Video summary

What Actually Went WRONG at Marcellus & What I'm Buying Now | ft. Saurabh Mukherjea

Main summary

Key takeaways

News and Commentary

Interview Overview

The video is an interview on the “India Opportunity Show” with Saurabh Mukherjea (Marcellus Investment Managers). The discussion centers on:

  • What went wrong for their “clean, well-run” / “coffee can investing” style during 2018–2022
  • How they’re approaching the market now
  • Views on AI’s impact across sectors, especially Indian IT
  • The case for investing in new-age/platform companies
  • Why global diversification matters

1) Can TCS (and Indian IT) reinvent itself?

Mukherjea argues that Indian IT faces a structural headwind from AI, not merely a cyclical downturn.

The core mechanism

  • Global hyperscalers are investing aggressively in AI (notably data centers and AI workloads).
  • As a result, some enterprise spend that might have gone to Indian IT services is instead diverted to hyperscalers and their ecosystems.

What reinvention might look like

  • Indian IT firms may need to remodel/reinvent, potentially:
    • Serving hyperscalers directly
    • Using AI internally to deliver new services
  • However, he expects the new addressable business to be smaller than the old IT services market—meaning not every firm will successfully transition.

Example: Tata resilience via business model

  • He contrasts TCS with another Tata-related franchise, Tata Elxsi, suggesting that if the underlying business model persists through technological change, a company can survive major drawdowns.

2) What went wrong for their strategy after strong early performance

Mukherjea says the strategy worked exceptionally well through 2018–2022:

  • He cites returns around the high-20s CAGR
  • Earnings growth closer to ~18%

Performance then deteriorated when fundamentals/valuation relationships shifted.

Key factors he highlights

  • COVID / timing uncertainty on trimming bets
    • “Quality” businesses (healthcare/essentials) performed well, so they retained exposure longer than later hindsight would have suggested.
  • Diwali 2021 as a valuation peak
    • Valuations ran up for their style; forward P multiples were allegedly around ~50.
  • Post-2022 challenges
    • “Revenge spending / unlock” was harder to capture because growth concentrated in premier franchises they owned, while second/third-rung companies grew quickly too—often at much lower valuations.
    • Government capex did not fit their style well:
      • They view many government contractors as having patchy/sketchy histories
      • They also cite accounting issues
    • Certain cyclical consumption shifts didn’t reward their specific portfolio construction as much as hoped.

An objective marker

  • The NSE Nifty 200 Quality 30 underperformed the broader Nifty 200 for roughly five years, aligning with the idea that “quality” was getting “thrashed.”

3) How they responded: three portfolio/process changes

After multi-year underperformance, they describe becoming “wiser/stronger” and implementing three main adjustments.

  1. Diversify more while staying concentrated

    • Portfolios remain concentrated (~15–20 stocks).
    • But they broaden sector exposure rather than relying mainly on two big themes (e.g., consumption + financial services).
    • Sectors include: healthcare, FMCG, discretionary consumption, financial services, export-oriented manufacturing, capital goods, internet, etc., while staying true to the “clean, well-run” philosophy.

  2. Act when price outruns earnings for specific stocks

    • They don’t necessarily exit fully.
    • Instead, they cut position size when the stock price runs far ahead of EPS compounding (to avoid valuation-compression pain such as P from 50 to 40).

  3. Exploit an expanding “consistent compounders” opportunity set

    • After COVID, and amid shifting government priorities (less direct public provisioning in some areas; more emphasis on defense/roads/railways), they see more room for private-sector franchises—especially:
      • Hospitals
      • Diagnostics
      • Private healthcare systems
    • They also see export manufacturing as potentially benefiting from:
      • Currency competitiveness
      • Trade arrangements like “China plus one”
    • This is framed as a potential new frontier for long-term compounding.

4) Reframing quality investing: resilience through attacks and drawdowns

Mukherjea argues that great franchises can underperform for long periods even when the business remains strong.

Global analogy

  • He uses Microsoft as an example:
    • Some years showed near-zero returns amid regulatory/legal and leadership transition periods
    • Later performance recovered

“Unusual Billionaires” audit

  • He references an earlier book (“Unusual Billionaires”) and notes:
    • Most named companies still beat the Nifty over the decade
  • The implication: long-term time and staying steady through noise can outperform.

What matters: competitive edge, not stock volatility

  • Competitive attacks and drawdowns are “normal” for quality companies.
  • The critical question is whether the firm can hold its competitive edge.
  • If yes, valuation declines can turn into opportunity.

5) Example: selling a “quality” franchise (and being wrong early vs later)

He discusses Relaxo:

  • It initially looked like a strong franchise:
    • Brand strength
    • Modes and performance indicators
    • ROC above cost of capital
  • But he attributes an error to a strategic misjudgment:
    • Post-COVID competitors undercut with cheaper products
    • Relaxo’s pricing/actions widened the gap
    • Share erosion followed, hurting earnings and the stock

He contrasts this with cases like Divi’s/DG Lab, where major drawdowns eventually recovered—emphasizing their need to assess whether problems are repairable vs structural.

6) View on “new-age”/platform companies and AI’s effect

Not “new-age vs old-age”

Mukherjea rejects a simplistic split. Core investing criteria remain the same:

  • Return on capital above cost of capital
  • Rational capital allocation
  • Durable comparative advantages

Example: CarTrade

  • Even if the IPO valuation looks rich, he argues the platform is singular/differentiated.
  • Its strength is tied to trusted counterparty interactions, such as:
    • Repossession auctions
    • Financing support and related processes

AI’s likely role

  • He disputes the claim that AI will instantly replace such marketplaces/platforms.
  • Instead, he argues AI will mainly improve:
    • Matching/search
    • user decision-making
  • But fulfillment and trust/risk management (counterparty/payment/escrow) still depend on the platform structure.
  • Therefore, AI could strengthen vertical platforms rather than destroy them.

He extends this logic to other vertical ecosystems built on trust and transactions.

7) Why global diversification matters (risk smoothing, not higher returns)

Mukherjea says global diversification isn’t primarily about earning more in the US; it’s about reducing risk through lower correlation across markets.

The “double-engine” idea

  • With India + global equities, returns may stabilize over time.
  • He frames this as “sleep better,” driven by:
    • reduced volatility
    • smoother risk over time

Practical enablers for Indian investors

  • He mentions infrastructure like GIFT City/IFSC, which can make global investing more cost/tax-efficient for Indians.

Globalization of life expenses

  • He notes many expenses are already effectively global:
    • imports
    • travel
  • This supports the idea that assets should also be diversified globally.

Presenters / Contributors

  • Saurabh Mukherjea — Guest; Marcellus Investment Managers
  • Rishi Shahu — Host / presenter
  • “India Opportunity Show” host persona — Interviewer (“Shi” in subtitles), who introduces and asks questions

Original video