Hedging Your Bets

Imagine two friends, Aarvi and Aman, find the same lucky coin. It's a special coin: when you flip it, it lands heads a little more often than tails - say 55 times out of 100 instead of a fair 50. Someone runs a game: bet some of your marbles on heads, and if heads shows you win as many marbles as you bet; if tails shows you lose what you bet. Both friends know the coin is good. Both are right about that. They have found a genuinely good bet.

Now watch what happens next, because this is the whole chapter in one picture. Aarvi bets a small handful of her marbles each round. Aman, excited that the coin is lucky, bets all his marbles every single round. They are playing the exact same game with the exact same good coin. And yet after an evening of flipping, Aarvi has slowly, steadily grown her pile - while Aman, at some point, hit a tails when everything was on the table, lost the lot, and had nothing left to bet with. Game over. Not because he picked a bad coin. He picked a great coin. He was ruined by how much he bet.

That's the idea we're going to sit with, and it surprises almost everyone. When people think about investing, they think it's all about the what - which company, which coin, which bet. They imagine that if you just pick well enough, the money takes care of itself. But there is a second question, hiding quietly behind the first, that matters every bit as much: how big should this bet be? You can be completely right about the coin and still lose everything if you size the bet wrong. How much you stake matters as much as what you stake it on - a good bet made too big can ruin you just as surely as a bad bet. Picking is half the game. Sizing is the other half, and it's the half almost nobody thinks about.

Let's slow down and really feel why Aman lost with a winning coin, because it isn't obvious and it's the beating heart of everything that follows.

Here's the trap. When you bet everything, you only need to lose once to be finished. It doesn't matter how good the odds are on each single flip. A 55%-heads coin still lands tails 45 times out of 100 - nearly half the time. So a tails is not some freak event you can pretend won't happen; it's coming, probably soon, and certainly within an evening of flipping. And the moment it lands while all your marbles are on the table, you go to zero. From zero, the lucky coin can't help you anymore, because you have nothing left to bet. The good odds are still there - but you're not.

This is the cruel thing about losing everything as opposed to losing something. If Aarvi bets a small amount and loses, she's a little poorer but still in the game; the coin is still lucky; she plays on and tends to climb back. But if you lose it all, there is no "play on." You've walked through a one-way door. On one side you can still recover; on the other, you simply can't, ever, no matter how good the bet was. Losing some money is a setback you recover from; losing all of it is a trapdoor you never climb back through - the two are different in kind, not just in size.

Grown-ups have a name for that trapdoor: ruin. And the whole point of sizing your bets carefully is to make sure that no single flip - no single bad month, no single company blowing up - can ever push you through it. A bet that could take everything, even with wonderful odds, is a bet you must never make at full size, because you don't play the game once. You play it flip after flip, year after year, and to keep playing you have to still be standing tomorrow. The average flip being in your favour is no comfort at all if one bad flip removes you from the table.

So if betting everything is madness, should you bet almost nothing? That doesn't work either - and the reason reveals that there is a sweet spot in the middle, which is the real secret of bet sizing.

Think about the two ends. If you bet far too much, you grow fast on good runs but you eventually hit the bad run that ruins you, and ruin cancels everything. If you bet far too little - a single marble each round out of a thousand - you're safe as houses, but your pile barely grows; the lucky coin is wasted on you because you never let it do any work. Neither extreme is smart. Somewhere between "so small it's pointless" and "so big it's suicidal" there is an amount that grows your pile the fastest without risking the trapdoor. That best amount depends on one thing above all: how big your edge is - how lucky the coin actually is.

A coin that lands heads 55 out of 100 deserves a bigger bet than a coin that lands heads only 51 out of 100, because your advantage is larger and steadier. A tiny edge deserves a tiny bet. A large, reliable edge earns a larger one. The size should follow the strength of your advantage, like the size of your umbrella following how hard it's actually raining. Bet bigger than your edge justifies and you drift toward ruin; bet smaller and you leave growth on the table. The craft is matching the bet to the edge.

Notice the shape of that hill. Going up the left side, betting a bit more helps you - more edge put to work, faster growth. But once you pass the peak, betting more hurts you, and it hurts faster and faster until, past the ruin line, the whole thing collapses. So "bet more" is good advice only until the peak, and terrible advice after it. The single most expensive mistake in this whole game is to be standing on the right side of that hill, feeling bold, thinking you're being aggressive and clever - when really you've walked past the peak and you're marching toward the cliff.

Let's put real numbers on the marble game and watch the sweet spot appear on its own. illustrative

Three friends each start with ₹1,00,000 and play the same lucky-coin game: a bet that pays back double on a win and loses your stake on a loss, and wins 55 times out of 100. It's a good bet - over many flips, the average flip makes money. Each friend picks a sizing rule and sticks to it, flip after flip.

• Aman goes all in. Every round he bets his whole pile. On a good streak he looks like a genius - ₹1,00,000 becomes ₹2,00,000 becomes ₹4,00,000, and he laughs at the other two. But heads-55 still means tails-45, so a tails is never far away. On the flip it lands - and it will - his entire pile vanishes. From ₹0 he cannot recover; the lucky coin is useless to a player with nothing to bet. Aman's most likely ending, played out again and again, is ₹0.

• Vikram bets a tiny sliver - just ₹1,000 a round, terrified of losing. He is perfectly safe; ruin can't reach him. But after a long evening his pile has barely nudged from ₹1,00,000 to maybe ₹1,08,000. The coin was lucky and he mostly wasted it, because he never let his edge do real work. Safe, but crawling.

• Aarvi sizes to the edge. She bets a sensible, modest fraction of her pile each round - enough that a good coin grows her meaningfully, small enough that no single tails can knock her out. When she loses, she's betting a fraction of a smaller pile next time, so she bends but never breaks. Round after round her pile climbs, unevenly but surely, and by the end of the evening she is comfortably ahead of both friends - far past crawling Vikram, and still in the game, unlike vanished Aman.

Look at what decided the winner. It was not who understood the coin best - all three knew it was lucky. It was who sized best. Aman was ruined by too much, Vikram was starved by too little, and Aarvi won by finding the middle. The lesson lands hard precisely because the coin was identical for everyone: with the very same good bet, the sizing rule was the difference between growing rich, crawling along, and going broke.

The marble game had one fixed coin. Real investing is messier: some of your bets are strong and some are weak, and the size should follow how good each one is. Let's watch a careful investor size two different bets. illustrative

Aayra has a portfolio of ₹10,00,000. Two ideas cross her desk in the same month.

The first is a plain, boring company she understands deeply - steady profits for years, little borrowing, honest owners, and a price low enough that she can see a real, defensible reason it's worth more than it costs. Her edge here is large and reliable. This deserves a proper bet, so she puts in ₹1,00,000 - a tenth of everything. Meaningful, but survivable if she's wrong.

The second is a tip about a company she barely understands - an exciting story, but her "edge" is really just a hunch, and she couldn't write down a solid reason it's mispriced. Her edge here is tiny and shaky. An honest sizing rule says: tiny edge, tiny bet. So even though the story is more thrilling than the boring company, she puts in far less - ₹20,000 - precisely because she's less sure. The excitement of a bet is not its edge. Often the more thrilling a bet feels, the less you actually know, and the smaller it should be.

See the discipline in that. Aayra didn't size by how excited she felt; she sized by how strong and certain her advantage was. The strong, dull bet got five times the money of the thrilling, flimsy one - the exact opposite of what her feelings wanted. And notice the quiet safety net: even her biggest single bet is only a tenth of her pile, so if the boring company somehow blows up, she loses ₹1,00,000 and still has ₹9,00,000 to play on. No single idea, however good it looks, is ever allowed to be big enough to ruin her. That's what "size by edge" means in practice: bigger when you're genuinely more sure, smaller when you're not - and never so big that being wrong ends the game.

There's a second quiet benefit hiding in Aayra's rule, and it's worth pulling into the light. Because each of her bets is only a slice of the whole, she can hold several of them at once - the boring company, the flimsy tip, and a handful of others - and a disaster in any one of them is just a dented tenth, not a shattered whole. Spreading her pile across many modest bets means she never has all her hope riding on a single flip of a single coin. Compare that with an investor who, feeling very sure, pours eight-tenths of everything into one "obvious winner." That person has quietly signed up for the all-in trap, dressed up as conviction: if the one company stumbles, there is no second bet standing to carry them. Sizing small enough to hold many bets isn't only about surviving each one - it's about never letting your whole future depend on being right about any single thing, because being sure and being right are not the same, and the gap between them is where ruin lives.

Now the deeper cut, and it's the part that separates people who merely know about sizing from people who actually survive. There's a clever bit of maths that can tell you the exact "perfect" bet size to grow a pile fastest - the very peak of the sizing hill. You might think: great, just bet that perfect amount every time. But the wise investor deliberately bets less than the perfect amount. Why on earth would you aim below the peak on purpose?

Because the "perfect" amount is calculated from your edge - and you do not actually know your edge exactly. You estimate it. When Aayra says the boring company is worth more than its price, that's her honest best guess, but it's still a guess. Maybe her edge is smaller than she thinks. Maybe the world changed and it's gone entirely. And here is the killer detail: the sizing hill is lopsided around its peak. Bet a little less than perfect, and you give up only a sliver of growth - you're just below the top, barely poorer. But bet a little more than perfect, and you fall down the steep right side toward ruin far faster than you'd expect. Overshooting is punished brutally; undershooting is punished gently.

So the safe move, when your edge is only a guess, is to aim below the peak on purpose - to bet, say, a quarter or a half of the "perfect" amount. You surrender a little growth in exchange for a big cushion against being wrong about your own edge. Put concrete numbers on it: suppose the maths says Aayra's boring company "perfectly" deserves ₹4,00,000 of her ₹10,00,000. Because her edge is an estimate and the right side of the hill is a cliff, she quietly bets a quarter of that - ₹1,00,000 - and sleeps soundly. If her edge was real, she still grows nicely. If her edge was half what she thought, ₹1,00,000 barely stings, while ₹4,00,000 could have been the bet that broke her. Because your edge is a guess and over-betting is punished far more harshly than under-betting, deliberately bet a fraction of the 'perfect' size - you trade a little growth for a lot of survival. Betting below perfect isn't timidity. It's respect for the fact that you can be wrong about how good your bet is.

There's one more force that ruins good, smart people - and it works entirely by pushing your bet size past the peak without you noticing. It's leverage: betting with borrowed money. Let's watch it happen in rupees. illustrative

Rohan has ₹1,00,000 and a genuinely sound idea - a company he understands, fairly priced. On its own, his ₹1,00,000 bet sits comfortably near the safe peak of his sizing hill. Then a helpful voice offers to lend him another ₹1,00,000 so he can bet ₹2,00,000 instead. If he's right, he wins on twice the money - thrilling. So he takes the loan.

Here's the sneaky thing borrowing does to the hill: it silently pushes his bet size to the right. Rohan still thinks he's making a ₹1,00,000-sized decision, but the lender has quietly turned it into a ₹2,00,000-sized bet. He was sitting nicely near the peak with his own money; the loan has shoved him past it, onto the steep, dangerous right side - and his edge hasn't improved by a single rupee. Same company, same odds, just a bet that's now too big for his pile.

Now the truly dangerous part, and here's where the rupees bite. When you borrow, you don't only need to be right about the company - you also need to survive until you're proven right, because a lender can demand their money back at the worst possible moment. Say the good company simply has an ordinary rough patch and the price dips 40% for a few months before recovering. With his own ₹1,00,000, Rohan would shrug: his stake is worth ₹60,000 for a while, he waits, it comes back. But on the ₹2,00,000 borrowed position, that same 40% dip is a ₹80,000 loss - and since ₹1,00,000 of the money was the lender's, Rohan's own ₹1,00,000 has been more than halved, tripping the loan's safety rope. The lender forces a sale at the bottom to get their ₹1,00,000 back. Rohan walks away with almost nothing - wiped out by a dip he correctly predicted would recover, on a company that then went on to do fine without him. The borrowed bet didn't just lose more; it ended his game on a day that, un-borrowed, would have been a mere bruise. That's the trapdoor from before, only leverage moved it right under his feet.

Think back to Aman losing everything on the lucky coin. Leverage is Aman's mistake wearing a respectable suit. A person betting all their own marbles at least feels reckless. A person borrowing to bet ₹2,00,000 on a "safe, sensible" company feels responsible - grown-up, even. But the maths is the same shape: the bet has been sized past the point where one bad flip can remove you. The lucky coin can't save a player with nothing left, and the good company can't save an investor who was forced to sell at the bottom to repay a loan.

The slip is almost never "I picked a terrible company." It's "I was so sure I was right that I bet too big." Confidence, not ignorance, is what pushes people past the peak of the hill.

Here's how it works on you. You do careful work, you find a genuinely good bet, and you feel - correctly - that it's good. But then a quiet, dangerous jump happens in your head: this bet is good, so I should put a lot into it. The bigger your conviction, the bigger the bet grows, until one idea is half your pile, or you've borrowed to make it bigger still. And the whole time it feels like strength, like backing your judgement, like the opposite of cowardice. That feeling is the trap. Sizing by how sure you feel is exactly how good investors blow up, because feelings run hotter than edges, and the hottest feelings often sit on the flimsiest facts.

Now the honest cautions, because "size carefully" can be pushed until it, too, becomes a mistake.

The first misreading is to think this chapter says bet small and you're safe, always. Not quite. Remember Vikram, who bet a tiny sliver and barely grew: betting too little is also a way to lose, just a slow and quiet one. A pile that crawls while prices rise around it is losing ground in real terms, year after year. The goal was never "bet as little as possible." It was to find the right size for your edge - big enough that a real advantage does real work, small enough that no bad flip can ruin you. Over-caution isn't a virtue here; it's the left side of the hill, and the left side still leaves growth on the table.

The second misreading is subtler and more important: all this careful sizing only helps if you actually have an edge. The lucky coin worked because it was genuinely lucky. If you size "by your edge" but you don't really have one - if the "good bet" is a hunch dressed up as analysis - then perfect sizing just makes you lose more slowly and more precisely. Sizing is the second half of the game, not the first. You still have to pick real bets. This chapter tells you how much to put on a good bet, not how to tell a good bet from a bad one; both jobs matter, and getting the sizing exactly right on a bad bet is no achievement at all.

And a last, quiet caution: the "perfect" bet size from any formula is only as trustworthy as the numbers you feed it. Feed it an edge that's too rosy, or forget that a bad year can be worse than any you've seen, and the maths will happily point you straight past the peak while looking rigorous. That's exactly why the wise move is to bet a fraction of what the formula says and to leave a wide margin for being wrong. The point of sizing isn't to squeeze out the last drop of growth. It's to make absolutely sure you're still holding marbles tomorrow - because the player who's still at the table is the only one who gets to keep playing the lucky coin.