Sambalpur Cashew Works: The ₹18 Lakh Shelling Machine
Sambalpur Cashew Works is a proprietorship that processes raw cashew into kernels and employs 40 workers on manual shelling. The proprietor, Ramesh, must decide whether to install an automatic roaster–sheller costing ₹18,00,000 with a five-year life and no salvage value, funded by ₹4,00,000 of his own savings and a loan at 11% a year. His accountant has set out three demand scenarios that, once established, are expected to persist: high demand giving annual net savings of ₹9,00,000 (probability 0.3), moderate demand ₹5,00,000 (0.5) and low demand ₹1,50,000 (0.2). A second option is to outsource roasting and shelling to a job-work unit at Cuttack, which gives a certain annual saving of ₹4,00,000 with no investment at all. Ramesh's daughter, a management graduate, warns that the probabilities are the accountant's guesses and that the family's free cash is only ₹1,20,000. Ramesh prefers the machine because 'it is a matter of pride for the town'. He wants the decision within a month, before the season begins.
1. Compute the expected annual saving of each option and state which one the figures favour. Show the working and note what the expected value hides.
Read the worked answer
Expected annual saving = Σ (probability × saving).
Machine: (0.3 × 9,00,000) + (0.5 × 5,00,000) + (0.2 × 1,50,000) = 2,70,000 + 2,50,000 + 30,000 = ₹5,50,000 a year. Over five years this is ₹27,50,000 against an outlay of ₹18,00,000, an expected net gain of ₹9,50,000.
Job work: ₹4,00,000 a year with certainty, ₹20,00,000 over five years, with no investment.
On the expected-monetary-value criterion the machine is ahead by ₹1,50,000 a year.
But the average conceals the spread. In a low-demand year the machine saves only ₹1,50,000 - below the job-work option and far below the loan repayment. The expected value also treats the probabilities as objective when they are the accountant's judgement, assumes each year's demand is independent when the question says demand persists, and ignores non-financial factors like control over quality and the 'pride' motive. The figures favour the machine; whether the firm should buy it is a separate risk question.
2. Using the loan repayment burden and the cash cushion, assess whether the expected value alone should decide. Recommend a course of action with reasons.
Read the worked answer
Annual repayment on ₹14,00,000 at 11% over five years: annuity factor = (1 − 1.11⁻⁵)/0.11 = 3.696, so repayment ≈ 14,00,000 / 3.696 = ₹3,78,700 a year, about ₹31,600 a month from month one.
Against that, a low-demand year yields savings of only ₹1,50,000 - a shortfall of about ₹2,29,000 in that year alone. The family's free cash of ₹1,20,000 is less than one year's gap, and a proprietor carries unlimited liability, so a run of two weak years would force distress borrowing or asset sales.
Recommendation: do not commit the full ₹18,00,000 on these figures.
- Take the job-work route for one season to build reserves and observe real demand at low risk.
- Explore lease or hire-purchase of the machine, which lowers the fixed annual outflow and preserves the option to exit.
- Seek advance supply contracts with two or three buyers so that part of the 'high demand' becomes a certain flow before the loan is signed. Once contracts and a season's data confirm demand, the machine becomes a sound investment rather than a gamble.
3. Ramesh's daughter calls his reasoning 'boundedly rational'. Explain the term in this situation and suggest two ways the firm can lower decision risk before the season.
Read the worked answer
Bounded rationality (Simon) means a manager cannot gather or process all information, cannot assign true probabilities, and has limited time and money, so he satisfies - accepts the first option that is good enough rather than searching for the optimum. Ramesh's 'pride' justification is a satisficing shortcut: it stops the search without testing demand. The accountant's probabilities are themselves bounded judgements, so the expected value is only as good as the guesses behind it.
Ways to lower risk:
- Search and trial: run a pilot - send part of the season's volume to job work while collecting actual price, yield and demand data from buyers; this converts guessed probabilities into observed ones.
- Sensitivity and break-even test: ask 'what must demand be for the machine to repay its loan?' Annual repayment is ₹3,78,700, so the machine must save at least that much every year. Ramesh can then judge how likely that is, and negotiate a moratorium or a lower-cost machine if the margin of safety is thin. A third practical step is to keep the fixed commitment small - lease rather than buy - so that being wrong is survivable.