Skewness and Asymmetry

Picture two games at a school fair, each with its own little stall.

At the first stall, you pay nothing to play, and almost every single time you win a small prize - a ₹2 toffee. Play ten times and you'll walk away with a toffee nine times out of ten, grinning. It feels wonderful. You win, and win, and win. But there's a tiny rule printed at the bottom of the sign in faint letters: on the rare tenth try, instead of a toffee, the stall takes ₹100 from your pocket.

At the second stall, it's the opposite. Almost every time you play, you lose a small ₹5 coin. Play ten times and you'll lose nine of them, feeling glum and foolish. But once in a while - that rare tenth try - the stall hands you a fat ₹500 note.

Now here's the question that this whole chapter is really about. If you could only stand at one stall all afternoon, which one makes you richer by evening?

Most people's hearts pull them toward the first stall, because winning feels like winning and losing feels like losing, and our feelings count how often we're right. But if you're patient and actually add up the rupees, the first stall quietly empties your pocket and the second stall quietly fills it. Being right nine times out of ten at the first stall isn't good enough, because the one time you're wrong is enormous. Being wrong nine times out of ten at the second stall isn't bad at all, because the one time you're right is huge.

That is the surprising heart of this chapter. When it comes to money, the thing that matters is not how often you turn out to be right. It's how much you make when you're right, weighed against how much you lose when you're wrong. A person can be right almost every day and still go broke. A person can be wrong most of the time and still grow rich. What decides it isn't the score of right-versus-wrong; it's the size of each outcome multiplied by how likely it is. What decides whether you win over time is the size of each gain or loss times its chance of happening - not the simple count of how often you turn out to be right.

Why does this feel so upside-down? Because from the time we're small, we're trained to keep score by counting correct answers. Ten out of ten on a spelling test is a perfect day. Getting most of the sums right earns the gold star. So our minds quietly decide that "being right most of the time" is the same thing as "doing well." In a classroom, it usually is.

But money doesn't keep score that way. Money keeps score in rupees. And rupees don't care one bit how many times you were right - they only care how much was at stake each time. The market never hands out a gold star for a high hit-rate. At the end of the year, nobody asks, "How often were you right?" The only question that pays your bills is, "How much more money do you have than you started with?"

This gap between the two scoreboards - the count scoreboard our feelings use, and the rupees scoreboard reality uses - is where an enormous amount of trouble hides. A strategy that scores beautifully on the count scoreboard ("I was right 95% of the time!") can be a disaster on the rupees scoreboard, if that shameful 5% is where all the big losses live. And a strategy that scores terribly on the count scoreboard ("I was wrong most months") can be wonderful on the rupees scoreboard, if the rare wins are big enough to carry all the small losses on their backs.

Once you really feel this, a strange freedom arrives. You stop needing to be right all the time. You stop being embarrassed by lots of small wrongs. You start asking a completely different, far better question about any bet in front of you - not "Am I likely to be right?" but "When I'm right, how big is the reward, and when I'm wrong, how big is the damage?" That single change in the question is the whole lesson, and everything below is just careful practice at asking it.

Let's slow right down and build the actual sum, because it's simpler than it sounds and, once you can do it in your head, you'll never look at a bet the same way again.

For any bet, there are usually two things that can happen: it goes well, or it goes badly. To judge the bet honestly, you don't look at either outcome alone. You look at both, and for each one you multiply how big it is by how likely it is. Then you add them up. That total - grown-ups call it the "expected value," but let's just call it the honest average - tells you what the bet is really worth if you could play it many, many times.

Let's do it for our two fair-stall games, in plain rupees.

The first stall (win small often): nine times out of ten you win ₹2, and one time out of ten you lose ₹100. So the honest average per play is: (₹2 × nine-in-ten) plus (minus ₹100 × one-in-ten). That's ₹1.80 minus ₹10, which comes to minus ₹8.20 every time you play. Every cheerful little win is quietly costing you money, because the rare big loss is far heavier than all the small wins put together.

The second stall (lose small often): nine times out of ten you lose ₹5, and one time out of ten you win ₹500. Honest average: (minus ₹5 × nine-in-ten) plus (₹500 × one-in-ten). That's minus ₹4.50 plus ₹50, which comes to plus ₹45.50 every time you play. Every glum little loss is actually a bargain, because you're buying a chance at a prize that dwarfs all those losses.

Look what just happened. The stall that wins almost every time is a slow leak, and the stall that loses almost every time is a quiet fountain. The count of wins pointed you at exactly the wrong stall. The honest average - size times chance - pointed you at the right one. This little sum is the referee that settles every argument in this chapter. Whenever your feelings shout "but I'm right so often!", make them sit down and do the sum.

Let's leave the fair and put real savings on the table, so you can feel this in rupees over a year rather than a game. illustrative

Meet Rohan. He has found what he proudly calls a "steady" way with his money. Every month he does something that, almost every month, gives him a small, reliable gain - think of it as a strategy that collects a little bit whenever markets are calm, which is most of the time. Month after month, a tidy ₹3,000 lands in his account. He keeps a diary, and in it there's a long, satisfying row of ticks: won, won, won, won. Eleven months in a row, ₹3,000 each. By November he's ₹33,000 up and telling everyone he's cracked it. His hit-rate is a glorious eleven out of eleven.

His cousin Aayra does the opposite kind of thing, and honestly she looks a bit silly next to him. Her method loses a small amount most months - a steady ₹2,000 drip out the door - while she waits for the rare, wild month when markets lurch and her position pays off big. Her diary is a miserable row of crosses: lost, lost, lost, lost. Eleven months in, she's ₹22,000 down. At every family lunch, Rohan is winning and Aayra is losing, and everyone can see it.

Then December comes, and December is strange. A sudden shock hits the market - the kind of month that arrives once in a while and can't be timed. In that one violent month, Rohan's "steady" method, the one that quietly collected ₹3,000 in all the calm months, turns around and bites: it loses him ₹90,000 in a single stroke. His eleven neat wins are swept away and then some. He ends the year down about ₹57,000, staring at eleven ticks and one cross that undid all of them.

That same wild December is exactly what Aayra was patiently paying ₹2,000 a month to wait for. Her position pays off ₹80,000 in that one month. After the eleven small losses, she ends the year up about ₹58,000, with eleven crosses and one glorious tick.

Now put the two diaries side by side. Rohan was right eleven times and ended the year poorer. Aayra was wrong eleven times and ended the year richer. If you'd judged them any month before December - by the count of wins, by who looked clever at lunch - you'd have backed Rohan every time, and you'd have been backing the leak over the fountain. The wins Rohan collected were real but tiny; the loss he took was rare but gigantic. The size beat the frequency. A long row of small wins can hide a strategy that is losing money, because the honest average is decided by the size of the rare big move, not by the count of the little ones.

The first example warns you off a trap. This one shows the good side of the same idea - how being wrong most of the time can be a wonderful way to invest, if your wins are shaped the right way. illustrative

Meet Arjun, who has ₹1,00,000 he can afford to be patient with. Instead of pouring it all into one thing that he hopes is right, he spreads it thin across ten small, cheap, long-shot ideas - ten tiny, early businesses, ₹10,000 into each. He knows, going in, that most of these will fizzle out. He isn't fooling himself that he's a genius stock-picker. He's arranging his bets so that each one can only lose a little, but a rare one might win a lot. That shape - small known loss, huge possible gain - is the thing he's actually buying.

Fast-forward five years and add up the wreckage honestly. Of the ten, six go to zero. Gone: ₹60,000 vanished. Three others just limp along and, when he finally sells, give back roughly what he put in: ₹30,000 in, ₹30,000 out, nothing gained. So on nine of his ten ideas, Arjun was "wrong" - six disasters and three duds. A ninety-percent miss rate. On the count scoreboard, he's a fool.

But the tenth idea - one small ₹10,000 bet - turns into a real business. It multiplies thirty times over those five years and is worth ₹3,00,000. Now tally the whole basket: ₹0 from the six dead ones, about ₹30,000 back from the three duds, and ₹3,00,000 from the single winner. That's ₹3,30,000 in hand from the ₹1,00,000 he started with - more than tripled - even though he was flat-out wrong on nine bets out of ten.

Here's the machinery to notice. Each losing bet could only ever cost Arjun ₹10,000 - the loss had a floor. But the winning bet had no ceiling; it was free to run to thirty times. When your losses are capped small and your wins are free to grow huge, you don't need to be right often. You need to be right big, even just once in a while. Arjun didn't get rich by picking well. He got rich by arranging his bets so that being wrong was cheap and being right was enormous. That is the honest-average sum working in your favour instead of against you. When each loss is small and capped but each win can grow without a ceiling, a low hit-rate can still be richly profitable, because one large win outweighs many small losses.

Now for the most dangerous version of all this, because it's the one dressed up to look the safest. It's the trap of the strategy that pays you a small, steady "income" - until, one rare day, it doesn't just stop paying but takes everything back at once. illustrative

Let me build the shape with a plain picture first. Imagine a boy who rents out his good cricket bat to friends for ₹5 a game. Coins come in, game after game - a lovely little income. It feels like free money falling from the sky. What he isn't counting is the rare game where a friend swings too hard and cracks the bat clean in half. On that one day, he doesn't lose ₹5. He loses the whole bat - worth, say, ₹800 - and with it every future ₹5 he was going to collect. All those neat little coins were, without his noticing, tiny down-payments the world was making toward one big bill it would hand him later.

Now the grown-up version, in rupees. Meet Aarohi. She sells other people a kind of protection - think of it as promising, for a fee, to cover someone's loss if the market suddenly falls hard. Because markets are calm most of the time, this feels like magic. Every month, buyers pay her a fee - a comfortable ₹4,000 a month - and almost every month nothing bad happens, so she simply keeps it. She does this for twenty months. Twenty times ₹4,000 is ₹80,000, and it arrives so regularly that she starts thinking of it as a salary. Her diary, like Rohan's, is a proud row of ticks. Her friends ask how she found such easy, steady income.

Then, in month twenty-one, the rare thing happens. The market drops violently - the once-in-a-few-years lurch that everyone forgets is even possible during the long calm. Now all those promises she sold come due at once, and she has to pay out. In that single month she owes ₹3,00,000. Twenty months of tidy ₹4,000 income - ₹80,000 in all - is wiped out four times over. She's not just back to zero; she's ₹2,20,000 in the hole, and worse, she never saw it as a risk, because for twenty months the diary said nothing but "won."

This is the cruellest shape in all of investing, because it turns our count-scoreboard instinct into a weapon against us. Twenty ticks in a row don't just feel safe; they actively lower your guard, tempt you to put in more money, and convince your friends to copy you - right before the one month that matters. The steady income was never really income. It was a slow collection of pennies laid down in front of a steamroller nobody could hear coming. A strategy that pays a small, steady sum most of the time can be quietly storing up one rare, violent loss that takes back all the little gains at once - so treat suspiciously smooth 'income' as a bill not yet delivered.

And notice how this is the exact mirror of Arjun. Arjun paid small steady losses to buy a rare huge win - a good shape. Aarohi collected small steady wins while selling away a rare huge loss - the same shape, flipped, and it's the flip that ruins you. The question to ask of any "steady income" is always: what am I quietly selling to be paid this smoothly, and what happens on the day it's called in?

Step back from single bets and look at a whole lifetime of them, because there's one more layer that makes all of this even more important than it first seems.

When you make many, many bets over years, they don't share the credit equally. It's almost never the case that each bet chips in a fair, even slice of your final result. Instead, a tiny handful of extreme outcomes - the rare enormous winners, and the rare catastrophic losers - end up deciding nearly the entire story. The vast, boring middle of your bets, the ones that did roughly nothing, barely move the needle. The ends do the work. Grown-ups call these rare extreme outcomes the "tails," and the plain truth is that the tails, not the average day, drive where you end up. Over a long run of bets, a very small number of extreme outcomes produce almost all of your final result, so most of your bets doing little is completely normal and not a sign of failure.

Let me make it concrete. illustrative Suppose Haridya makes twelve careful, patient investments over ten years. When she tallies them at the end, here's roughly how they land: five went nowhere and gave back about what she put in. Four lost small amounts. Two did nicely - a comfortable double each. And one turned into a monster, multiplying twenty-five times over the decade. When she adds up her whole fortune, almost all of it traces back to that single monster bet. The five flat ones and four small losers, all nine of them together, are a rounding error next to it. Nine out of twelve "didn't work," and it didn't matter in the slightest.

This is why the whole chapter fits together. Because a few extreme outcomes decide everything, your single most important job is arithmetic on the extremes. On the losing side, make sure no single rare loss can be so big it ends the game - that's the Aarohi lesson. On the winning side, own things where a single rare win can grow without a ceiling - that's the Arjun lesson. Look after the tails, and the boring middle can take care of itself.

The slip is almost never greed for a wild gamble. It's the opposite - it's the comfort of the smooth, steady win, and the quiet pride of being right a lot.

Here's how it gets you. You find something that pays a little, regularly, and it works. Month after month, tick after tick. Being right so often feels like skill, so you trust it more, and - this is the dangerous step - you put more money behind it precisely because it's been so reliable. Your friends see your steady run and copy you. Everyone leans harder on the thing exactly as it collects the most pennies, which is exactly the moment before the steamroller. The very smoothness that should make you suspicious is the thing that makes everyone relax. We are wired to trust what has been calm and right; the market is happy to be calm and right for a long time before it isn't.

Now the honest cautions, because this idea, like any sharp tool, can cut you if you swing it carelessly.

First, "prefer cheap bets with rare huge upside" is not a licence to throw money at any long-shot with a thrilling story. The Arjun example only worked because two boring things were true: each loss was genuinely small and capped, and the winner had a real, honest path to growing huge. Plenty of long-shots have neither - they can lose more than you expect and their upside is a fantasy. If you buy lottery-shaped bets that are simply overpriced, you get all the losing and none of the winning. The shape has to be real, not just exciting. A rare huge upside you overpaid for is just a slow, expensive way to lose.

Second, being wrong most of the time is only fine if you've made sure your losses stay small. Arjun could be wrong nine times out of ten because each wrong cost him ₹10,000 and no more. If even one of his "small" losses had secretly been unlimited - the Aarohi trap in disguise - the whole method would collapse. So the low-hit-rate approach depends entirely on the losses being truly capped. Check that floor before you comfort yourself with "I only need one to work." A strategy that's wrong most of the time and has uncapped losses is not brave; it's just a faster road to ruin.

Third, don't take "tails drive everything" as an excuse to stop thinking, hold on to hopeless bets forever, or refuse to ever admit a mistake. "One of these will be the monster" is a lovely comfort to whisper to yourself while a genuinely dead bet drains more money. The idea says a few extremes decide the outcome - it does not say every one of your losers is secretly a future monster. You still have to tell an early, promising business from a broken one, and let the truly dead ones go. The tails do the work, but only if you keep planting real seeds and pulling out the rotten ones - not if you water every dead twig hoping it's the giant.

And a final, quiet caution: none of this lets you know in advance which bet is the monster or when the steamroller arrives. This chapter helps you shape your bets so the extremes are on your side - losses capped, wins uncapped, no single catastrophe possible. It does not hand you a crystal ball. You arrange the odds well and then you stay in the game long enough for the rare, decisive outcomes to show up. That's the whole trick, and it's humbler than it sounds: not being right often, just being shaped right and alive when it counts.