Deciding on Machinery Replacement with Relevant Costs Analysis

A ten-year-old lathe machine is still running. It hasn’t broken down, but it’s slower, guzzles more electricity, and needs a mechanic’s visit every other month. Should the company replace it with a new one? The instinct is to look at how much was originally paid for the machine and feel reluctant to “waste” that investment. But that instinct is exactly what relevant cost analysis warns against. When it comes to machinery replacement decisions, only the costs and benefits that will actually change in the future matter. Everything else is noise.

What makes a cost “relevant” to a replacement decision 🔗

In management accounting, a cost is relevant only if it is a future cost that differs between the alternatives being compared. If a cost stays the same no matter what you decide, or if it has already been spent, it has no bearing on the choice ahead of you. This is the essence of relevant costing, which asks a simple question: will this amount actually change the numbers on the bank statement depending on what you decide?

Applied to machinery replacement, this means you compare the old machine and the new machine only on the basis of what each will cost or earn from today onward. The purchase price you paid for the old machine years ago has nothing to do with today’s decision.

The sunk cost trap 🔗

The original purchase price of the existing machine, along with its book value and any remaining depreciation, is a sunk cost. It was spent in the past and cannot be recovered or changed by any decision you make now, so it plays no role in whether replacement is worthwhile. AccountingCoach explains this well with the example of a company deciding whether to buy a new machine: once the money on the old one is spent, the only question left is whether the extra spending required today will generate enough future savings to justify itself.

This trips up even experienced managers. Consider a machine that originally cost ₹3 lakh, now carries a book value of ₹50,000, and could be updated at extra cost, or sold today for ₹75,000. Neither the ₹3 lakh purchase price nor the depreciation charge on the books represents an actual future cash flow, so both are irrelevant to the replacement decision. What matters is what happens to cash flow from this point forward.

Relevant vs irrelevant costs at a glance 🔗

The factors that actually decide the outcome 🔗

Once the sunk costs are set aside, the real replacement decision comes down to a handful of factors. Each one needs to be estimated honestly, because an overly optimistic projection can make a poor replacement look attractive on paper.

Operating costs, old versus new 🔗

Compare the running costs of the two machines: power consumption, consumables, labour, maintenance contracts, and the frequency of breakdowns. An older machine often costs more to run each year even if it was cheaper to buy originally. The relevant figure is the difference in these costs going forward, not the absolute cost of either machine.

Technological advancement 🔗

Newer machinery frequently brings higher precision, lower wastage, and better safety compliance. Failing to account for technological obsolescence can mean sticking with equipment that quietly erodes competitiveness. Rapid technological change can render existing equipment or processes outdated, and recognising this early helps avoid production bottlenecks that are harder to fix later.

Return on capital employed 🔗

A replacement decision is, at its core, a capital budgeting decision. The extra capital committed to the new machine should earn a return that clears the company’s cost of capital. Standard techniques such as net present value and internal rate of return are used to test this: a replacement is generally worth pursuing when its net present value works out positive, or when its internal rate of return exceeds the cost of capital. Even a simpler payback period calculation, which shows how quickly the extra investment is recovered through savings, is useful for smaller decisions.

Demand for the output 🔗

A faster or more efficient machine is only valuable if there is enough demand to absorb the extra output, or if the cost savings alone justify the switch even at current volumes. If demand for the product is falling, spending on higher capacity may not pay off, no matter how efficient the new machine is.

Opportunity cost 🔗

Every rupee spent on new machinery could have been used elsewhere, whether to fund another project, reduce debt, or earn interest. This is the opportunity cost of capital, and it is the value of the benefit given up by choosing one course of action over the next best alternative. There is also an opportunity cost tied to the old machine itself: if it can be sold today, that sale value is money forgone by choosing to keep using it, and it belongs firmly on the relevant side of the analysis.

A simple worked example 🔗

Suppose a garment unit is deciding whether to replace an older stitching machine. The old machine can be sold today for ₹40,000. Keeping it means annual running costs of ₹1,20,000. A new machine costs ₹4,00,000 to buy and install, but cuts annual running costs to ₹70,000 because of lower power use and fewer breakdowns.

The relevant cash outflow today is ₹3,60,000 (₹4,00,000 minus the ₹40,000 recovered from selling the old machine). Against this, the unit saves ₹50,000 every year in running costs. Ignoring the time value of money for a moment, the payback period works out to just over seven years. Applying a discount rate to those annual savings and comparing the resulting present value against the ₹3,60,000 outlay would show whether the investment clears the firm’s required rate of return, which is precisely how NPV-based analysis is used in replacement decisions. Notice that the original price paid for the old machine never enters this calculation at all.

Mistakes to watch out for 🔗

Chasing the sunk cost: Letting the amount already spent on the old machine influence the decision is the single most common error. The money is gone regardless of what you decide next.

Ignoring tax and depreciation benefits: New machinery often qualifies for depreciation allowances that reduce taxable income. These are genuine future cash effects and belong in the analysis, unlike the book value of the old asset.

Treating fixed costs as if they change: Allocated overheads that stay identical under both options are irrelevant, even though they appear on the profit and loss statement. Only costs that move because of the decision should be counted, a distinction highlighted clearly in discussions of relevant versus irrelevant costs.

Overlooking qualitative factors: Employee morale, safety improvements, and downtime during installation don’t always show up as neat numbers, but they still affect the real outcome of a replacement decision.

Bringing it together 🔗

Machinery replacement decisions look complicated because they involve big numbers and long time horizons, but the underlying logic is straightforward once sunk costs are removed from the picture. What’s left is a clean comparison: future operating costs, the capital required, the return that capital can generate, demand for the output, and the opportunities forgone by choosing one path over another. Get this comparison right, and the decision practically makes itself.

What do you think? If a machine has almost no resale value left, does that change how much weight technological obsolescence should carry in the decision? And how should a business factor in uncertain future demand when the payback period on a new machine stretches beyond five or six years?

References

1. https://www.accaglobal.com/us/en/student/exam-support-resources/fundamentals-exams-study-resources/f5/technical-articles/relevant-costs.html
2. https://www.accountingcoach.com/blog/what-is-a-sunk-cost
3. https://fastercapital.com/content/Replacement-Projects–How-to-Analyze-Replacement-Projects-in-Capital-Budgeting.html
4. https://pages.stern.nyu.edu/~adamodar/pdfiles/acf3E/ch6.pdf
5. https://www.accountingverse.com/managerial-accounting/relevant-costing/relevant-and-irrelevant.html