Chapter V

Chapter V: The Compounding Method

40 min · 12 ideas

"Investing is simple, but it isn't easy." — often credited to Warren Buffett; he has said it many times, but others wrote it before him

This chapter is the heart of the book. I've taken a lifetime of what I've learned, what I've done and where I went wrong, and boiled it down into a method that any family can follow.

1. My Investment Method in One Sentence

Take the money you save each month and invest a fixed amount at regular intervals in passive broad-market index ETFs; hold for the long run and never sell at the highs; and when a big drop or a crash comes, add more from new income.

Every phrase in that sentence is there for a reason:

This method grows out of classic value investing. Benjamin Graham and Warren Buffett taught us that "price is what you pay; value is what you get," and to buy more when things are cheap. John Bogle taught us not to look for the needle in the haystack, but to buy the whole haystack. And after using these ideas myself for more than a decade, and paying for my share of lessons, I finally realized: for an ordinary family, the best value investing is to buy more of the whole market in times of fear, and then wait patiently.

I, too, once believed that you should "sell when prices get too high." I even developed my own method for spotting bubble peaks, using trading volume, the number of accounts trading and margin debt. Later I tested it against nearly a century of US data, nearly sixty years of Japanese data and more than thirty years of Taiwanese data. I found that over the long run, it actually left you poorer. I tell the whole story in Section 6 of this chapter.

2. Reading the Market's Temperature

When I was young, I put a great deal of effort into studying "trends" and "timing": where interest rates were heading, whether the US dollar was strong or weak, whether individual investors had lost their heads. I hoped to use all of it to sell at the highs and buy back at the lows. I haven't thrown that knowledge away, but I use it differently now. I used to use it to guess tomorrow. Now I use it to see clearly how cold or hot today is, so my emotions don't carry me away.

This section has three parts: the temperature of money (monetary conditions), the temperature of people (psychology), and, once you can read the temperature, what to do about it (timing).

1. The Temperature of Money: Interest Rates, Exchange Rates and the US Dollar

(1) Interest rates are the gravity that pulls on every asset.
The image is Buffett's, from a 1999 Fortune article: interest rates "act on financial valuations the way gravity acts on matter." The higher the interest paid on deposits and government bonds, the less attractive stocks and real estate become, and the more easily their prices get dragged down. The lower interest rates go, the more money flows into risky assets. That's why rate-hiking cycles often come with stock market corrections, while rate cuts often inflate asset prices.

But don't treat this as a formula. Rate cuts often come because the economy has already weakened. Several times in history, after a central bank began cutting rates, corporate profits fell and unemployment rose, and only then did the stock market truly hit bottom. "Rate cuts = good news" is only half true.

(2) Taiwan's stock market and the NT dollar often move together.
When foreign investors buy Taiwanese stocks, they first have to convert US dollars into NT dollars, so the NT dollar rises. When they sell and take their money out, the NT dollar falls. That's why big rallies in Taiwanese stocks often come with a stronger NT dollar, and big falls with a weaker one. Watching the exchange rate gives you a rough sense of whether international money is coming in or heading out.

(3) The strength of the US dollar affects the whole world.
When the US raises interest rates, when its economy is relatively strong, or when something goes wrong in the world and everyone runs for safety, money flows back to the United States. The dollar strengthens, and emerging-market stocks and currencies often come under pressure together. When the dollar weakens, money flows back out to the rest of the world. For a Taiwanese family, this also means that when you hold overseas assets, the exchange rate itself is a risk, and also a form of diversification.

(4) Long-term government bonds and stocks don't always move in opposite directions.
Many people assume that when stocks fall, bonds will rise. But in 2022, as inflation heated up and the Federal Reserve raised rates rapidly, US stocks and long-term Treasury bonds fell hard together. The role bonds should play, and the detour I took with long-term US bonds, are covered in Section 7 of this chapter and in Appendix 2.

2. The Temperature of People: Fear and Greed

In the short run, the stock market can't be predicted. But one thing is almost certain: the crowd will swing back and forth between greed and fear.

• When it's hot: people around you who never touch stocks start asking what you're buying, the news reports new highs every day, account openings and trading volume soar, and more people borrow money to buy stocks. This is the moment the older generation meant when they said, "Even the shoeshine boys are giving out stock tips." (The line goes back to an anecdote about Joseph P. Kennedy Sr., who is said to have sold his stocks before the 1929 crash because even a shoeshine boy was giving him tips.)
• When it's cold: nobody wants to hear about stocks, the news is full of analysis saying "it will fall further," trading volume dries up, and many people sell at a loss and swear they'll never touch stocks again. In a bear market, people go through denial, anger, bargaining and depression before they finally reach acceptance (borrowing the psychiatrist Elisabeth Kübler-Ross's "five stages of grief") — and the market often starts quietly recovering just when everyone is most discouraged.

These swings in mood can be roughly measured with numbers: how far prices are above their long-term average, how far they've fallen from the peak, whether trading volume has exploded, whether margin debt has surged, and how high the fear index is. That is where this book's "market thermometer" comes from.

3. Once You Can Read the Temperature: What to Do

I used to treat the thermometer as a "sell signal." After backtesting long-run data from the US, Japan and Taiwan (see Section 6), I replaced that with the table below:

The thermometer is for keeping yourself in check, not for predicting the market. It only describes how cold or hot things are; whether to add more depends on whether you have new income, not on the temperature. What really makes wealth grow is still the fixed amount you invest every month, and the patience never to interrupt compounding.

4. The Temperature at the Time of Writing (October 2026)

As I write this edition, stock markets around the world are on the hot side:

• US stock prices are about 1.85 times their 10-year average, in the top 6% of all readings since 1881. The total market value of US corporate equity is about 2.5 times GDP, the highest since records began in 1952.
• Taiwanese stocks (monthly average) are about 2.7 times their 10-year average. The total market value of listed companies is about 4.8 times GDP over the last four quarters, also a record high.
• Yet the St. Louis Fed's Financial Stress Index stands at about −0.5, calmer than normal, and the 10-year US Treasury yield is above 5%, higher than the Fed's benchmark rate.

By my old method, I would have sold everything by now. And I already saw one "sell signal," back in 2024 — after which the TAIEX climbed from just over 20,000 points to nearly 50,000. So here's what I'm doing today: keep investing a fixed amount as usual, don't borrow, don't add extra, and save my strength for adding more in the next real crash.

This is what I myself was doing when I wrote the book, not buy-or-sell advice for you. For later numbers, see the thermometer page on the website, which only describes how hot or cold the market is.

3. Buying Low

Howard Marks: "It's not what you buy, it's what you pay that counts."

1. Value Investing

Graham said: "An investment operation is one which, upon thorough analysis, promises safety of principal and an adequate return." Taken apart, that's three things:

1. Thorough analysis: understand what you're buying. For a broad-market ETF investor, that means understanding which companies it owns, how much it costs and which market it tracks.
2. Keeping your principal: buy things that won't disappear, and don't pay too much for them.
3. Earning an adequate return: let time and compounding work for you.

2. Value vs. Price

Buffett said: "Price is what you pay; value is what you get." The lower the price you pay for the same thing, the more you get.

It has always struck me as strange. The same person who waits for the big annual sales to buy clothes or appliances does exactly the opposite with stocks: the higher prices climb, the bolder they get about buying, and the moment there's a discount, they run away. Marks has a way of putting it that I fully agree with: the point isn't "buying good things," but buying things well.

3. Mr. Market

"Mr. Market" is Graham's metaphor. I picture him as a neighbor with wild mood swings. Every day he knocks on your door with a price. When he's cheerful, he names a sky-high price; when he's gloomy, he'll sell for next to nothing. Whatever he quotes, you're free to ignore it. He'll be back tomorrow anyway.

This metaphor reminds me of one thing: his price is only an offer, not an order. When he's desperate to sell cheap, I can take a little more off his hands; when he's hawking at high prices, I don't have to dance to his tune. What I should worry about isn't his moods, but catching them.

4. Mean Reversion

Over the long run, the stock market heads upward. But day to day, prices get pushed around by a crowd's emotions, and they often overshoot, both up and down. André Kostolany compared it to an old man walking his dog: the dog (the stock price) races ahead one moment and lags behind the next, but there's a leash around its neck, and in the end it keeps circling back to the old man (the underlying businesses). When watching the market makes you anxious, think of that leash.

5. Margin of Safety

A margin of safety, put plainly, is "a cushion you leave for yourself." If you figure something is worth NT$100 and pay only NT$70, that NT$30 in between is your protective pad for when you're wrong. The cheaper you buy, the thicker the pad, and even if your estimate is off, you won't get badly hurt. The "30% off" here is just an illustration, not a fixed formula.

For the market as a whole, when is the margin of safety at its greatest? It's when everyone is most afraid. In his 1997 letter to shareholders, Buffett offered an analogy: if you're going to be buying hamburgers all your life, you should hope beef gets cheaper; likewise, if you'll keep buying stocks for many years to come, you should welcome falling prices.

6. A Market Crash Is an Opportunity, Not a Threat

Most people believe that "if you want to earn more, you have to take more risk." Marks sees it exactly the other way around: buy too expensively, and your risk is high while your return is low; buy cheaply, and your risk is low while your return is actually higher. Both are decided by the price you pay.

Here's how much the timing of a purchase matters, using actual data for the US S&P 500:

(Annualized total return, including dividends, in nominal US dollars. Data: Shiller's long-run S&P 500 data; my own calculations.)

It was the same S&P 500, bought just two years apart, yet 15 years later the annualized returns differed by nearly six percentage points.

7. Buying Low Also Means Low Costs

Investor's net return = market return − costs

Bogle called this "the relentless rules of humble arithmetic." Suppose the market returns 8% a year over the long run, and you invest NT$10,000 in the first year and let it compound for 50 years:

• With no costs: about NT$469,000;
• In a low-cost index fund (costs of about 0.1% a year): about NT$448,000;
• With frequent stock trading or an actively managed fund (fees plus trading costs of about 2.5% a year in total): about NT$145,000.

It's the same compounding either way. Rolling on your returns, it helps you; rolling on your costs, it bites you. And the longer the time, the more both are magnified. Choosing a low-cost broad-market ETF is another way of "buying low."

8. Dollar-Cost Averaging, a Lump Sum or Waiting for a Dip: Which Is Best?

This is one of the questions people ask most often. The answer has three layers:

1. If you already have a sum of money in hand, investing it all at once usually beats spreading it out. The market rises over the long run, so the sooner you invest, the longer you take part in the gains. A 2012 study by the US fund company Vanguard (using historical data from the US, the UK and Australia) found that investing a lump sum beat investing in installments about two-thirds of the time.
2. When you have income every month, dollar-cost averaging simply means "invest whenever you have money." It isn't about guessing the lows; it's about not having to guess.
3. Deliberately waiting for a low before investing usually does worse. Data scientist Nick Maggiulli compared "dollar-cost averaging" with "buying only after a drop." Even assuming you knew exactly where every bottom would be, the wait-for-a-drop strategy still lost to plain dollar-cost averaging more than 70% of the time.

I checked this myself, using data for the US from 1934 to 2026, Japan from 1968 to 2026 and Taiwan from 1997 to 2026:

The conclusion: don't keep money on the sidelines waiting for a crash. Hoarding cash for the lows won't leave you better off, and you may miss the gains.

9. In a Crash, Add More from "New Income"

So how do you act on the idea that "a crash is a chance to add more"? The key is that the extra money should come from new income (a year-end bonus, a raise, a trip you skip, and so on), not from money you've deliberately been holding back from investing.

Again using historical data, I tested this: "keep investing a fixed amount as usual, and in any month when the index is more than 20% below its peak, put in double."

Why did Japan come out worse? Because after its bubble burst in 1989, the Japanese stock market took 34 years to get back to its peak. Over that long a stretch, the money added "in the crash" stayed stuck in a stagnant market.

This gives us two important reminders:

1. Adding more in a crash assumes that the market will come back in the long run. A single country can lose decades (Japan is the example), so it's best to spread across the whole world. That's also why, for our family fund, I chose an index ETF that invests in stocks around the world.
2. Add only what you can afford: use spare money that won't affect your daily life, and never borrow to add more.

4. What to Own: Only Passive Broad-Market Index ETFs

John Bogle: "Don't look for the needle in the haystack. Just buy the haystack!"

1. Judge by "Total Return"

Whether an investment is good comes down to one thing: whether it can deliver a good annualized return, steadily and over the long run.

Annualized return = (1 + total return)^(1 ÷ number of years) − 1

Only by converting returns over different periods into "how much per year, on average" can you compare them fairly. What matters is the total return — price changes plus reinvested dividends — not just how much is paid out.

2. Why Not Pick Individual Stocks

• Over the long run, most stocks lose their shareholders money. Of the 26,168 companies listed in the US between 1926 and 2019, 57.8% reduced their shareholders' wealth; the stock market's wealth as a whole was created by a tiny number of companies. (Bessembinder, 2021)
• Companies, like people, are born, grow old, fall ill and die. Look at America's largest companies by market value: the list from twenty years ago and today's list have almost completely turned over. The horse-carriage, typewriter and photographic-film industries of the past have all disappeared. The investment research firm Dimensional studied US stocks since 1927. Of individual stocks bought at any point in time, about 18% were delisted for poor performance within 20 years (not counting those delisted because they were acquired), and only about one in five both survived and beat the market.
• A good company isn't necessarily a good stock. Good companies that everyone knows about are often already very expensive. Paying too high a price for the high growth you're hoping for is what Professor Jeremy Siegel calls the "growth trap."
• Even the experts can't win by picking. More than 90% of actively managed funds trail the market index over the long run.

3. Why the "Broad Market"

There are many kinds of index ETFs, but I buy only those that track a broad-market (market-cap-weighted) index covering the whole market:

• The index itself won't disappear: the index regularly swaps out declining companies and brings in rising ones. In effect, it automatically "drops the weak and keeps the strong," giving you the "stability of unchanging conditions" that compounding needs. An ETF's assets are held separately by a custodian, but some ETFs that were too small have been liquidated and delisted (investors were paid out at net asset value), so choose ones that are large enough.
• It shares in the growth of the whole economy: population growth, technological innovation, mild inflation and market competition make a country's total corporate profits grow over the long run.
• It costs the least: there's no need for frequent trading, and expense ratios are usually just a few hundredths to a few tenths of a percentage point.

Why not buy sector, thematic or high-dividend ETFs?

• Sector ETFs (semiconductors or technology, for example): at heart, they're a bet on a single industry. They're highly concentrated, highly volatile and prone to the growth trap. Today's star industry may not still be a star twenty years from now.
• High-dividend ETFs: high dividends don't mean high returns. Take Taiwan's two most representative funds: over the 10 years to October 13, 2026 (including distributions, at market price), the ETF tracking the Taiwan 50 Index returned about 756%, and the high-dividend ETF about 399% (MoneyDJ; results vary a lot depending on the period). A high-dividend ETF follows a "strategy" index that picks stocks by rules, which strays from the whole point of "buying the entire market," and its fees and turnover are usually higher too. Its payouts also bring taxes and the National Health Insurance supplementary premium. (This applies to Taiwan.)

I've bought sector ETFs and individual stocks myself in the past. I write about that experience in Appendix 1, "Why I Ended Up Buying Only the Broad Market."

4. Better Yet, Spread across the Whole World

A single country can lose decades: after its bubble burst in 1989, the Japanese stock market took 34 years to get back to its peak. No one can guarantee which country will do best over the next thirty years.

So for our family fund, I chose an index ETF that tracks the whole world's stock markets. It buys thousands of companies around the globe in one go, so we don't depend on the rise or fall of any one country.

5. Five Checkpoints for Choosing an ETF

1. Expense ratio: the lower the better. A long-run difference of 1 percentage point adds up to a big gap after thirty years.
2. Tracking error: the less it strays from the index it tracks, the better.
3. Size and liquidity: a fund that's big enough and trades enough has a narrow bid-ask spread, and it's less likely to be delisted.
4. Premium: don't buy when the market price is clearly above the net asset value. A premium eats straight into your return.
5. Taxes and distributions: with US-listed ETFs, a 30% tax is withheld from dividends up front; with ETFs listed in Taiwan that invest overseas, look at the distributions and the fees. (This applies to Taiwan.) Before you choose, work out the total return after taxes and fees. (From what I could find as of October 2026, the US Congress has not yet finished passing the United States-Taiwan Expedited Double-Tax Relief Act, so the US still withholds 30% of the dividends paid to Taiwanese investors. If the act takes effect, the rate is expected to fall to 10%–15%; for the latest status, go by the announcements of Taiwan's Ministry of Finance. Also, once a household's combined overseas income reaches NT$1 million in a year, it must be counted in the basic income amount under Taiwan's alternative minimum tax.)

This book explains only the principles for choosing; it doesn't recommend any particular fund company's products. Which fund you actually buy is for you to decide, after comparing the options in light of your own situation.

6. Gold and Other Assets

Buffett says gold is an unproductive asset: own an ounce of gold today, and a hundred years from now you'll still own one ounce. Businesses, by contrast, keep generating earnings and paying dividends. In his 2011 letter to shareholders, he made a comparison. With the value of all the world's gold, you could buy all the farmland in the United States plus more than a dozen of the world's most profitable oil companies, and still have money left over. A hundred years later, the farmland and the companies will have produced untold value, while the gold will have produced nothing at all. (His numbers: about 170,000 metric tons of gold worldwide, worth about US$9.6 trillion, could buy all 400 million acres of US farmland plus 16 ExxonMobils, with about US$1 trillion to spare.)

Over the long run, stocks have returned far more than bonds, cash and gold. Professor Siegel's analysis of US data going back to 1802 found that, after inflation, stocks returned roughly 6.5% to 7% a year over the long run, a figure known as the "Siegel constant." I ran the numbers myself using Professor Robert Shiller's public US data (1871 to 2026): after inflation, stocks with dividends returned about 7.8% a year and 10-year government bonds about 2.6%, while cash left sitting idle lost about 2.3% of its purchasing power each year.

5. Long-Term Holding: Time Is a Multimillionaire Maker

Buffett: "Life is like a snowball. The important thing is finding wet snow and a really long hill."

You can always earn more money, but time, once used, is gone. And time, of all things, is what compounding feeds on most.

1. Compounding Takes Long-Term Persistence

Many people are set on finding the method with the highest return. Morgan Housel, author of The Psychology of Money, offers a good reminder: rather than chasing a big win all at once, find a method that's pretty good and that you can keep doing, and then stick with it for decades. Compounding doesn't need explosive power. It needs endurance.

Buffett says his snowball got so big because he started rolling it very early, from the top of a long hill. As for how to get a hill that long, he said the trick "is either starting very young or living to be very old." There's also a saying often credited to Charlie Munger (I couldn't find a primary source for it) whose meaning I fully agree with: the first rule of compounding is never to interrupt it unnecessarily.

Housel did the math: at the time he was writing the book, US$84.2 billion of Buffett's US$84.5 billion net worth had been accumulated after age 50. As he puts it, "His skill is investing, but his secret is time."

2. Most of the Wealth Comes Late

There's a classic example that has circulated widely: suppose you invest NT$14,000 at the end of every year and earn 20% a year. By year 5 you'd have about NT$100,000; by year 20, about NT$2.61 million; by year 30, about NT$16.55 million; and by year 40, more than NT$100 million. The last five years alone would add over NT$60 million. (This example just shows the shape of compounding. The long-run return of a broad-market index is far below 20%, so real-life growth would be much slower.)

In this example, the rate of return never changes from beginning to end; the only thing that changes is the size of the snowball. For the first twenty years it looks as if you're running in place, and that's exactly when many people give up. Push through, and that's when the snowball really starts to put on weight.

There are many stories about compounding. The story goes that in 1626, Native Americans sold the island of Manhattan for goods worth about US$24. If that US$24 could have compounded at 6% a year until today, it would be roughly US$300 billion.

3. The Longer You Hold, the Better Your Odds of Making Money

Over time I came to see investing as rolling dice. Roll once, and it's hard to say whether you'll win or lose; roll enough times, and the results lean toward the side that favors you. What decides your odds isn't how smart you are, but how long you hold on. Using US S&P 500 data since 1928 (including dividends), I calculated how often you'd end up with a positive return if you bought at any point and held for different numbers of years:

(Data: Shiller's long-run S&P 500 data, January 1928 to September 2026; my own calculations.)

4. A Product's Return Isn't Its Investors' Return

In the ten years to the end of 2009, the US fund CGM Focus earned about 18% a year on average, making it the top performer of the decade. Yet Morningstar calculated that its investors, on average, actually lost about 11% a year. The fund made money while the people who bought it lost money, and the difference came down to when they got in and out: they rushed in after big gains and fled after deep falls.

So when I choose a product, I don't ask, "Which one has gone up the most in the past?" I ask, "Can I hold on to it?" A broad-market index ETF holds few surprises, but precisely because it's so plain, it's the easiest thing to hold with peace of mind for a lifetime.

5. Overcoming "Myopic Loss Aversion"

Losing NT$1,000 hurts far more than gaining NT$1,000 feels good. That's human nature. Stock prices rise and fall every day, so the more often you look, the more often you see "a loss," the more uneasy you feel, and the more your hand itches to sell. In 1997, behavioral economist Richard Thaler and his colleagues ran an experiment: the more often people saw their investment results, the less money they dared to put into stocks, and the less they earned in the end. It isn't that Buffett has no fear of falling prices. It's that his eyes are on compounding ten or twenty years out, so the ups and downs right in front of him naturally look small.

Here's a simple method for you: don't check stock prices every day. Once your monthly investment is set up, checking once or twice a year is plenty.

6. Why I No Longer Sell at the Highs

In this section, I want to be honest with you about one stretch of my own journey.

1. I Used to Believe "If the Price Is Too High, Sell"

For many years, my investment system included a rule: "sell in a bubble." I closely studied the Taiwan stock market's bubbles of 2000 and 2008 and put together a set of indicators for spotting the top:

• Price: the TAIEX at 150% or more of its 10-year moving average (what I called the "10-year-line ratio");
• Volume: monthly trading volume at 4 or more times its usual level, a "sky-high volume" month;
• Crowd interest: the number of accounts that traded that month hitting a record;
• Leverage: US margin debt hitting a record.

I believed that when these indicators peaked together and began to slide, the bubble was ending and it was time to sell everything. Then, once volume had shrunk into a "volume trough" and crowd interest had fallen to around 40% of its peak, I would buy it all back.

Looking back, this method called both 2000 and 2008 with great accuracy. But what happened in 2024 forced me to rethink it.

2. One Misjudgment

In July 2024, the TAIEX reached 24,416 points, and the 10-year-line ratio hit a towering 193%, higher than in 2000 or 2022. The index, trading volume and the number of accounts trading all set new records at the same time and then turned down. By my rules, that was the signal to sell everything. The TAIEX did drop to just over 17,000 points at one point in April 2025, but it then climbed steadily, and by October 2026 it was near 50,000 points, double where it stood when I issued my "sell signal."

3. Testing It with More Countries and Longer Periods

Taiwan offers too few examples, so I added data for the US from 1934 to 2026 and Japan from 1968 to 2026. Together with Taiwan from 1997 to 2026, I used them to test the rule "sell when the price is too far above its 10-year average" (investing the same amount every month):

(US figures include dividends and are adjusted for inflation, with cash earning Treasury-bill interest. Japan uses the OECD share price index and Taiwan uses the central bank's TAIEX series; both are price indexes without dividends. Taiwan's dividend yields are high, so once dividends are counted, holding comes out even further ahead.)

The results are clear:

• "Sell when pricey, buy back when cheap" greatly reduces wealth over the long run, because the market can stay in "pricey" territory for many years. People who sell often end up waiting more than a decade, even twenty years, and miss huge stretches of gains.
• The conclusion is the same in all three markets, even counting Japan's 1989 bubble, the biggest in history. That's because the rule had already sold out decades before that bubble, missing Japan's great rise in the 1970s and 1980s.
• The signal means something only at the extremes: when US prices were more than twice their 10-year average, the median real return over the next 5 years was −3%. But such moments are rare, and even when one comes, it's hard to use it to come out ahead of simply holding.
• The more complicated "alert plus moving-average break" approach does come out slightly ahead, but the margin is small, the results differ every time, and it's hard for an ordinary family to stick with it year after year.

4. Even Buffett Couldn't Do It

In his letter to shareholders for 2003, Buffett admitted that failing to sell several big holdings during the dot-com bubble was a major mistake. But notice that he was talking about individual stocks priced far above their intrinsic value. He has never advised ordinary people to try to guess the top of the whole market. Instead, he has urged them again and again to buy low-cost index funds.

In 2022, Jeremy Grantham judged that the US stock bubble had further to burst; instead, US stocks went on to new highs. If even the very best investors find it hard to call the top accurately, ordinary people should rely on it even less.

5. What I Do Now

• I don't sell because things are "too expensive." The market rises over the long run, and highs are followed by higher highs.
• When the market overheats, I do only three things: I don't borrow or raise my bets; I check whether my stock allocation has gone above my original target; and I keep investing my fixed amount at the usual pace, without adding extra.
• When the market panics and crashes, I add more from new income.
• When do I sell? Only in two situations: when I truly need the money to live on (for example, withdrawals in retirement), or when my asset mix needs adjusting (rebalancing).

I'm writing this down so my family will know: methods can be corrected, and beliefs must stand up to the facts. Over a lifetime of investing, the most important thing isn't proving you were right, but being willing to change when you're wrong.

7. Don't Lose Money

"Rule one: never lose money. Rule two: never forget rule one." — often quoted as Warren Buffett's words

"Don't lose money" doesn't mean prices will never fall. It means don't make mistakes that permanently destroy your principal. A temporary drop in prices is only a fluctuation on paper, as long as what you own is the whole market (the index itself won't disappear) and you aren't forced to sell. Real losses come from buying things that can disappear, from borrowing and getting wiped out by a margin call, or from selling at the very bottom in a panic.

1. Mistakes Are Unavoidable, but Avoid the Fatal Ones

• Don't put all your eggs in one basket: if a single stock runs into trouble, you could lose everything;
• Don't follow the crowd: the star stocks and star industries everyone is scrambling to buy already have plenty of hope packed into their prices;
• Don't overpay: pay too much, and even the best thing won't make you money.

Put simply: invest for the long run in low-cost broad-market index ETFs.

2. Circle of Competence: Don't Buy What You Don't Understand

Buffett's point is simple: don't buy what you don't understand. In his 1996 letter to shareholders, he wrote of the circle of competence: "The size of that circle is not very important; knowing its boundaries, however, is vital." For most families, the best circle of competence is this: I don't understand every company, but I do understand that "the market as a whole grows over the long run."

Reading is the best way to widen your circle of competence. Buffett is famous for his love of reading, and Munger said you should go to bed every night a little wiser than when you got up that morning.

3. Emotional Intelligence: Patience, Endurance, Persistence and Discipline

I love the phrase "still waters run deep." In investing, what counts isn't who thinks fastest, but whose mind is calm enough.

• Patience: when everyone else is selling and you're still buying and still holding, it's bound to feel uncomfortable, because it goes against instinct. But that's investing: the discomfort comes first, and the money comes later.
• Endurance: how long does it take to get back to the old high after a crash? A year or two if it's quick, more than ten years if it's slow. Endurance means you keep investing every month through those days when you can't see the end.
• Persistence and discipline: investing is "simple but not easy." The method takes one sentence to explain; the hard part is not letting it drift for decades.

When I read Getting Rich Slowly (a Chinese book; the title is translated), the passage on patience struck a deep chord with me. In my own words: patience isn't doing nothing. It's being willing, at the very moment you doubt yourself most and most want to give up, to walk back onto the right road you started on. And that is exactly what most people in the market can't do.

4. Patience over Haste: Beware of Scams, and Never Borrow, Gamble or Use Leverage

Not borrowing to invest and not using leverage means keeping a way out for yourself. Things won't always go smoothly, and the only predictable thing about the stock market is that it's unpredictable.

In his letter to shareholders for 2017 (published in February 2018), Buffett wrote that no one can predict how far the market might fall in the short run. Even if you've borrowed only a little and your position isn't immediately threatened, your mind can be rattled by frightening headlines and breathless commentary, and when your mind is unsettled, it's hard to make good decisions.

Borrowing to invest means putting three things on the table at once (an article interpreting Buffett's 2017 letter sums it up this way, and I agree):

1. You lose your judgment: the urge to borrow usually comes when prices have risen so much that you're itching to get in, or when your gains have made you overconfident. That is exactly when your head is least clear.
2. You lose your composure: people who haven't borrowed see a drop as a sale; for people who have, a drop means a margin call. They're frightened half to death, and they may be forced to sell at the very bottom.
3. You lose your opportunity: when the real crash comes, your pockets are already empty.

Kostolany warned again and again: don't borrow money to buy stocks. Even if your view is right, you may not hold out until the moment you're proven right. There's an old saying in investing (one Howard Marks often quotes, too): "There are old investors, and there are bold investors, but there are no old bold investors."

And beware of scams: anything that guarantees profits or high returns, or asks you to join a chat group and buy stocks on a "teacher's" say-so, is a scam. (See Appendix 3 for details.)

5. Diversification and Asset Allocation

Diversification works on three levels:

1. Across companies: buy the whole market, not a single stock;
2. Across countries: the whole world's stock markets, not a single country's;
3. Across time: invest a fixed amount regularly and keep buying all your life.

The Nobel Foundation is a good example of asset allocation. When Alfred Nobel died in 1896, he left an estate of about 31 million Swedish kronor. In its early years, the foundation could invest only in deposits and government bonds. Not until 1953, after inflation and taxes had badly shrunk its assets, was it allowed to invest in stocks and real estate. Since then it has invested according to long-term allocation targets and rebalanced regularly, and it has now been awarding prizes for more than a hundred years. At the end of 2025, the foundation's investment capital was about 6.7 billion Swedish kronor.

6. Where Should the Money Go? Cash, Deposits and Bonds

With a "never sell at the highs" approach, most of your money will stay in stocks for the long run. But every family still needs some money that isn't in the stock market:

About bonds, I want to stress two things:

1. Long-term bonds are not a "capital-protected" product. When interest rates rise, long-term bond prices can fall sharply. As US interest rates climbed from 2020 to 2023 (the Fed raised rates again and again in 2022–2023), the ETF tracking US Treasuries with 20 or more years to maturity (TLT) fell about 52% in price, from its August 2020 high to October 2023. Long-term bond prices are driven by too many factors, so they aren't something ordinary families should use to guess interest rates or trade for profit. I took a detour here myself: I bought too early, and luckily sold when I was close to breaking even including interest, losing only a little over NT$10,000 (see Appendix 2, "Detours I Took").
2. Foreign bonds carry exchange-rate risk. For a family whose income and spending are both in NT dollars, US-dollar bonds can lose money when the NT dollar strengthens, even if the interest is good. Money for emergencies or short-term needs is best kept in NT dollars.

Buffett's instructions for the trust that will manage his estate are to put 90% in a low-cost S&P 500 index fund and 10% in short-term government bonds (Berkshire Hathaway's letter to shareholders for 2013). Our family fund follows a similar spirit: the great majority goes into a global stock index, a small portion into global bonds, and we check the proportions every few years.

7. Enjoy Solitude; Don't Let the Media and the News Sway You

What TV, the internet and bank wealth advisers eagerly recommend is usually what they want to sell. Their income comes from your buying and selling, not from your profits. When you hear a recommendation, first ask: "Why are they telling me this?" Once you've thought that through, you won't be afraid to walk a different road from everyone else, on your own.

Key Takeaways

1. My method: invest a fixed amount regularly in passive broad-market index ETFs, hold for the long run, never sell at the highs, and add more from new income in a crash.
2. For the market as a whole, the spirit of buying low means buying more in times of fear; low cost is a low price too.
3. Don't hoard cash waiting for a crash. The extra money should come from new income, and it's best spread across the whole world.
4. Buy only broad-market (market-cap-weighted) indexes: no individual stocks, sector funds or high-dividend funds. When choosing an ETF, look at fees, tracking error, size, premium and taxes.
5. The longer you hold, the better your odds of making money. Since 1928, US stocks held for 20 years or more have never lost money after inflation — but that is the US record. In Japan, someone who bought at the 1989 peak was still deeply in the red 20 years later. The past is no guarantee of the future.
6. "Sell when pricey" has greatly reduced wealth over the long run in the US, Japan and Taiwan alike. I've dropped that rule.
7. Don't lose money: don't bet everything on one thing, don't borrow, don't use leverage, and diversify. Keep money you'll need soon in NT-dollar deposits or short-term bonds.