense becomes ₹65,000 and management profit falls to ₹25,000. Do not invent additional sales from unused upstream capacity.
Ten-step verification table
| Step | Verification | Evidence to retain |
|---|---|---|
| 1 | Define the management decision and period | Plan, product population and assumptions |
| 2 | Validate short-run cost behaviour | Material and genuinely incremental-cost evidence |
| 3 | Identify actual binding resources | Capacity, cycle-time, yield and setup working |
| 4 | Support saleable demand | Customer orders and realistic demand caps |
| 5 | Compute throughput per unit | Selling price less stated material cost |
| 6 | Rank by constrained-resource return | Consistent bottleneck time units |
| 7 | Build the feasible allocation | Demand, capacity and other constraints |
| 8 | Reconcile total throughput to profit | Sales, materials and operating-expense bridge |
| 9 | Test improvements incrementally | Additional feasible sales, costs and new constraints |
| 10 | Reconcile to accounts and monitor results | Inventory-policy bridge and actual-versus-plan review |
Sixteen practitioner FAQs
1. Is throughput the same as contribution?
Not in the stated model; throughput deducts material while other costs are treated separately.
2. Can labour always be ignored in a decision?
No. Identify genuinely incremental or avoidable cost under the actual facts.
3. Why divide by bottleneck time?
It measures return from the resource limiting the feasible plan.
4. Does the product with highest unit throughput always rank first?
No. Its scarce-resource time may be much larger.
5. What limits production after ranking?
Real saleable demand and all applicable operational constraints.
6. Does unsold output create sales throughput?
No. Do not equate making inventory with earning sales.
7. Are multiple constraints always solved by one ranking?
No. More complex constraints may require a different optimisation working.
8. What is cost per factory hour here?
Period factory operating expense divided by available bottleneck hours on the stated basis.
9. Does TPAR change the ranking within one common factory model?
Its common denominator normally preserves the scarce-hour return ranking.
10. Does a ratio of one automatically mean discontinue?
No. Assess avoidable costs, capacity use and the actual decision.
11. Is operating expense deducted for each unit and again in total?
Avoid deducting the same cost twice.
12. Can this model replace statutory inventory valuation?
No. Maintain the applicable AS or Ind AS accounting and a management reconciliation.
13. Does adding upstream capacity necessarily increase sales?
Not while the same downstream constraint still binds without other benefits.
14. Can a bottleneck improvement create a new constraint?
Yes. Recheck the entire system and demand after the change.
15. Is a one-period benefit a full investment appraisal?
No. Longer-term cash flows, financing and risk may need separate analysis.
16. What makes the MIS actionable?
Supported cost behaviour, demand and capacity inputs, a feasible plan, a profit bridge and monitored actual results.
Conclusion
Use throughput per bottleneck hour to allocate genuinely scarce capacity, then enforce demand and cost behaviour. Reconcile the feasible plan to both management profit and the financial accounts before acting.