Practice cases: Singapore and Southeast Asia
Six full cases set in Singapore and Southeast Asia: wealth-management fees, a new airline route, buy-now-pay-later in a super-app, moving production to Vietnam, certifying palm oil smallholders in Malaysia, and a weekly mobile data pass in the Philippines.
Key takeaways
- Worked case: Kallang Private Bank: wealth revenue fell while assets grew.
- Worked case: Straits Wings: Singapore to Bangkok.
- Worked case: Nusa super-app: should it launch buy-now-pay-later?
- Worked case: Move production from Malaysia to Vietnam?
- Worked case: Kinabatu Mill: should it pay to certify its smallholders?
Six full cases set in Singapore and Southeast Asia: wealth-management fees, a new airline route, buy-now-pay-later in a super-app, moving production to Vietnam, certifying palm oil smallholders in Malaysia, and a weekly mobile data pass in the Philippines.
Cover the solution and run each case out loud, ideally with a partner playing the interviewer. Ask your own clarifying questions, state a hypothesis, build a structure, and do the math on paper before you look. Then compare your synthesis with the one given, and read the strong and weak candidate notes. Each case is labeled Starter, Standard, or Stretch.
Case 1: Kallang Private Bank: wealth revenue fell while assets grew
Worked case
Standard: Kallang Private Bank: wealth revenue fell while assets grew
The prompt
Kallang Private Bank's wealth-management unit in Singapore saw profit fall from SGD 160 million to SGD 110 million even though client assets grew. The exhibit shows the main figures. What happened, and what should the bank do?
Difficulty: Standard. Format: interviewer-led, with an exhibit. Industry: Banking. Region: Singapore. Interview length: about 30 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- How is revenue earned?Answer: Fees charged as a percentage of client assets.
- Did clients leave?Answer: No, assets grew 5 percent.
- Did anything change in what clients hold?Answer: Yes, many moved money from managed portfolios into low-cost funds.
A hypothesis to say out loud: Assets grew but revenue fell, so my hypothesis is that the average fee rate fell because clients moved into cheaper products.
The structure
- Revenue = client assets x average fee rate
- Client assets
- Key: Average fee rate
- Costs
The exhibit
| Measure | Last year | This year |
|---|---|---|
| Client assets (SGD billion) | 40 | 42 |
| Average fee rate (%) | 0.9 | 0.75 |
| Costs (SGD million) | 200 | 205 |
Working it through
1. Revenue last year
SGD 40,000 million of assets at 0.9 percent.
Revenue last year (SGD million):40,000 × 0.009 = 3602. Revenue this year
SGD 42,000 million at 0.75 percent.
Revenue this year (SGD million):42,000 × 0.0075 = 3153. Profit this year
Minus costs of SGD 205 million.
Profit this year (SGD million):42,000 × 0.0075 - 205 = 1104. Asset growth effect
Extra assets at last year's fee rate.
Asset effect (SGD million):(42,000 - 40,000) × 0.009 = 185. Fee rate effect
This year's assets times the change in fee rate.
Fee rate effect (SGD million):42,000 × (0.0075 - 0.009) = -636. Curveball: where did the fee rate go?
Interviewer: "About SGD 7 billion moved from managed portfolios charging 1.2 percent into low-cost funds charging 0.3 percent." Revenue lost from that move:
Revenue lost from the shift (SGD million):7,000 × (0.012 - 0.003) = 63
What the exhibit shows
Assets grew, but the average fee rate fell by 0.15 points, which outweighed the growth.
The recommendation
Profit fell because clients moved into cheaper products, not because they left. First, the lower fee rate cost SGD 63 million, far more than the SGD 18 million gained from asset growth. Second, the whole fee-rate drop is explained by about SGD 7 billion moving from managed portfolios into low-cost funds. Third, costs rose slightly, making the fall worse. Kallang Private Bank should offer advice worth paying for, such as advisory packages for larger clients, serve smaller clients through low-cost digital advice to protect margins, and review whether relationship managers are paid in a way that encourages the shift.
Risks: Clients may keep choosing low-cost funds whatever the bank offers; Rivals may compete hard on fees.
Next steps: Segment clients by size and product use; Design an advisory package and test it with 200 clients.
A strong candidate
Split revenue into assets and fee rate, sized both effects, and used the curveball to explain the fee rate drop.
A weak candidate
Said "assets grew, so the business is healthy" and blamed costs, which barely moved.
Case 2: Straits Wings: Singapore to Bangkok
Worked case
Starter: Straits Wings: Singapore to Bangkok
The prompt
First, estimate how many passengers fly between Singapore and Bangkok each year. Then: Straits Wings, a low-cost airline, wants to add two daily round trips (four flights a day). Is it worth it?
Difficulty: Starter. Format: market-sizing opener, then a business question. Industry: Airlines and travel. Region: Singapore and Thailand. Interview length: about 25 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- Both directions, and only direct flights?Answer: Yes, both directions, direct flights only.
- How many flights and how big?Answer: About 25 flights a day each way, around 180 seats each, about 85 percent full.
- For the business question, what are the fare, costs, and expected load?Answer: Average fare SGD 120 one way; each flight costs about SGD 16,000 to operate; we expect about 80 percent of seats filled.
A hypothesis to say out loud: This is a busy route, so the market is a few million passengers a year. For a new entrant, my hypothesis is that the extra flights work only if they fill well above three quarters of seats.
The structure
- Size the route, then test one flight's economics
- Flights x seats x load factor x days
- Key: Revenue and profit per flight
- Break-even load factor
Working it through
1. Passengers a year
25 flights each way, 180 seats, 85 percent full, 365 days.
Passengers a year:25 × 2 × 180 × 0.85 × 365 = 2,792,2502. Revenue per flight
180 seats, 80 percent full, SGD 120 each.
Revenue per flight (SGD):180 × 0.8 × 120 = 17,2803. Profit per flight
Minus SGD 16,000 of cost.
Profit per flight (SGD):180 × 0.8 × 120 - 16,000 = 1,2804. Yearly profit
Four flights a day for a year.
Yearly profit (SGD):4 × 365 × (180 × 0.8 × 120 - 16,000) = 1,868,8005. Break-even load factor
Cost divided by revenue from a full flight.
Break-even load factor (%):16,000 ÷ (180 × 120) × 100 = 74.076. Curveball: rivals cut fares
Interviewer: "The largest rival responds by cutting its fare to SGD 100, and we must match." Profit per flight at 80 percent full:
Profit per flight at SGD 100 (SGD):180 × 0.8 × 100 - 16,000 = -1,6007. New break-even
At SGD 100, the flight must be this full to break even.
Break-even load factor at SGD 100 (%):16,000 ÷ (180 × 100) × 100 = 88.89
The recommendation
The route carries about 2.8 million passengers a year, and the new flights are only narrowly worth adding. First, at SGD 120 and 80 percent full, four daily flights earn about SGD 1.9 million a year. Second, break-even is about 74 percent full, so there is only a small cushion. Third, if rivals cut fares to SGD 100, each flight loses about SGD 1,600 and would need to be about 89 percent full. Start with one daily round trip at times rivals do not serve well, and add the second only if load factors stay above 80 percent at current fares.
Risks: A fare war on a busy route; Airport slot costs at popular times.
Next steps: Check which departure times are least served; Sell the first month of seats before committing the second round trip.
A strong candidate
Sized the route cleanly, then found the break-even load factor and showed how a fare cut changes it.
A weak candidate
Said the route is big so there is room for more flights, without checking whether each flight makes money.
Case 3: Nusa super-app: should it launch buy-now-pay-later?
Worked case
Standard: Nusa super-app: should it launch buy-now-pay-later?
The prompt
Nusa, a ride-hailing and payments app in Indonesia that reports in US dollars, is considering "buy now, pay later": shoppers pay in four monthly parts with no interest, and merchants pay a fee. Should Nusa launch it?
Difficulty: Standard. Format: candidate-led, with interviewer dialogue. Industry: Technology and financial services. Region: Indonesia. Interview length: about 40 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- How many users and how much would they spend?Answer: About 2 million users, each making 4 purchases a year of about USD 60.
- How does the product earn money?Answer: Merchants pay a 4 percent fee on each purchase. Funding costs are 1.5 percent of purchase value, credit losses 2.5 percent, and operations 0.5 percent.
- Are there rules to consider?Answer: Yes. Indonesia's financial regulator (OJK) has buy-now-pay-later rules from 2026, including a minimum age of 18, a minimum income of about IDR 3 million a month, and a cap on repayments relative to income. Assume about 2 million users qualify and a licensed partner does the lending.
A hypothesis to say out loud: Buy-now-pay-later margins are thin. My hypothesis is that it loses money at typical loss rates, and only works if the app's own data can cut credit losses.
The structure
- Profit = purchase value x (merchant fee - funding - losses - operations)
- Purchase value (users x purchases x size)
- Merchant fee income
- Key: Costs: funding, credit losses, operations
- Wider benefit to the app
Working it through
1. Purchase value
Candidate: "2 million users x 4 purchases x USD 60."
Purchase value (USD a year):2,000,000 × 60 × 4 = 480,000,0002. Fee income
The 4 percent merchant fee.
Fee income (USD a year):480,000,000 × 0.04 = 19,200,0003. Costs
Funding 1.5 percent, losses 2.5 percent, operations 0.5 percent.
Costs (USD a year):480,000,000 × (0.015 + 0.025 + 0.005) = 21,600,0004. Profit
Candidate: "So it loses money."
Profit (USD a year):480,000,000 × (0.04 - 0.015 - 0.025 - 0.005) = -2,400,0005. Break-even loss rate
Candidate: "Losses must stay below the fee minus funding and operations."
Break-even loss rate (%):(0.04 - 0.015 - 0.005) × 100 = 26. Curveball: using the app's data
Interviewer: "Our ride and payment history could cut losses to 1.5 percent." Candidate: "Then profit becomes:"
Profit at 1.5 percent losses (USD a year):480,000,000 × (0.04 - 0.015 - 0.015 - 0.005) = 2,400,0007. Wider benefit
Interviewer: "Users of the product also use our other services more, adding about USD 3 of contribution per user a year."
Extra contribution (USD a year):2,000,000 × 3 = 6,000,0008. Break-even with the wider benefit
Candidate: "Counting that benefit, losses could reach:"
Break-even loss rate with cross-selling (%):(480,000,000 × (0.04 - 0.015 - 0.005) + 2,000,000 × 3) ÷ 480,000,000 × 100 = 3.25
The recommendation
I recommend that Nusa launch, but only with underwriting based on its own data and a strict loss limit. First, at typical 2.5 percent losses the product loses about USD 2.4 million a year, and it breaks even only at 2 percent. Second, if Nusa's ride and payment history cuts losses to 1.5 percent, it earns about USD 2.4 million, plus about USD 6 million from users spending more across the app. Third, the main risk is the loss rate, so set the stop-loss at the standalone 2 percent, not the 3.25 percent that counts wider benefits. Start with users who have 12 months of history.
Risks: Loss rates may rise in a downturn; OJK eligibility rules may leave fewer qualifying users than assumed, and rules may tighten further; Merchants may resist a 4 percent fee.
Next steps: Test a loss model on past user data; Launch with 50,000 users and small limits.
A strong candidate
Built profit as a margin on purchase value, found the 2 percent break-even loss rate, and made the launch depend on it.
A weak candidate
Focused on how popular the product is with young shoppers and never checked the loss rate.
Case 4: Move production from Malaysia to Vietnam?
Worked case
Standard: Move production from Malaysia to Vietnam?
The prompt
A Malaysian electronics contract manufacturer is considering moving production to Vietnam. The exhibit compares cost per unit. Should it move, and how much?
Difficulty: Standard. Format: interviewer-led, with an exhibit. Industry: Industrials and supply chain. Region: Malaysia and Vietnam. Interview length: about 30 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- How many units and at what price?Answer: About 5 million units a year, sold at about USD 35 each.
- What does a move cost?Answer: About USD 12 million one time to move everything, or USD 6 million to move 40 percent.
- What do customers want?Answer: The main customer, a US electronics brand, wants supply from more than one country.
- Do US tariffs differ by country of origin?Answer: They can, and they change often. For this case assume equal rates for goods from Malaysia and Vietnam (illustrative), but flag tariffs as a risk.
A hypothesis to say out loud: Labor is cheaper in Vietnam, but materials and logistics may cost more. My hypothesis is that the saving is small and shrinking, so moving part of the volume beats moving all of it.
The structure
- Saving per unit x units versus the cost of moving
- Cost per unit in each country
- Yearly saving and payback
- Key: How the saving changes over time
- Customer value of two-country supply
The exhibit
| Cost item | Malaysia | Vietnam |
|---|---|---|
| Labor | 4 | 2.5 |
| Materials | 20 | 20.5 |
| Logistics | 1 | 1.8 |
| Overhead | 3 | 2.7 |
Working it through
1. Cost per unit in Malaysia
Labor, materials, logistics, and overhead.
Malaysia (USD per unit):4 + 20 + 1 + 3 = 282. Cost per unit in Vietnam
Cheaper labor and overhead, dearer materials and logistics.
Vietnam (USD per unit):2.5 + 20.5 + 1.8 + 2.7 = 27.53. Yearly saving
5 million units at USD 0.50 each.
Saving (USD a year):5,000,000 × (28 - 27.5) = 2,500,0004. Payback of a full move
USD 12 million one time.
Payback (years):12,000,000 ÷ (5,000,000 × (28 - 27.5)) = 4.85. Curveball: wages grow at different speeds
Interviewer: "Wages are rising about 8 percent a year in Vietnam and 3 percent in Malaysia." Saving per unit after three years:
Saving per unit in year 3 (USD):0.5 - ((4 - 2.5) - (4 × 1.03 × 1.03 × 1.03 - 2.5 × 1.08 × 1.08 × 1.08)) = 0.22166. Partial move with a customer premium
Interviewer: "The customer would pay 1 percent more on units made in a second country." Moving 40 percent of volume for USD 6 million, payback on the saving plus the premium is:
Payback of a 40 percent move (years):6,000,000 ÷ (5,000,000 × 0.4 × 0.5 + 5,000,000 × 0.4 × 35 × 0.01) = 3.53
What the exhibit shows
Vietnam saves USD 1.50 on labor but gives most of it back on materials and logistics, leaving a saving of USD 0.50 per unit.
The recommendation
Move about 40 percent of volume, not all of it. First, the full move saves only USD 0.50 per unit, a 4.8-year payback. Second, faster wage growth in Vietnam cuts that saving to about USD 0.22 per unit within three years, weakening the case for a full move. Third, a 40 percent move earns the saving plus a 1 percent premium for two-country supply, paying back in about 3.5 years while keeping the Malaysian plant and its skills. Revisit a larger move only if Vietnamese suppliers lower material costs.
Risks: Start-up quality problems in a new plant; The customer premium may not last; US tariff rates by country of origin can change and could favor either country.
Next steps: Agree the premium in the next customer contract; Find local suppliers in Vietnam to reduce material and logistics costs.
A strong candidate
Compared full cost per unit, not just labor, tested wage growth, and used the customer's need for two-country supply to shape a partial move.
A weak candidate
Saw that labor is 40 percent cheaper in Vietnam and recommended moving everything.
Case 5: Kinabatu Mill: should it pay to certify its smallholders?
Worked case
Standard: Kinabatu Mill: should it pay to certify its smallholders?
The prompt
Kinabatu, a palm oil mill in Malaysia, buys fruit from its own estates and from independent smallholders. Its two largest buyers pay a premium for certified sustainable oil. Should Kinabatu pay for its smallholders to become certified?
Difficulty: Standard. Format: candidate-led, with interviewer dialogue. Industry: Agriculture and commodities. Region: Malaysia. Interview length: about 30 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- How much fruit does the mill process, and how much oil does it get?Answer: 300,000 tonnes of fresh fruit bunches a year, with an oil extraction rate of 20 percent (illustrative).
- Where does the fruit come from?Answer: 60 percent from the group's own estates, which are already certified, and 40 percent from about 2,000 independent smallholders, who are not.
- What do buyers pay for certified oil, and what does certification cost?Answer: A premium of MYR 150 per tonne of crude palm oil. Certifying a smallholder costs about MYR 1,500 once, for training and records, and MYR 300 a year for audits (illustrative).
A hypothesis to say out loud: The premium is earned on every tonne of oil, while the cost is per farmer, so my hypothesis is that certifying the smallholders pays back quickly if they keep selling to the mill.
The structure
- Value = premium on newly certified oil - certification cost
- Key: Oil from smallholder fruit
- Premium per tonne of certified oil
- Certification cost: one time and yearly
- Risks: buyer rules and side-selling by smallholders
Working it through
1. Total crude palm oil
Candidate: "300,000 tonnes of fruit at a 20 percent extraction rate."
Crude palm oil (tonnes a year):300,000 × 0.2 = 60,0002. Oil from smallholders
Candidate: "40 percent of the fruit comes from smallholders."
Smallholder oil (tonnes a year):300,000 × 0.4 × 0.2 = 24,0003. Premium on smallholder oil
Candidate: "MYR 150 on every tonne."
Yearly premium (MYR):300,000 × 0.4 × 0.2 × 150 = 3,600,0004. One-time cost
Candidate: "2,000 smallholders at MYR 1,500 each."
One-time certification cost (MYR):2,000 × 1,500 = 3,000,0005. Yearly cost
Candidate: "Audits at MYR 300 per smallholder."
Yearly audit cost (MYR):2,000 × 300 = 600,0006. Payback
Candidate: "The one-time cost divided by the yearly premium after audits."
Payback (years):2,000 × 1,500 ÷ (300,000 × 0.4 × 0.2 × 150 - 2,000 × 300) = 17. Curveball: buyers want all of it certified
Interviewer: "The buyers say that from next year they will pay the premium only if all the mill's oil is certified." Candidate: "Then the premium on the estate oil is at stake too:"
Premium on estate oil at risk (MYR a year):300,000 × 0.6 × 0.2 × 150 = 5,400,0008. If some smallholders sell elsewhere
Interviewer: "What if a quarter of the smallholders' fruit goes to other mills once they are certified?" Candidate: "Yearly premium on the rest, minus audits for all 2,000:"
Yearly net gain with side-selling (MYR):300,000 × 0.4 × 0.75 × 0.2 × 150 - 2,000 × 300 = 2,100,000
The recommendation
Kinabatu should pay to certify its smallholders, and tie the support to supply agreements. First, the smallholders' 24,000 tonnes of oil would earn about MYR 3.6 million a year of premium, against MYR 3 million once and MYR 0.6 million a year of cost, so it pays back in about one year. Second, if buyers pay the premium only on fully certified supply, certifying smallholders also protects about MYR 5.4 million a year on the estate oil. Third, even if a quarter of the fruit is sold to other mills, the net gain is still about MYR 2.1 million a year. Offer certification with a multi-year supply agreement and share part of the premium with smallholders, so they have a reason to stay.
Risks: Smallholders may fail audits in the first year; Buyers may cut the premium as more certified oil reaches the market.
Next steps: Pilot with 200 smallholders near the mill; Agree the premium and its duration with the two buyers in writing.
A strong candidate
Found the oil volume first, compared the premium with one-time and yearly costs, and saw that the curveball makes certification a way to protect the whole premium.
A weak candidate
Compared the MYR 3 million cost with nothing and called certification too expensive.
Case 6: Isla Mobile: should it launch a weekly data pass?
Worked case
Stretch: Isla Mobile: should it launch a weekly data pass?
The prompt
Isla Mobile, a mobile operator in the Philippines, sells prepaid data mostly as daily passes. It is considering a weekly pass at PHP 99. Using the table below, would the weekly pass raise data revenue, and at what price?
Difficulty: Stretch. Format: interviewer-led, with an exhibit. Industry: Telecom. Region: Philippines. Interview length: about 35 minutes. The company is fictional and all figures are illustrative.
Clarifying questions, with the interviewer's answers
- How do prepaid users buy data today?Answer: With a daily pass of PHP 20 for 1 GB. The exhibit shows the segments (illustrative).
- What is the new product?Answer: A weekly pass of PHP 99 with 7 GB.
- Who would buy it?Answer: Market research suggests every heavy daily buyer would switch, 25 percent of light daily buyers would switch, and 2 percent of prepaid users who buy no pass today would start (illustrative).
A hypothesis to say out loud: A weekly pass gives heavy users a discount and asks light users to spend more. My hypothesis is that it pays only if enough light users and new users take it to make up for heavy users paying less.
The structure
- Change in weekly revenue = gains from upgraders and new buyers - losses from heavy users
- Key: Heavy daily buyers who switch and pay less
- Light daily buyers who switch and pay more
- New buyers among prepaid users with no pass
- Price: test a higher weekly price
The exhibit
| Segment | Users (million) | Daily passes a week | Weekly spend (PHP) |
|---|---|---|---|
| Light daily-pass buyers | 4 | 3 | 60 |
| Heavy daily-pass buyers | 2 | 6 | 120 |
| Prepaid users with no pass | 14 | 0 | 0 |
Working it through
1. Pass revenue today
Light buyers: 3 passes a week at PHP 20. Heavy buyers: 6 passes a week.
Weekly pass revenue today (PHP):4,000,000 × 3 × 20 + 2,000,000 × 6 × 20 = 480,000,0002. Heavy buyers switch
Each heavy buyer now pays PHP 99 instead of PHP 120 a week.
Change from heavy buyers (PHP a week):2,000,000 × (99 - 6 × 20) = -42,000,0003. Light buyers switch
A quarter of light buyers pay PHP 99 instead of PHP 60 a week.
Change from light buyers (PHP a week):4,000,000 × 0.25 × (99 - 3 × 20) = 39,000,0004. New buyers
2 percent of the users who buy no pass today start buying the weekly pass.
Revenue from new buyers (PHP a week):14,000,000 × 0.02 × 99 = 27,720,0005. Net change at PHP 99
The three effects together.
Net change in revenue at PHP 99 (PHP a week):2,000,000 × (99 - 6 × 20) + 4,000,000 × 0.25 × (99 - 3 × 20) + 14,000,000 × 0.02 × 99 = 24,720,0006. Curveball: price it at PHP 119
Interviewer: "Research says that at PHP 119 heavy buyers still all switch, but only 15 percent of light buyers and 1.5 percent of non-buyers take it." Net weekly change at PHP 119:
Net change in revenue at PHP 119 (PHP a week):2,000,000 × (119 - 120) + 4,000,000 × 0.15 × (119 - 60) + 14,000,000 × 0.015 × 119 = 58,390,0007. A year at PHP 119
Over 52 weeks.
Yearly gain at PHP 119 (PHP):(2,000,000 × (119 - 120) + 4,000,000 × 0.15 × (119 - 60) + 14,000,000 × 0.015 × 119) × 52 = 3,036,280,000
What the exhibit shows
Heavy buyers already spend PHP 120 a week, more than the planned PHP 99 pass, so each heavy buyer who switches pays less.
The recommendation
Launch the weekly pass, but at PHP 119 rather than PHP 99. First, at PHP 99 heavy buyers switch and pay PHP 21 less each week, costing about PHP 42 million a week, which light upgraders and new buyers only just outweigh, for a net gain of about PHP 25 million a week. Second, at PHP 119 heavy buyers lose only PHP 1 each, and even with fewer light and new buyers the gain rises to about PHP 58 million a week, or about PHP 3 billion a year. Third, the gain depends on new and light users, whose response is the least certain number, so test before a national launch. Launch at PHP 119 in two regions for a month, compare take-up with the research, and watch network load, since weekly passes invite heavier use.
Risks: Rivals may launch a cheaper weekly pass; Heavier data use on weekly passes may need more network capacity.
Next steps: Run the two-region test with a control region; Measure the share of light buyers who switch at each price.
A strong candidate
Split the effect by segment, saw that heavy users pay less, sized each group, and used the curveball to find a better price.
A weak candidate
Multiplied 20 million users by PHP 99 and called it a large new revenue stream, ignoring the revenue that heavy buyers already bring.
Sources for this lesson (1)
- Recognized public explanations of case-interview concepts and frameworks
My notes on this lesson
0 of 5,000 characters. Saves automatically.
Completing lessons builds your skill levels and your readiness.
Spotted something wrong or out of date? Report a mistake. We check every report and correct the page.