He Built A Live Quant Stock Portfolio That Shocked Me

ELI5 / TLDR

Rishabh Nahar runs Qode Advisors, a Mumbai shop managing about ₹6,700 crore by rules instead of gut feel. The pitch: most professional fund managers lose to a dumb index, so build a slightly smarter index — pick companies that are growing, not overpriced, and already rising — and follow it mechanically for years. On a live screen he layers four filters onto a 10-year backtest and watches the return climb from 19% to 29% a year. Around it sits a worldview about money-printing, gold, and currency depreciation that explains why he never holds cash and rotates between equities and gold.

The Full Story

What “quant” actually means here

The word sounds like robots trading every millisecond. Nahar spends the first ten minutes killing that image. To him, quant means rules-based, emotion-free investing — nothing more exotic than writing down a checklist and obeying it.

It is not that Rishabh has done any… the formula throws out the stock and that stock ends up doing really well. Quant investing is more than that. It is a set of rules that you have to work with.

He and two college friends — all commerce graduates, none could code — started Qode in 2016 out of a two-bedroom flat in Dadar. They taught themselves coding so they could stop staring at screens all day. The motivation was temperament, not mathematics: he couldn’t sit in front of a terminal making subjective calls forever, so he tried to turn judgment into a process. The label he prefers is “systematic investor.”

The numbers he quotes: roughly 23–24% CAGR since 2018 for most investors in his funds; the flagship strategy launched November 2024 (near the market top) has annualized about 31% while the Nifty went slightly negative. Treat the precise figures as marketing — this is his own podcast appearance — but the philosophy underneath is coherent.

Why bother — the case against active managers

The hook for going systematic is a single statistic. Of 81 PMS schemes (portfolio management services — discretionary managers for wealthy clients) with a 10-year track record, only 26 beat their benchmark. That’s 32%. Two-thirds of professionals, with all their factory visits and management meetings, lost to the index.

So what is the index? Nahar’s sharp observation: the Nifty is itself a quant strategy — a dumb one. It just ranks the 50 biggest companies by market value and weights them, with zero regard for whether a business is any good.

It’s actually a very silly quant indicator… If such a silly index can beat active managers 68% of the time, let’s make a smarter index.

That smarter index is his whole business.

Building the model live, filter by filter

This is the segment that gives the video its title. Using a backtesting tool (Portfolio Visualizer — software that replays a strategy against historical data to see how it would have done), he constructs a portfolio in real time, watching the 2013–2024 return change as he adds rules. Backtesting just means: pretend you’d run these rules for the last decade, and measure the result.

Start dumb. Buy every company between ₹500 crore and ₹20,000 crore in market value — roughly 3,000 small- and mid-cap firms — equally weighted. Result: 19% CAGR, but a brutal 61% drawdown in the March 2020 crash. (“Drawdown” = how far the portfolio fell from its peak.)
Add quality. Keep only businesses that are actually growing — positive revenue growth, expanding margins, rising earnings, or a return on capital above ~12%. The universe shrinks to ~1,128 firms. Return jumps to 24%.
Add valuation. Refuse to overpay — cap the price-to-earnings and price-to-book you’ll accept. Return to ~26%.
Add momentum. Only buy stocks whose price has been climbing over the past 6–12 months. Return reaches ~29%.

He’s careful not to oversell the momentum filter, and the reasoning is the most honest part. Momentum works because insiders and the wider industry often know good results are coming before they’re announced.

Lenovo itself told me, we are going to get the full price… That same laptop we bought for 40,000 last year, we bought for 75,000 this year.

The price runs up before the earnings print. Momentum tries to ride that. The refined version uses volatility-adjusted momentum — a stock that climbed steadily 100→200 scores better than one that lurched 100→300→100→200, because the smooth riser reflects real earnings rather than speculation.

The five questions, not one

His best point cuts against how everyone talks about investing. People obsess over which stock. Nahar insists there are five decisions, and stock selection is only the first:

Which stock you bought, how much you want to buy, when to buy it, when to sell, and how you manage it. These five elements will make you a successful investor.

A tip with no position size and no exit rule is worthless. If someone gives you a 10x stock but you put 1% of your portfolio in it, it never moved your life. The quant framework forces all five to be answered — which, he argues, is the real reason to systematize.

Risk on, risk off — and a valuation gauge

The model above stays 100% invested in small-caps forever. The real fund doesn’t. Nahar’s headline principle is risk management: don’t be a hero in a bull market, be a hero in a bad one. Anyone makes money in a bull market just by staying exposed.

His tool for sensing the weather is a valuation spread indicator. For every one of ~5,000–6,000 listed companies, he checks whether today’s price-to-book sits above or below its own 15–20 year median, scores each 1 or 0, and aggregates. In November 2024 the reading hit ~74% — three-quarters of companies trading expensive — so he avoided microcaps entirely and shifted toward large-caps, gold, and bonds. That kept his investors out of the 2025–26 drawdown. By the time of recording the gauge had fallen to the mid-40s (neither cheap nor dear). The analogy he likes: drive in fifth gear on an open road, first gear in heavy traffic.

He stresses this has to be mechanical, because nobody can do it by feel for twenty years. At a market top, every newspaper is selling the India growth story; at a bottom, everyone is despairing. The number doesn’t care.

The macro worldview: why he never holds cash

A big chunk of the conversation is macro, and it’s where to be most skeptical. The thesis: central banks have only two tools — interest rates and printing money — and every crisis (2008, 2020) ends the same way, with a flood of liquidity that inflates every asset. He invokes Ray Dalio’s “long-term debt cycle” and the US debt pile (~$38–40 trillion, ~120% of GDP) to argue more printing is inevitable, possibly larger than 2020’s.

The conclusion he draws is the practical part: never sit on cash, because cash is the thing being debased. He rotates between equities and gold, sometimes land. His personal split is roughly 60–70% equities, 30–40% gold.

His gold argument is genuinely clarifying. Indian gold returned ~11% a year (pre the recent rally) while global gold did ~7%. The 4% gap is pure rupee depreciation. Owning gold is owning an exit from the rupee.

The moment you transfer your rupees to gold, you are foolproof of the rupee… It is a storehouse of wealth, future-proof against currency depreciation.

He drives it home with extremes: gold in Turkish lira returned 5,420% since 2016 versus 269% in dollars — the difference is the lira collapsing. The “best-performing stock market in the world” is the Karachi exchange, for the same sad reason. The lesson isn’t to chase those markets; it’s that currency exposure is a silent tax, and bonds — which pay a fixed 7–8% — are the worst place to be when your currency is eroding faster than that.

Jim Simons and Ed Thorp

The last stretch is hero-worship, but useful. Jim Simons ran the Medallion Fund at ~66% gross CAGR (~39% after its famous fees), started seriously investing around 50, and was a mathematician who hired no finance people — only PhDs in physics and astronomy. Nahar copies the staffing: ~75% of his 50-person team is non-finance, because “finance can be taught in three months; problem-solving can’t.” Ed Thorp gets a retelling — card-counting at blackjack (the 21 / Beat the Dealer story), then a roulette-predicting computer hidden in a shoe, then a 30-year, ~30% investing record. The shared idea: the market, unlike a casino, lets you keep your edge and compound it, and it won’t throw you out for winning.

He name-checks the Kelly criterion (a formula for how much to bet given your edge and odds) and Sanjay Bakshi’s probability-based position sizing — both pointing back to his refrain that how much matters as much as what.

The closing advice, asked of every guest: start investing as early as possible (Buffett began at 11), and — his personal addition — marry early. Compounding rewards time more than genius; 26% for ten years is already 10x.

Key Takeaways

Quant ≠ high-frequency trading. Here it means rules-based, emotion-free investing — a checklist obeyed for years, rebalanced roughly annually, not daily churn.
The benchmark is itself a quant strategy. The Nifty mechanically ranks companies by market cap with no quality screen — a “silly” factor model that still beats two-thirds of active managers.
68% of 10-year PMS schemes lose to their benchmark (26 of 81 beat it). The case for systematic investing is that beating the index is rare and hard.
Factor stacking on a 2013–2024 backtest: market-cap universe (~19% CAGR) → + quality/growth (~24%) → + valuation cap (~26%) → + momentum (~29%).
A 60% win rate is enough. Of 30 stocks, ~18 in the green still produced 29% — winners outweigh losers. Even Buffett runs a ~60–70% hit rate.
Volatility-adjusted momentum rewards stocks that rose smoothly, filtering out speculative round-trips; it tries to capture real earnings momentum, not PE re-rating.
Five decisions, not one: what to buy, how much, when to buy, when to sell, how to manage. A stock tip without position size and exit is useless.
Valuation spread indicator: share of all listed companies trading above their own long-run median price-to-book — a market-wide cheap/expensive gauge (~74% = expensive in Nov 2024, ~45% at recording).
Risk-on / risk-off dictates exposure: small-caps when cheap, shift to large-caps/gold/bonds when the gauge is extreme. Protect the downside; the upside takes care of itself.
Never hold cash, rotate equities and gold. Cash is the asset being debased by money-printing; gold is an exit from rupee depreciation (Indian gold’s ~4% premium over global gold is the currency loss).
Price chases earnings. They track EPS growth, not price targets; a rally driven by PE expansion (liquidity) without earnings growth — the 2024 momentum-ETF trap — is one they avoid.
Hire problem-solvers, not finance grads — Simons-style. Finance is learnable in three months; the scarce skill is solving problems.

Claude’s Take

This is one of the better finance interviews of its genre, mostly because Nahar keeps deflating his own mystique. He repeatedly says the model is simple, the filters are obvious, an IQ of 120–130 is enough, and the win rate is only 60%. That candor is rare on a channel whose audience speaks in “X” returns.

The live model build is the real value. Watching the CAGR climb 19→24→26→29% as quality, valuation, and momentum filters stack on is a clean education in factor investing, and it lands without a single equation.

Now the BS filter. The 29% backtest deserves heavy suspicion. A backtest is a strategy fitted to the exact decade it’s measured on, and Nahar admits as much — he warns about “curve fitting” and “survivorship,” and notes the displayed model isn’t the one they actually run (“we have a better version”). That hedge cuts both ways: it’s honest, but it also means the impressive number on screen is a demo, not a track record. He himself says the live results sit lower (slippage assumptions of ~1%, real drawdowns of 15–20%). Anchor on his actual claim — ~23–24% CAGR since 2018 — not the 29% screen. And every figure here is self-reported on a promotional podcast; the ₹6,700 crore AUM, the 31% flagship number, the “we exited Trent at the right time” anecdotes — none are audited in this format.

The macro section is the weakest. The “central banks will print, so own everything” thesis is directionally defensible but has been an evergreen prediction for fifteen years; “more printing than 2020, probably double” is a guess dressed as inevitability. The gold-and-currency-depreciation point, by contrast, is genuinely good and survivorship-free — the lira and Karachi examples are vivid and correct.

Score: 7/10. Substantive, well-explained, refreshingly self-aware about its own limits — but it’s a fund manager pitching his fund, the headline backtest is a fitted demo rather than a verified record, and the macro is hand-wavy. Worth watching for the factor-stacking walkthrough and the “five decisions” framing; worth discounting everything with a number attached to it.

He Built A Live Quant Stock Portfolio That Shocked Me | Ft. Rishabh Nahar | Kushal Lodha #349