Skip to content

← Money

Risk and Reward

Chapter 21 · Owning a Share of Things

In this chapter 5 sections
  1. Fair odds and good odds
  2. What a fall does to a sum
  3. How long the money can wait
  4. Lending or owning
  5. Try this

Suppose someone offers you a bet on one toss of a coin. Heads, you win Rs. 1,500. Tails, you lose Rs. 1,000. The odds are in your favor: play it many times and you would come out ahead by about Rs. 250 a toss. Most people turn it down all the same.

Daniel Kahneman, a psychologist who won the Nobel Prize in economics in 2002, spent years with his colleague Amos Tversky studying how people choose between bets like this one, asked in dollars. In his Nobel lecture he summed up what they found. Most people refuse a bet with even chances to win and lose unless the possible win is at least twice the possible loss. Losing a thousand rupees hurts more than winning a thousand pleases, so a chance of losing has to be paid for with a bigger chance of winning.

That one finding explains most of this chapter. Nobody takes a chance for nothing. If one way of keeping money is sure and another isn’t, the uncertain one has to offer more on average, or people keep their money in the sure one. The extra it offers is the reward for taking the risk. It is only an average: some who take the chance do much better than the sure thing, and some do worse. If they all did better, there would be no risk, and if there were no risk, there would be no reason to pay extra. So a promise of high return with no risk is a promise of something that can’t exist. Chapter 12 explains why from the side of whoever pays the interest, and Chapters 17 and 18 show what quick money and high rates cost the people on the other side of them.

Fair odds and good odds

A bet is fair when, played many times, it leaves nobody ahead: win Rs. 1,000 on heads, lose Rs. 1,000 on tails. The average result is nothing. A careful person refuses a fair bet with money they care about, and for the reason Kahneman found: the loss weighs more than the gain. Chapter 19 turns this round. There, people paid more than their average loss to get rid of a risk, because a household would rather lose a little for certain than face a small chance of ruin. Here, someone is paid to carry a risk instead.

A bet is good when, played many times, it leaves you ahead, like the coin at Rs. 1,500 against Rs. 1,000. Good odds still aren’t a promise. One toss can lose. Ten tosses can lose. Only over many tosses does the average start to show, and how many tosses you can afford to play before it does is the whole question.

Money kept anywhere is a kind of bet. A deposit at a bank is about as near to a sure thing as an ordinary person can find. Nepal’s commercial banks paid an average of 3.07% a year on deposits in mid-October 2026, and Rs. 1 lakh put in comes back as Rs. 1 lakh plus that interest. A share in a company is a bet with good odds, if the company and the economy grow, and nobody promises the odds. What a share brings in is the dividends it pays plus whatever its price has done, and either can be small, or nothing, or worse.

So risk, in money, doesn’t only mean the chance of losing. It means not knowing. A thing whose result could be anywhere from a big loss to a big gain is risky even if it usually does well, and the wider the range of what could happen, the more risky it is. A sum that grows by exactly 3% has no risk at all and a small reward. A sum that might grow by 30% or shrink by 20% in a year has a lot of both.

How much risk a person can carry isn’t the same for everyone. It depends on what else they have to fall back on, how steady their income is, how soon they will need the money, and how they are made: some people sleep through a fall in prices and some can’t. A family with one income and a loan to repay can afford less risk than the same family with two incomes and none.

What a fall does to a sum

Ups and downs have an odd arithmetic, and it is worth getting straight, because it surprises nearly everyone. Say Rs. 1 lakh falls by half, to Rs. 50,000. To get back to Rs. 1 lakh it now has to double, a rise of 100%, because the rise is counted from the smaller sum. The deeper the fall, the steeper the climb back:

A fall of needs a rise of
10% 11%
20% 25%
25% 33%
50% 100%
75% 300%
90% 900%

The same arithmetic means that a rise and a fall of the same size don’t cancel. Rs. 1 lakh that gains 50% in one year and loses 50% the next ends at Rs. 75,000. Lose first and gain after, and it still ends at Rs. 75,000. Each change multiplies whatever is there at the time: a gain of half multiplies it by 1.5, a loss of half by 0.5, and 1.5 × 0.5 is 0.75, in either order.

This is why the average year can flatter a thing that goes up and down. Take the made-up investment in the widget below, which in any year does one of eight things, each as likely as the others: it loses a quarter, loses a tenth, stays the same, or gains 10, 15, 20, 25 or 40%. Its average year adds 9.4%. But kept up year after year, those eight kinds of year grow money by only about 7.6% a year, because the years multiply rather than add, and a loss multiplies by less than a gain of the same size gives back. The steadier a thing is, the smaller this gap. A deposit has none at all: its average year and what it grows by each year are the same number.

Nepal’s own share market shows the climb back in real numbers. Nepal Rastra Bank, the central bank, records the NEPSE index, the number that follows the prices of the shares listed on the Nepal Stock Exchange, four times a year, in the middle of January, April, July and October. In mid-July 2021 it stood at 2,883. By mid-October 2022 it was 1,858, about 36% lower, and a fall of 36% needs a rise of 55% to undo it. Those are readings taken four times a year, so the market’s best and worst days can fall between them: its record close came a month later, in August 2021, as Chapter 23 tells, and Chapter 24 follows the fall. Five years on, in mid-July 2026, the index was still about a tenth below where it had stood in July 2021.

How long the money can wait

Time doesn’t remove risk, but it changes it. Over one year, a share’s price can do almost anything. Over many years, the good years and the bad have more chance to even out, and the reward for risk, if there is one, has more time to show.

America’s record is the one every book on the subject refers to, because it is long and carefully kept. Take the shares of America’s largest companies together, with their dividends, year by year from 1928 to 2025: 98 calendar years, before rising prices, costs and tax. They lost money in 26 of those years, about one year in four, and in 1931 they lost more than two-fifths. Take every stretch of ten years in a row instead, and the shares ended lower than they began in only 5 of 89 stretches. Over every stretch of twenty years they ended higher, and ahead of short-term loans to the American government, about the safest place there is for dollars.

That record comes with a warning built in. It is famous partly because it turned out so well, and nobody knew in 1928 that it would. Japan’s is the other standard case. Tokyo’s main share index closed 1989 at 38,915.87, at the end of a decade of soaring prices. Then it fell, on and off, for nearly twenty years: in March 2009 it closed at 7,054.98, more than four-fifths down. It didn’t close above its 1989 level again until 22 February 2024, 34 years later. Someone who put money in at the end of 1989 and needed it back within thirty years had waited and still lost. That index counts prices only, so the dividends paid along the way made the wait somewhat kinder than it looks.

And a whole market is the gentle case. One company can do far worse, because it can fail outright. Morgan Housel, an American who writes about money, puts it in four words, “Long tails drive everything”: the rare extremes at either end decide most of the result. In an essay called “Tails, You Win” he quotes a study of the companies in one broad American index from 1980 to 2014. Forty in every hundred “lost at least 70% of their value and never recovered”. At the other end, about seven in every hundred did so well that, as Housel reads it, “effectively all of the index’s overall returns came from 7%” of them. Most of a market’s gain comes from a few big winners, and most single companies are not one of them. That is why spreading money over many companies matters so much, which is the subject of Chapter 25.

The widget makes the same point with made-up numbers, so that nothing in it is a forecast. A hundred savers each put Rs. 1 lakh into that up-and-down investment, and each has their own run of good and bad years. Beside them sits a deposit paying a steady 5%, a round rate like Chapter 10’s, above what banks pay today.

Try it

A hundred savers, and how long they can wait

Each square is a made-up saver. Their yearly results, from a quarter lost to 40% gained, and the deposit’s 5% are made up, not any market’s history.

less than they put inmore, but behind the depositahead of the deposit
Less than put in
24 of 100
Behind the deposit
34 of 100
Deposit at 5%
Rs. 1.05 lakh

After 1 year, the unluckiest saver has Rs. 75,000 and the luckiest Rs. 1.4 lakh. The deposit has Rs. 1.05 lakh, the same for everyone.

After one year, 24 of the hundred have less than they put in. Slide the years along, and the red squares thin out, unevenly, since a run of bad years can push a few more under for a while: after ten years 12 are still below Rs. 1 lakh, after twenty, 5. Waiting helps a great deal. But the red never quite goes, and even after twenty years nearly a third of the savers would have done better in the deposit. The spread grows too. After twenty years the luckiest saver has nearly Rs. 35 lakh and the unluckiest less than half a lakh. A longer wait makes a loss less likely and the range of results much wider.

So the question to ask of any sum is when it will be needed. Say Sabina, who is made up, has Rs. 60,000 saved for a course that starts next year. If that money sits where it could fall by a third, there is a real chance it will be Rs. 40,000 on the day the fees are due, and the course won’t wait for prices to recover. Money that will be needed soon, or that might be needed suddenly, like the fund for surprises of Chapter 9, has no business being anywhere it can fall. A deposit in a licensed bank won’t fall in rupees, and is guaranteed up to a limit if the bank fails (Chapter 12). It can still fall behind rising prices, as Chapter 10 explains, and that small, slow loss is the price of being sure. Only money that can wait years, and wait through a bad patch without anyone needing it, can afford a chance of falling, and even then nothing promises it a good result. None of this is advice to put money anywhere. It is how the arithmetic works.

Lending or owning

All the paper people hold money in comes in a few families, and knowing which family a piece of paper belongs to tells you most of what to expect from it. Finance calls them financial assets. A field or a cow is a real asset, the thing itself. A financial asset is a claim on someone: a promise, written on paper or kept in a computer, that some person, bank, company or government will pay you. Its value is only as good as the promise and whoever makes it.

Five questions sort out almost any of them. What does it pay, and is that fixed in advance? How sure is it? How quickly can it be turned back into cash without losing value, which finance calls liquidity? How long does it run before it ends? And are you lending or owning?

The last question matters most. Finance calls the two sides ownership and creditorship, a creditor being anyone who is owed money. When you lend, the borrower promises you a fixed return: so much interest, and your money back when the agreement says. A deposit is lending to a bank. A bond or a debenture is lending to a government or a company, usually for years, with interest paid along the way. The lender knows in advance what to expect, and if the borrower gets into trouble, lenders are paid before owners. Lending has a fixed return, and its main risk is that the borrower can’t pay.

When you own, nobody promises you anything. A share is a slice of a company, and its owner gets whatever is left after everyone else is paid: workers, suppliers, lenders, the tax office. In a good year that can be a great deal. In a bad year it can be nothing, and if the company fails, the owners are last in the queue. Owning has a variable return, and that is exactly why, on average, it has to offer more. Some papers sit between the two, and Chapter 22 meets the commonest of them, the preferred share.

The time a paper runs splits them another way. Short-term papers are repaid within a year: a call deposit, a three-month fixed deposit, or a treasury bill, which is a loan to the government for under a year. Long-term ones run for years: a ten-year bond, or a share, which has no end date at all and is ended only by being sold to someone else. Short and fixed is the calm end. Long and variable is the lively end. Units of a mutual fund, a pool of many people’s money spread over many papers, take on whatever their pool holds.

The next chapters take the two big families one each. Chapter 22 is about owning, what a share is and what it might be worth; Chapter 25 comes back to lending, with bonds and debentures, and to spreading money across many things at once.

What a reader should keep from all this is short. Nobody pays extra for nothing, so a higher return always comes with a chance of doing worse, and a promise of high return with no risk is either mistaken or a trap. A fall takes a bigger rise to undo, and a thing that swings grows more slowly than its average year suggests. Time makes a loss less likely but never impossible, and it widens the range of results. So the first question about any sum isn’t which place pays most, but when the money will be needed.

Try this

  • Write two lines: “Money that will be needed in the next three years” and “Money that can wait longer”. Put a number on each, for yourself or for a made-up family.
  • Take any fall in a price or an index in the news and work out the rise that would undo it: divide 100 by what is left after the fall. A fall of 20% leaves 80, and 100 ÷ 80 = 1.25, a rise of 25%.
  • Find NEPSE’s index today on the Nepal Stock Exchange’s site and compare it with the readings in this chapter. Is it above or below July 2021’s?

Found a mistake? Please tell us.