The Contrarians Playbook When to Exit | Rajeev Thakkar | ET Alpha Wealth Summit

ELI5/TLDR

Investors obsess over when to buy. The real discipline is knowing when to hold and when to sell. Rajeev Thakkar, CIO of PPFAS, walks through six reasons NOT to sell (boredom, chasing news, booking profits, trailing returns, buyer’s remorse, option envy), then lays out five legitimate reasons to exit: you need the cash, fundamental error, business is dying, valuations are absurd, or a better opportunity came along. The throughline: low portfolio churn compounds wealth; high churn feeds brokers.

The Full Story

The Asymmetry of Selling

Most investors spend years learning to buy. Selling is lonelier, harder, and rarely discussed. Thakkar frames the discipline in three phases: learning to identify good investments, learning to stay invested (arguably the bigger part), and finally learning when to walk. The gap matters because a buy mistake costs you capital; a sell mistake costs you compounding.

He opens with a provocative ask: if you have one crore and want to become a dollar billionaire (at a 100-rupee exchange rate = 10,000 crores), you need 14 doublings. That’s the stakes. Every stock you sell prematurely is a missed doubling; every stock you hold through a drawdown is psychological torture.

The Six Terrible Reasons to Sell

Boredom. You’ve done the work—met management, read the reports—and bought at a good price. Six months of sideways movement, and you’re restless. So you sell and move elsewhere. Thakkar is blunt: investing isn’t entertainment. If you want entertainment, go to Vegas. Investing is for wealth creation, not dopamine hits.

Chasing News. This is where fortunes evaporate. In February 2020, a tech investor with inside knowledge heard (via a congressional contact) that COVID would lock down the world. He sold his tech positions before the crash—correctly. But after the crash, when everyone feared further falls, he didn’t buy back. Tech was the biggest winner of lockdowns. His prescient news call ended up costing him everything he exited.

The lesson: every news stimuli you follow = a potential sell decision made on incomplete information. Most institutional investors filter through dozens of signals daily. Retail investors can’t. If you follow the news like a sword, you’ll cut yourself.

“Booking Profits.” A stock goes from 100 to 150. You feel the itch: I’m already in profit, let me lock it in. Thakkar calls out the phrase itself as propaganda invented by finance ministers and stock brokers. The only winners from periodic profit-booking are the finance ministry, brokers, and Ram Mohan (the exchange). Exchanges and discount brokers have been among the best-performing “stocks” precisely because investors churn. Retail investors, by contrast, haven’t made money in two years—not despite high activity, but because of it.

The car analogy: 95% of driving is looking forward. You glance in the rear-view mirror only when changing lanes. In investing, people spend 95% of their time looking backward at returns, projecting them forward, and panicking. The 2024 equity bubble happened because people extrapolated 3–4 years of returns into infinity. Now they cry.

Trailing Returns. This stock is down 5% over two years; it’s a poor performer, time to exit. Wrong. The index had flat years. Some companies have sideways periods. Hindustan Unilever spent 6–7 years sideways. Infosys took 7 years (2000–2007). Even recent darlings like BAT took 7–8 years. In a diversified portfolio, some holdings will be flat while others boom. What matters: do you beat the index post-tax over 10–20 years? If not, move to an index fund. If yes, stop beating yourself up about individual sideways positions.

Buyer’s Remorse. You buy a shirt, see someone else’s shirt, think you chose wrong. Same with houses, cars, stocks. A stock you own looks stale; another stock looks shinier. So you sell to chase the new one. The problem: if you’re in a good business with decent valuation, long-term prospects, excellent management, and huge TAM, you’re in a rocket ship. These come 3–5 times in a career. Just hold. Don’t ditch a rocket for a distant light.

Valuation Irrationality. This is the only reason Thakkar says feels legitimate, but he adds a caveat: it’s the hardest to time. Selling when “valuations are absolutely nuts” is like saying you’ll time the peak. A querent mentions buying Nvidia at $10 (split-adjusted 2019) and watching it rise 3,000–4,000%. Now it’s 30% of his portfolio, valued at 33x earnings, and he doesn’t know if he should hold. Thakkar’s answer is brutal: if one stock becomes 100% of your portfolio value (because it went up 1,000x while others flatlined), you’ll lose sleep. Not a way to maximize wealth. A way to not lose your mind.

His practical rule: trim gradually. If a position is at 70–80x earnings and you feel it’s overcooked, sell 10% per month, letting it fall to a comfortable weightage. At least your original cost comes out, plus gains. You keep a piece of the upside but don’t bet the farm.

The Five Legitimate Reasons to Sell

You Need the Cash. Family office needs capital for an acquisition in a familiar sector? Sell something. This is the cleanest reason to exit—and where people have maximum emotional resistance. They average down instead, hoping the problem goes away. The liquor baron’s airline was the cautionary tale. If you’ve made a fundamental error in due diligence (miscalculated entry price, business model flawed), exit at the first chance. Capital locked in bad ideas can’t chase good opportunities. Bonus: a loss can offset other income for tax purposes. Take it.

Fraud. If you discover fraud, exit immediately.

Business Model Disruption. When the internet arrived, should you exit newspapers? Renewables vs. thermal power. EVs vs. petrol cars. These aren’t yes-or-no calls. They’re maybes. The key: you can’t decide this once and forget. Even not deciding is a decision to stay invested. Every 3–6 months, especially in listed securities, you have to revisit. Thakkar isn’t invested in newspapers, but he is invested in fossil fuels (believing they’ll matter 10–20 years). Is disruption in 5 years possible? Yes. So he monitors. He reads evidence. He recalibrates. Kodak was great until digital cameras killed it. No amount of past success insulates you from industry death.

Absurd Valuations. A pharma founder built a huge generics + outsourcing franchise in India, then got an offer from an American buyer at compelling valuations. He sold. Made sense. Early Infosys or Wipro investors in the mid-90s could have exited in early 2000s at prices that baked 15–20 years of future growth into valuations. At some point, the price is so divorced from reality that even a great company should be sold. Exit, pay taxes, hunt for better opportunities.

Better Opportunity, No Cash. You own a reasonably valued stock. Something comes along that’s 5x better. You have no dry powder. Sell the first one, pay the taxes and brokerage, and deploy into the new opportunity.

The Portfolio Lens

Here’s where Thakkar pivots from theory to reality: if you’re in India and paying taxes, you want low churn. Institutions and people in zero-tax jurisdictions (Dubai, etc.) can play rummy—evaluate every hand and trade. You can’t. Every trade is a 15% tax friction (in India’s structure). High activity starves you.

The north star: post-inflation, post-tax returns of 7%+. If you’re not beating 7%, something’s wrong. You’ve already won; you don’t want to lose it all in a blowup (Buffett’s line: “It is foolish to risk what you have and need to get what you don’t have and don’t need”). Capital preservation comes first. Growth comes second.

Diversification matters because most people make wealth from one sector, one employer, one company. Their net worth is concentrated. In your investment portfolio, diversify. Don’t take concentrated bets there. Salaried employees have NPS, EPFO, and PPF—tax-efficient compounding vehicles. The rich have mutual funds and closed-end funds (PPF equivalent for HNIs). As long as you don’t redeem within a fund, taxation is exempt. Use that. Stay put. Churn every 18 months and you’ll pay 15% every time. That kills returns.

Key Takeaways

Boredom is not a reason to sell. Sideways movements happen in every great stock. Infosys, HUL, and BAT all had years of flatness. Compound discipline beats entertainment addiction.
News is noise on a daily timescale. Someone with advance COVID intelligence sold tech in February 2020 and missed the entire rally. His edge was real; his decision cost him everything. Don’t chase every stimulus.
“Booking profits” is jargon invented by finance ministers and brokers. It’s a sign that you’re churning for the wrong reasons. The only winners are the exchanges and brokers.
The rear-view mirror is a dangerous compass. Extrapolating past 3–4 years of returns into the future is how bubbles form. Look forward, not backward.
Sideways is not death. Holding through flat years is harder than selling, but necessary. Test yourself: beat the index post-tax over 10–20 years? Keep going. Underperforming? Switch to an index fund.
Cutting losses is underrated. People who are disciplined about exiting mistakes outperform people who are great stock-pickers but hold forever. Capital locked in dud ideas can’t chase opportunities.
Trim giant positions gradually, not in panic. When one stock becomes 20–30% of your portfolio (and you’re uncomfortable), sell 10% per month. Your original cost comes out. You keep the upside. You sleep.
Valuations matter, but they’re fuzzy. 70–80x earnings = probably expensive. 15–20x earnings of future growth already baked in = extremely expensive. 33x earnings with 30%+ growth = maybe defensible. Use heuristics, not precision.
Disruption is real, but monitor it. Fossil fuels probably matter 10–20 more years, but could be disrupted in 5. Keep reading evidence. Recalibrate every 6 months.
Tax drag is real and underestimated. Low churn + tax efficiency = the difference between good and great returns. In India, high activity kills post-tax wealth.
Post-inflation, post-tax returns of 7% = the bar. If you’re not clearing it, you’re either taking too much risk or not investing well. Both are problems.

Claude’s Take

Thakkar is refreshingly clear because he doesn’t pretend certainty exists where it doesn’t. He’ll say “fossil fuels probably matter 10–20 years, but maybe not”—and then act on that probabilistic view by monitoring evidence. He doesn’t demand binary answers to fuzzy questions.

The talk is strongest on why not to sell. Boredom, news-chasing, and trailing returns are compounding killers that retail investors do constantly. His Kodak example and the COVID story cut through the noise of post-hoc rationalizations. The February 2020 investor was right and still lost—a perfect illustration that timing based on news (even true news) is a fool’s game.

The Q&A with Nvidia shows his pragmatism. He admits the difficulty of calling peaks and doesn’t pretend he can. His answer—trim gradually to a comfort level—is humble and actionable. Not “hold for ever” or “get out at the top,” but “own enough to feel the upside, not so much you lose sleep.”

The weaker part: vagueness on when valuations are actually excessive. “Absolutely nuts” is subjective. He gives markers (70–80x earnings), but an investor in Nvidia at 33x earnings growing at 30%+ might not feel it’s nuts—and he might be right. The framework doesn’t resolve that tension; it just flags that you have to monitor.

The tax and churn discussion is specific to India and high-net-worth investors. For international readers or salaried folks, some of it is local. But the core principle—high activity kills post-tax returns—holds everywhere.

Worth watching if you’re a value investor wrestling with holding periods or if you’ve ever sold a stock and watched it 10x afterward. Thakkar doesn’t promise you’ll get it right; he promises that discipline beats activity.

Score reflects solid framework, concrete examples, and useful psychological calibration—but also the reality that no one fully answers “how do you know when valuations are crazy?” That’s the problem every investor lives in.