Building a portfolio core with index ETFs

Open a typical Indian retail portfolio and you find eleven stocks, three mutual funds and an ETF someone recommended. Each position was defensible when it was bought. Together they add up to nothing in particular — usually a concentrated bet on large-cap financials and IT, held by someone who believes they are diversified.

The problem is not stock selection. It is that no one ever decided what the portfolio was supposed to be. A core fixes that by reversing the order: decide the shape first, then fill it.

What a core actually is

The core is the part of your portfolio that expresses no view. It is broad market exposure, held for years, that you do not trade and do not second-guess.

The satellite is everything else — individual stocks, sector bets, tactical positions, whatever you have a specific opinion about.

The point of the split is not that the core outperforms. Often it will not. The point is that it makes the satellite measurable. When your core is a Nifty 50 ETF, every stock you pick has an unavoidable benchmark sitting next to it in the same portfolio view. After two years you will know, without a spreadsheet, whether your selection added anything.

Most people avoid finding that out. It is the single most valuable thing a core does.

Which indices actually diversify

Here is where Indian portfolios go wrong most often. Adding funds is not the same as adding diversification.

• Nifty 50 — the fifty largest companies, heavily weighted to financials and IT. This is the default core.
• Nifty Next 50 — the following fifty. Genuinely different companies, more volatile, and the most common sensible extension of a Nifty 50 core.
• Broader market indices — several hundred names, which mostly means the top fifty plus a long tail that barely moves the needle.
• Sector and thematic indices — banking, IT, pharma, consumption. These are satellites wearing index clothing. A banking ETF is a sector bet, not diversification.

The overlap problem. A Nifty 50 ETF and a broad-market ETF share their largest holdings almost entirely. Owning both feels like diversification and delivers close to none — you have bought the same top thirty companies twice and paid two spreads for the privilege.

Before adding any index to a core, ask what it holds that the existing core does not. If the honest answer is "not much", it is a second unit of the same thing.

A defensible core

For most people, one broad index ETF is a complete core. Two is the sensible maximum.

• Simple: a single Nifty 50 ETF. Unglamorous, and adequate for the large majority of investors.
• Slightly broader: Nifty 50 with a smaller allocation to Nifty Next 50. This picks up genuine mid-cap exposure without overlap.
• With a non-equity sleeve: the above plus a fixed, modest gold allocation for the diversification equity indices cannot provide. See gold, silver and debt ETFs for how that shelf behaves.

Beyond that you are adding complexity without adding exposure. Three equity index ETFs in one portfolio almost always means substantial duplication.

Sizing, honestly

The core-satellite split is a decision about temperament, not optimisation.

• New investor: most of the portfolio in the core. The satellite is where tuition gets paid, and tuition should be affordable.
• Experienced, with a track record you have actually measured: a larger satellite is defensible — if the measurement supports it.
• Experienced, without measurement: treat yourself as a new investor. An unmeasured track record is not a track record.

The failure mode is a satellite that grows by accretion. Each new stock feels small on its own, and eighteen months later the core is a quarter of the portfolio. Write the target split down. Check it against reality quarterly. It will have drifted, and the drift will always be toward more satellite.

The rebalancing rule is the whole strategy

Everything above is setup. This is the part that generates the benefit.

Rebalancing means periodically returning to your target weights — selling what has grown beyond its share, buying what has shrunk below it. It forces you to sell strength and buy weakness, which is correct and which nobody does voluntarily.

Make it mechanical:

• On a schedule. Once or twice a year, on a date you fix in advance. Not when it feels right.
• Or on a band. When any holding drifts more than a set percentage from target. Fixed in advance, again.
• Never on a headline. The moment rebalancing becomes a judgement call, it stops being rebalancing and becomes trading with extra steps.

The schedule is the mechanism. Choosing a date removes the decision, and removing the decision is the entire point.

Executing it with ETFs

The ETF core has one recurring practical cost: every contribution is a market transaction, with a spread and a possible premium attached.

That makes execution discipline part of the plan:

• Contribute less often, in larger amounts. Quarterly beats monthly for an ETF core. Fewer transactions, less spread paid, and the same money invested.
• Check Price vs NAV before each purchase. A premium paid on every quarterly contribution compounds against you across a decade. See how to read an ETF page.
• Limit orders, mid-session. Never at the open, never at the close, never a market order.
• Keep the core off the daily screen. Put satellites on your watchlist; leave the core alone. It is designed to be ignored.

Or do not use ETFs for the core at all. If your plan is a fixed monthly contribution and you want it automated, an index mutual fund executes at NAV, handles fractional amounts, and runs without you. There is no spread, no premium, and no monthly decision to get wrong. Model the contribution with the SIP calculator and let it run.

The honest position: for a purely automated monthly core, the index fund is the better tool. ETFs earn the core slot when you are contributing in lumps, want a specific entry price, or prefer everything in one demat view. Both are defensible. Pretending the ETF is automatable when it is not is where people quietly stop contributing in month five.

The one-page version

1. Decide the split — core versus satellite — and write it down.
2. Pick one broad index ETF, two at most, with no material overlap.
3. Add a small non-equity sleeve if you want diversification equities cannot give you.
4. Fix a rebalancing date and put it in the calendar.
5. Contribute quarterly, with limit orders, after checking price against NAV.
6. Measure the satellite against the core annually, and act on what you find.

Step six is the one that gets skipped and the one that pays. If your stock picking has not beaten the index sitting beside it in the same account, the index should get a larger share. That is not defeat. That is the portfolio working as designed.