old-Portfolio Management Services


Rob Kirby, in a note written in 1984 (in the Journal of Portfolio Management), narrated an incident involving his client’s husband. The gentleman had purchased stocks recommended by Kirby in denominations of US$5000 each but, unlike Kirby, did not sell anything from the portfolio. This process (of buying when Kirby bought but not selling thereafter) led to enormous wealth creation over a period of about ten years.

The wealth creation was mainly on account of one position transforming to a holding worth nearly $1m in Xerox. Impressed the power of this approach – of buying great companies and then letting them for ten years – Kirby coined the term “Coffee Can Portfolio”.

Nearly 40 years later, Peter Thiel hit upon the same insight but this time in the context of the VC investing – great VC portfolios are defined by a couple of blazing winners at the end of ten years. Thiel calls this the “Power Law”.


Marcellus Investment Managers in December 2018 launched a PMS strategy – Consistent
Compounders to invest in a concentrated portfolio of heavily moated companies that can drive healthy earnings growth over long periods of time.

Portfolio construction involves a two stage process:

  1. a filter based approach to create an investible universe of 30-35 stocks
  2. in-depth bottom-up research of such companies in the universe to assess sustainable
    competitive moats to build a portfolio of 10-20 stocks that deliver healthy compounded earnings growth over long periods of time.

Such a portfolio is monitored for sustainability of moats on a continuous basis through extensive
primary research Repeating the filters annually helps keep the investible universe updated and also such a universe is continuously researched for developing or strengthening of moats to augment the portfolio

  • The Filter based approach
  • The Power of a Filter based approach
  • So how do we (as fund manager) add value
We create a list of stocks using a twin-filter criteria of double digit YoY revenue growth and return on capital being in excess of cost of capital, each year for 10 years in a row.Next, we build a portfolio of such stocks each year and hold each of these annual iterations of portfolios for the subsequent 10 years (without any churn).The bar chart on the right shows the backtesting performance of such a filter based portfolio.
There are two conclusions from this exercise:

  • This filter based portfolio delivers returns of 20-30% p.a. and 8-12% outperformance relative to the Sensex.
  • The volatility of returns of such portfolios, for holding periods longer than 3 years, is similar to that of a Government Bond

Returns here (both for our portfolio and for the Sensex) are on a Total Shareholder Return basis i.e. all dividends are included in the returns.

Unique DNA of these companies: By “filtering in” companies with a history of very consistent fundamentals over very long time periods, the portfolio is skewed towards companies with a DNA built around relentlessly deepening their competitive moats despite disruptive changes taking place both inside as well as outside the organization. More often than not, such DNA sustains over the subsequent 5-10 years investment horizon of the filter based approach.Power of compounding: Holding a portfolio of stocks untouched for 10 years allows the power of compounding to play out, such that the portfolio becomes dominated by the winning stocks while losing stocks keep declining to eventually become inconsequential.Avoiding the pitfalls of psychology and reducing transaction costs: Being patient with a portfolio helps cut out ‘noise’ of trying to time entry / exit decisions. With no churn, this filter based approach also reduces transaction costs. Consider two data points: (a) In a portfolio with 70% churn (average churn of large cap mutual funds), 20bps broking cost and 30bps impact cost, churn reduces the terminal value of the portfolio (after 10 years) by 10% (i.e. a drag of 120bps on the 10-year CAGR); and (b) deferring the 10% long term capital gains tax payable on the portfolio by 10 years enhances the terminal value of the portfolio by 8% (i.e. 100bps increase in the 10-year CAGR) vs a portfolio where capital gains are paid each year.
The Consistent Compounders Portfolio combines our deep-dive stock-specific research with the benefits of the filter-based approach explained earlier, to help generate outperformance of 4-5% per annum over and above these filter-based portfolios. This is achieved via 3 factors:Portfolio concentration: The filters might give a longer list of stock which dilutes the reliance of the portfolio on outstanding companies. We narrow the portfolio down to 12-15 ultra-high quality stocks. So, how do we do that?Ignorable consistency in historical fundamentals: Eg. Many housing finance companies which form part of the filter-based portfolios, are examples of 10 years of consistent fundamentals delivered due to unsustainable macro tailwinds for the Housing Finance Companies from low cost money market funding and a booming real estate market in the country – neither of which to our mind is sustainable.Excusable blips in historical fundamentals are forgiven: For example, Nestle’s Maggi episode ensured that revenue growth of Nestle India dropped below 10% in FY15. Similarly, the fall in crude oil prices to below US$30 per barrel caused a 6% product price cut by Asian Paints in FY17 which led to its revenue growth dropping below 10% YoY in FY17. Manual intervention in portfolio construction analyses the nature of these blips and might include such stocks in the portfolio.

Fee Structure

Marcellus offers Consistent Compounders Portfolio with zero fixed fees

The Consistent Compounders PMS comes with ZERO entry load/exit load and with no lock-in. Our clients can choose any of the following fee structures:

    1. a fixed fees model (2% p.a. fixed fees + zero performance fees) or
    2. a variable fees model (zero fixed fees + performance fees of 20% profit share above a hurdle of 8%, no catch-up)*
    3. 3.a hybrid model (1% p.a. fixed fees + performance fees of 15% profit share above a hurdle of 12%, no catch-up).

High water mark applies for performance fees

Minimum investment: INR 50 lacs


Marcellus’ Consistent Compounders PMS performance as on April 30

Portfolio Fundamentals-Q4FY20

Marcellus’ Consistent Compounders PMS performance