40% CAGR. 13 Years. The Engineering System Used To Find 50X Multibagger Stocks

ELI5/TLDR

A Pune-based fund manager — a mechanical engineer who blew up his account to zero three separate times before he learned to read a balance sheet — explains the system he’s used to compound roughly 40% a year for 13 years. The trick isn’t finding winners; it’s throwing away most of the market first. He deletes whole sectors (anything cyclical, anything with dodgy promoters, anything expensive), then hunts for the boring, under-researched specialist hiding inside a sector everyone ignores. His favourite shape of business is the contract manufacturer — the company that quietly makes things for big global brands and earns fat, durable margins doing it.

The Full Story

Three trips to zero

The guest (the transcript never cleanly states his name — it’s a SEBI-registered PMS founder, ex-Lehman Brothers, ex-Access Holdings) didn’t start as an investor. He did mechanical engineering at Pune, an M.Tech at IIT Kanpur in 2002 working on “rapid prototyping” — what everyone now calls 3D printing. He tried to build businesses out of it: prototyping jewellery moulds, then turning MRI and CT scans into 3D models for doctors. Both were too early. The doctors told him plainly: you’re ahead of your time, nobody will pay.

He drifted into finance through the back door — software jobs at financial-services firms, then a CFA, then in 2008 a private-equity seat at Access Holdings. But before that, from 2005 to 2008, he traded futures and options. He lost everything. Not once.

“I invested X amount it become zero. Then again I put X amount again it become zero. Then again put X amount again it become zero. So it was not one or two, it’s three times it has become zero. And it is only because of trade.”

His advice on trading is one of the few unhedged things he says: don’t. “You will lose your time, you will lose your money, and you will lose your mental peace.” The PE job is where he learned the difference between watching a screen and owning a business.

The single obsession: don’t lose the money

When he started managing money for a family office in 2012, he didn’t begin with a return target. He spent six months on one question: how do I make sure the capital never gets wiped out, no matter what the world does?

“You can’t blame the market. But you should be able to protect the investor’s money.”

That defensive instinct became the spine of the whole system. Out of 5,000-plus listed companies, the first move is deletion, not selection.

Elimination first

Three filters knock the universe down to maybe 100–150 names:

He throws out cyclicals — steel, cement, sugar, construction, real estate — anything whose earnings boom and bust with a commodity or a capex cycle. He throws out any company or promoter with a corporate-governance stain. And he throws out anything expensive, however good the business. His logic on that last one is worth keeping: the only two risk dials you actually control are how much you allocate and the price you pay on the way in. Everything else is the market’s to decide.

Then: right sector, right business, right price

What survives the cull gets judged on three things.

Right sector means a decadal theme — a runway of ten years, not two. He doesn’t necessarily hold for ten years; he just wants enough road that a valuation mistake gets bailed out by time. A two-year theme is dangerous because the crowd starts leaving a year before the story ends.

Right business is the heart of it. Every sector, he argues, splits into a commodity half and a specialist half. Pharma has generics (commodity) and CDMO (specialist). Chemicals have bulk versus speciality. Even defence — currently the market’s favourite word — hides plenty of plain fabrication shops that will never earn more than 20% gross margin no matter who they sell to.

“Any business which works on the tendering process is not a good business because there is always a race to the bottom… So that L1 is the bad word for the investment.”

He won’t touch banks or NBFCs at all. His reasoning: money itself is a commodity, so any lender can manufacture 50% growth simply by loosening its credit standards. He’ll buy the technology sold to banks, never the bank.

Right price rests on what he calls the two gods of valuation: growth and return on capital. Growth comes from the size of the opportunity. Return on capital comes from a sustainable competitive advantage — and he leans hard on “sustainable.” A company with both deserves a premium multiple; a company with only one doesn’t. His cautionary tale is infrastructure in 2008: huge opportunity, but returns on capital barely above the cost of borrowing, so when projects slipped and interest piled up, the maths flipped negative and the companies went bankrupt.

The re-rating trade, explained simply

The cleverest stretch is his account of where the multiple actually comes from. A multiple isn’t fixed by the business; it’s the investor’s perception of future growth and returns. So the money is made by buying a company before the big institutions notice it.

His chain: a company that is under-researched must be under-owned, and an under-owned company must be undervalued. When the earnings keep landing and the business reaches a certain scale, analysts start covering it, institutions start buying, and the P/E expands — because a foreign fund borrowing at 4–5% is happy with a 15% return and will therefore pay up for a business growing 25%. You don’t just earn the profit growth; you earn the re-rating on top.

The CDMO thread running through everything

Nearly every winner he names is some flavour of contract manufacturer — a business that makes things, to spec, for demanding global customers, under long relationships that are hard to win and hard to lose.

Suven Pharma — bought in 2013 at a ~200 crore market cap, when nobody in India used the word CDMO. What flagged it: the company expensed all its R&D instead of capitalising it to flatter profits — a sign of an honest, bold management. Underneath sat a cash-gushing CDMO funding a decades-long bet on an Alzheimer’s molecule. He bought at a 5–6 P/E and rode it to roughly a 50-bagger.
Manorama Industries — 2019, ~200 crore market cap. He read about it in a magazine, visited the plant, and took a significant stake the next day — the fastest decision of his career. The edge: it was the only Indian firm procuring sal seed (the raw material for cocoa-butter equivalent), supplying chocolate giants who audit suppliers for years. Big opportunity, real entry barrier.
Laurus Labs — a pure operating-leverage bet. They’d doubled their gross block but weren’t using it yet. He waited 18 months, asking management the same utilisation question on every call, until the capacity filled, profit jumped ~10x, and the stock followed.
Deepak Nitrite — bought near ₹150 as its giant phenol plant (India’s first at scale) was about to commission. Commissioning slipped a year, then operating leverage plus a move into higher-margin phenol derivatives took it to ~₹3,000 — roughly 20x in three years.
Axiscades — an engineering-R&D business that, after a management change (the new boss came from Tata Technologies), pivoted into defence CDMO. Azad Engineering is his cleaner defence/aerospace CDMO bet, supplying Honeywell, Rolls-Royce, Safran, Siemens and Mitsubishi.

The leverages, stacked

He frames the whole return as a stack of leverages firing in sequence: product-mix leverage (margins rise as you sell richer products), operating leverage (fixed costs shrink as a share of growing revenue), financial leverage (interest and depreciation shrink as a share of revenue), and finally valuation leverage (the multiple re-rates as return on capital climbs). When all four fire, gross profit grows faster than revenue, EBITDA faster than gross profit, PBT faster than EBITDA — and the P/E expands on top of all of it. That compounding-of-compounding is the multibagger.

Where he got it wrong

To his credit, he names the mistake. He got “carried away” with chemicals and pharma during Covid, extrapolating pandemic demand and the India-plus-one story too far. When demand normalised, Deepak Nitrite fell from ~3,000 to ~2,000 and Laurus from ~700 to ~300, and the portfolio printed roughly -26% in FY23. He also admits one chemical position caused “massive erosion of wealth.”

Knowledge first, wealth second

His closing sermon: everyone obsesses over compounding wealth; nobody talks about compounding knowledge. Wealth, he says, is just the by-product. The line he draws is between knowledge of a business (what happened, what’s happening) and understanding of it (what’s coming in three to five years, and what to do about it) — the difference between an investor and a promoter. Mediocre returns come from knowledge; multibaggers come from understanding. He runs his PMS with a ₹25 crore minimum cheque, partly to keep out investors who want to book a quick 30% and leave. His own target, stated flatly: 50x in ten years, every time.

Key Takeaways

Eliminate before you select. From 5,000+ stocks, cut all cyclicals (steel, cement, sugar, construction, real estate), anything with a governance blemish, and anything richly priced — leaving ~100–150 names.
The only two risk dials you control are allocation and entry price. Be strict on both; never overpay even for a great business.
Three-part selection: right sector (a 10-year “decadal” runway), right business (the specialist, not the commodity, half of every sector), right valuation.
Two “gods” of valuation: growth (from size of opportunity) and return on capital (from a sustainable competitive advantage). You need both; one alone doesn’t justify a premium.
Tender/L1 businesses are bad businesses — a structural race to the bottom. Avoid “defence” names that are really just fabrication shops capped at ~20% gross margin.
No banks or NBFCs — money is a commodity, growth can be faked by loosening credit. Buy the tech vendor to banks instead.
The re-rating chain: under-researched → under-owned → undervalued. Buy before institutions discover it; the P/E expands when they do.
A multiple is the investor’s perception of future growth and ROC, not a property of the business. Rising ROC re-rates the stock.
Four stacked leverages drive a multibagger: product-mix, operating, financial, and valuation leverage firing in sequence.
The repeatable pattern is CDMO — contract manufacturers (pharma, food, engineering, defence) with sticky global customers and durable margins.
Operating-leverage setup: buy a company that has built capacity it isn’t using yet, then wait for utilisation; track it via the same utilisation question every concall.
Honest accounting is a signal — Suven flagged itself by expensing R&D rather than capitalising it to flatter profits.
Compound knowledge, not just wealth. “Understanding” (what’s coming next) beats “knowledge” (what already happened).

Claude’s Take

This is a genuinely good investing interview — coherent, internally consistent, and unusually concrete about why each bet worked. The elimination-first discipline, the commodity-vs-specialist lens, and the “multiple is perception” framing are all worth keeping. He even names his biggest mistake (extrapolating Covid demand into a -26% year), which most fund managers performing for a camera will not do. That candour is what pushes this above the usual finance-podcast fare.

But run the standard filter. This is survivorship storytelling told in reverse: a fund manager narrating his five best outcomes — Suven 50x, Deepak 20x, Laurus 10x — and fitting a clean framework over them after the fact. We never hear the names that fit the same framework and went nowhere (he admits one chemical bet caused “massive erosion of wealth” but doesn’t name it). A system that produces five 10–50-baggers and one near-total loss can still be excellent — but you cannot judge that from the winners alone, and the “40% CAGR / 50x target” headline numbers are self-reported, from a family-office track record that only became a SEBI-registered PMS in 2022. None of the pre-2022 record is independently audited here. Treat the CAGR as a claim, not a fact.

Two more cautions. First, “buy the under-researched small-cap before institutions notice” is the single most over-told story in Indian investing right now, and he himself warns the small/mid-cap space is full of narrative-driven valuations that “will come down” — so the strategy he’s selling is the one whose hunting ground he says is currently overpriced. Second, the transcript is an auto-translated Hinglish mangle, and several specifics are unreliable: the guest’s own name never comes through cleanly (rendered as “Mr. Jven”); company names are garbled (“Loris/Lotus Labs” = Laurus Labs, “Monorama” = Manorama, “Access kids/Axis CAD” = Axiscades, “DV/DVS” = Divi’s, “Deepak 9” = Deepak Nitrite); and a few figures are explicitly hedged by the speaker himself (he says he’s “not remembering honestly very correctly” on Deepak’s plant size). The shape of each thesis is trustworthy; the exact numbers and the missing track-record context are not. Score: 7.

40% CAGR. 13 Years. The Engineering System Used To Find 50X Multibagger Stocks | Kushal Lodha #21