Is Buying the KOSPI Index Diversification? Actually, It's All-In on Semiconductors

A few days ago, the YouTube algorithm pushed an interview my way. The title was provocative.

“The One Treasure I Gained from Having a $700,000 Lunch with Warren Buffett.” I clicked, and there was Mohnish Pabrai, an Indian-American investor, talking about Korea.

The comment section was already a festival, celebrating his high praise for Samsung Electronics and SK Hynix. Everyone was cheering for his remark: “Never sell Samsung or Hynix.”

However, after watching the entire original video, I felt a bit strange.

What Pabrai emphasized most wasn’t actually the semiconductor stocks themselves.

It was about how the KOSPI index functions differently than people think, and what he would buy if he were Korean.

But when I searched for domestic articles, they all converged on a two-line summary: “Hold Samsung/Hynix for life” and “Korea’s population decline is the biggest risk.”

Everyone transcribed the praise for semiconductors, but the conditions attached to that praise and the alternatives mentioned afterward seemed to have vanished.

So, this post isn’t about semiconductors.

I want to dig into the part of the interview that is genuinely useful for novice investors—the structural story that was relatively under-reported in domestic media.

Is Buying the KOSPI Index Really Diversification?

There is usually one reason to buy an index fund:

To avoid individual stock risk and spread investments across the entire market. But the point Pabrai made was that this premise doesn’t fit the KOSPI well.

Since the market cap of Samsung Electronics and SK Hynix effectively dominates the entire index, buying a KOSPI index fund or ETF isn’t really diversifying across the economy; it’s similar to making a concentrated bet on two memory semiconductor companies.

It makes sense. Even looking at the KOSPI 200 composition, the weight of these two semiconductor giants is overwhelming, and if this sector wobbles, the entire index follows.

There is a significant gap between the perception that “buying the index means diversification” and the reality that “in practice, you are all-in on one sector.” You could call this an index illusion.

What’s interesting is that this diagnosis isn’t pessimistic about semiconductors.

On the contrary, Pabrai referred to Samsung, SK Hynix, and Micron as “companies selling shovels in an AI gold rush,” evaluating that thanks to long-term contracts with Big Tech clients, their revenue for the next few years is already largely secured.

In fact, Pabrai even mentioned that selling Samsung and SK Hynix in the past was a painful mistake that violated his principles—a point most domestic articles picked up as their headline, focusing solely on his praise for semiconductors.

The problem is what comes next. Is it realistically possible for a retail investor to track and respond to the subtle signals of the memory semiconductor industry, facility investment cycles, and technology transition timing in real-time?

Pabrai himself admitted that this area is highly difficult. He meant that just because you like a stock, it doesn’t mean anyone can deeply understand it—a caveat that was almost entirely omitted in domestic reports.

Then What Should You Buy? — The Berkshire Alternative

This is where Pabrai actually played his card: Berkshire Hathaway Class B (BRK.B). He stated that if he were Korean, he would spend less than he earns and, with the leftover money, accumulate Berkshire monthly like an index fund instead of the KOSPI.

Why Berkshire instead of KOSPI? His logic was that it is already diversified through dozens of subsidiaries spanning railroads, energy, insurance, manufacturing, and consumer goods, and it holds massive amounts of cash, making a market crash a structural opportunity.

There was also a specific reason he chose it over the S&P 500.

He argues that the US stock market is currently in a high-valuation zone, and there is a possibility that the index itself will experience a stagnation period where it fails to rise significantly for the next decade.

Indeed, there were periods from 1965 to 1982, and from 1999 to 2010, where the S&P 500 returns were essentially 0% for over a decade.

I think we should take this talk of stagnation with a grain of salt.

No one can be certain about future returns, and Pabrai might be speaking from a position of promoting his own fund. However, the logical structure that “assets with large cash reserves and relatively reasonable valuations are advantageous during stagnation periods” is hard to refute.

Here is a comparison of KOSPI and Berkshire along a few axes:

Of course, Berkshire isn’t a panacea. There is constant talk about Warren Buffett preparing for retirement, and long-standing concerns that the company has become too large to generate excess returns as it once did.

This is Pabrai’s personal perspective, and it is safer to interpret it as the principle that “you should take a look at how an asset you buy under the name of an index is actually composed” rather than taking it as advice to buy a specific stock.

The Name Rick Guerin, and Why Leverage is Absolutely Forbidden

Personally, the most impressive part of this interview wasn’t about semiconductors or Berkshire, but the story of a man named Rick Guerin.

In the early 1970s, there was a third investor alongside Warren Buffett and Charlie Munger: Rick Guerin. The three of them exchanged investment ideas, and Guerin’s track record was just as impressive as Buffett’s.

It is said that when Pabrai asked Buffett during their 2007 charity lunch what had happened to Rick Guerin, Buffett replied that while he and Munger knew they would eventually become very wealthy and were in no rush, Guerin was in a hurry.

During the 1973–74 period, when the US stock market crashed by nearly half, Guerin was using margin loans—investing with debt—and eventually faced a margin call.

Needing cash, the last asset Guerin could sell was the Berkshire Hathaway stock he held.

He sold it to Buffett at around $40 per share. That same stock is now worth over $700,000 per share.

It’s a laughable story when looking at the numbers, but the reason this anecdote is terrifying is that it wasn’t because Guerin was incompetent.

Many evaluate that his skills were on par with Buffett and Munger. Yet, for one single reason—he was using leverage—he was forced to sell assets at a time and price he didn’t want.

The core of this story seems to be that no matter how skilled you are, if you are in debt, you end up selling not when “you want to sell,” but when “the creditor tells you to sell.”

Seeing similar patterns repeat in the Korean market, this doesn’t just sound like an old American story.

We have all seen in the news cases of people who scrape together wedding funds or business capital, put them into surging thematic stocks, and collapse in an instant.

Pabrai described this by saying that human nature doesn’t change, and if you don’t learn from history, tragedies are bound to repeat.

213 Checklists, Ultimately Narrowed Down to Three

Pabrai says he runs a 213-item checklist when buying a stock. Hearing the number makes it sound like some grand formula, but it actually summarizes into three main branches.

First is the exclusion of leverage, as mentioned. Since over 20 of the 213 items are used to check risks related to debt, its weight is significant.

Second is a sustainable economic moat. In capitalism, if a business looks profitable, competitors flock in an instant to erode excess profits.

The key is whether there are entry barriers that don’t easily crumble over time, like the Coca-Cola brand or Mastercard’s payment network.

His comparison of Hyundai Motor and Mastercard was interesting. He noted that while Hyundai is a great company, it is in a position where it must constantly engage in a price war with Toyota, Tesla, and BYD, whereas Mastercard has a clear moat but is always traded at a high price because everyone already knows its value.

Therefore, he said the real opportunity is when a company has a great moat but the market is terrified and dumps it at a bargain. He cited an experience from the 2019 currency crisis in Turkey, where he bought a warehouse company with $800 million in net assets for a market cap of $15 million while foreign investors were fleeing. He said the company’s value jumped over a hundredfold in seven years.

Third is management. No matter how good the moat is, it’s useless if the person defending the castle is incompetent or greedy. He looks at whether they treat shareholders as partners and where they reinvest the money they earn.

In fact, this is the most difficult area for a retail investor to verify, as it’s not easy to judge the true character of management just from public disclosures or IR materials.

Where Should a Beginner Start?

Pabrai advises beginners not to start with stock analysis. Instead, he suggests taking steps—a process of going from kindergarten to university in his own words.

The kindergarten stage is simply starting by regularly accumulating core assets.

Spend less than you earn, never borrow money, and consistently invest the surplus.

Moving to the elementary school stage, he suggests observing the brands you actually consume in your daily life.

His logic is that this is one of the few areas where retail investors can get ahead of institutions. He cited Burger King as an example. When they changed the Whopper’s bun and mayonnaise, sales jumped 10% in one quarter. He argues that consumers who eat at the stores feel the change before analysts write their reports. The story of Duan Yongping, the founder of Oppo and Vivo, holding Apple stock as his largest position, was cited in a similar context.

The middle and high school stage is patience. Training to hold on to cash and wait until a good company is obviously cheap because the market is scared.

And only in the university stage should you attempt to add individual stocks you have high conviction in with a small amount of money, while maintaining your core asset allocation.

If the order is reversed—that is, if you go all-in on individual stocks without conviction—the risk of repeating the Rick Guerin story increases.

There is Only One Reason to Sell

Finally, his criteria for selling were interesting.

Pabrai cited selling his Ferrari stake too early as his most painful mistake. He bought a 1% stake in Ferrari for less than $10 million, but sold it in a hurry once the share price neared its fair value. Considering Ferrari’s current market cap, that 1% would have been worth a billion dollars.

He explained that being too faithful to the Benjamin Graham-style principle of buying undervalued assets and selling when they reach fair value actually worked against him.

Another example he used was the Nifty 50.

In the late 1960s, there was a trend in the US of unconditionally buying 50 blue-chip stocks. The calculation is that even if 49 of them later failed and their value went to zero, if you had just kept Walmart and never sold it, the total portfolio return would have overwhelmed the S&P 500.

Ultimately, it’s the logic that a few super-blue-chip companies drive most of the long-term returns; if you invert this, selling those few companies early is the most expensive mistake you can make.

So when should you sell?

Pabrai’s criteria is only one: when permanent decline is confirmed.

A drop in the stock price or noise in the media is not a reason to sell.

When American Express stock halved due to a loan fraud case in 1964, Buffett reportedly went to restaurants himself to check if people were still paying with Amex cards instead of judging by the news. After confirming that brand trust remained intact, he poured 40% of his fund’s assets into it.

If you apply this standard to Samsung Electronics and SK Hynix, it means as long as you can continue to answer “yes” to questions like “Will the memory 3-company oligopoly be maintained in 10 years?” or “Will memory remain an essential good in the AI era?”, the correct answer is to do nothing and not be swayed by stock price fluctuations. Conversely, if you can’t answer those questions with confidence, it might be a sign that the problem isn’t the stock, but that you don’t properly understand the industry.

Having summarized this, I can understand a little better why domestic reports focused so heavily on semiconductors.

“Never sell Samsung or Hynix” is a much more provocative sentence that generates more clicks. But what Pabrai himself repeatedly emphasized wasn’t a specific stock, but rather an attitude of not using leverage, an attitude of not touching areas you don’t understand, and an attitude of holding on until the reason to sell becomes clear.

How actionable this advice is depends on the individual, but it was at least content too valuable to pass up just by looking at the headlines.