سرمایه‌گذاری در جنوب‌شرق آسیا و مکعب یخی که آب می‌شود

A private equity firm was weighing an acquisition: a small Thailand-based company with a niche rail-electrification technology, operating across five countries, holding 80% market share in a market growing at 10% a year, except that 80% share was shrinking by roughly ten points annually as two competitors gained ground fast. The case handed me the full financials and asked for a valuation and a yes-or-no. No outcome given, no "right answer" printed anywhere in the material.

I worked it cold, my first instinct landing on caution, a business with an eroding position in a fast-moving technology sector reads as risky by default, before a single number gets touched.

The central question

When a calculation produces a conclusion that feels clean and reassuring, is that a reason to trust it more, or a reason to check it harder?

Key ideas

Then came the math, and it went wrong in the middle, on my side. I applied the ten-point annual share decline as a relative drop, 80%, then 72%, then 64.8%, instead of a straight point-by-point subtraction: 80%, then 70%, then 60%. The relative version, combined with a growing total market, produced a finding that looked almost elegant: profit staying roughly flat even as market share collapsed, the market's growth quietly offsetting the erosion. It was wrong, and I built a real argument on top of it before catching the error and redoing the five-year projection correctly. The honest version showed profit cut nearly in half over five years, not flat at all.

That correction mattered less than what came after it. Once the real numbers were in front of me, the actual decision wasn't about the trend, it was about time value. A shrinking asset today isn't worth today's price; it's worth what its future cash flows are worth once you account for the fact that money next year is worth less than money now, more so for a business this volatile. Discounting the corrected five-year profit stream at a rate that reflected the real risk here, not a stable industry's discount rate, something much higher, produced a present value well below the company's current annual profit. That gap is the whole case. It's the difference between what a nervous buyer sees on the label and what the business is actually worth once its own trajectory is priced in honestly.

My synthesis

The final call: buy, but only as a discount play, not a growth story. A shrinking company priced accordingly can still be a legitimate bet, the way a value investor buys something struggling because the price already reflects the struggle, but only if there's a credible plan to arrest the decline, not just a bet that a lower price makes yesterday's problems tomorrow's bargain. The more important takeaway sits underneath that call, though: the error and the correct answer were separated by one bad assumption about how a percentage decline compounds, and the wrong version was the one that felt more satisfying to arrive at.

Connections

This joins the "worked cold" series, Pepsi A.M., Heinz, Bud Light, Real Burger World, and ESPN, but shifts the failure mode from persuasive reasoning to persuasive arithmetic. The pattern is the same: something felt right before it was actually checked.

Questions I still have

Is there a habit of checking that would have caught a relative-versus-absolute error like this one before it got built into a full argument, rather than after? And how much should the emotional feel of a result, clean, reassuring, elegant, function as a warning sign in itself, independent of whether the underlying math is later confirmed to be wrong?

Final perspective

What actually stuck from this case wasn't the valuation method, which is fairly standard. It was catching a real, substantive arithmetic error mid-argument, watching it produce a conclusion that felt clean and reassuring, and having to throw that conclusion out once the correct numbers told a harder story. The elegant answer and the correct answer aren't always the same thing, and the elegant one is more dangerous precisely because it doesn't feel like it needs checking.