The Nifty 50 is 10 stocks plus 40 footnotes

Pulled the NIFTY 50 weights this morning. Wanted to test something.

The NIFTY 50 gets sold as a "diversified" basket of 50 stocks. Look at the weights and the math tells a different story.

HDFCBANK alone is 12.8 percent of the index. RELIANCE 9.4. ICICIBANK 8.1. By the time you add the next 7 names you've crossed 58 percent of every rupee. The remaining 40 stocks share what's left. Average position outside the top 10 works out to about 1 percent. Per-stock factsheets sit at rupeecase.com/stocks/

That isn't a 50 stock portfolio. That's 10 stocks plus 40 footnotes. And it isn't a bug in the NIFTY 50, it's the design.

Cap-weighting compounds the existing winners. Whichever stock went up most last year quietly bought more of itself in your fund. The index re-weights monthly to reflect market cap. Which is just a polite way of saying "buy more of what already won." A walk through this math sits inside /learn/ at rupeecase.com/learn/

The fix isn't to abandon the universe. The 50 NIFTY names are still the strongest 50 large caps in the country. The fix is changing the rule that picks which names you actually own and how much of each. Methodology open at rupeecase.com/strategies/methodology

Three questions I keep asking when I look at a portfolio.

What percent is in the top 10 names? Above 50 percent and you are concentrated, not diversified. You can match yours against ours at rupeecase.com/compare

Are you choosing the rule, or is the index calendar choosing for you? Cap-weight is a passive rule that runs on autopilot. Equal-weight, momentum, factor screens are rules you actually pick.

What is the average size of your smaller positions? Below 2 percent and those names are too small to matter on the upside, too small to teach you anything on the downside. The 28 question risk profile at rupeecase.com/risk-profile surfaces this.

If your "50 stock portfolio" is actually 10 stocks plus 40 footnotes, change the rule. Not the wrapper.

RupeeCase Nifty 50 Equal Weight holds all 50 names at equal weight and rebalances by the rule, not the index calendar. rupeecase.com/strategies/nifty-50-equal-weight

How many stocks you really own: the effective count

There is a single number that turns "10 stocks plus 40 footnotes" into arithmetic. It is called the effective number of stocks. Square every weight, add the squares, and divide one by the total. A portfolio of 50 names at exactly 2 percent each scores 50. A portfolio with one stock at 100 percent scores 1. Everything real sits somewhere in between.

Take the shape from the opening as a worked example. Put 58 percent in the top 10 names, 5.8 percent each to keep the sum simple, and spread the other 42 percent across 40 names at 1.05 percent each. The squares add up to 0.0336 from the top 10 and 0.0044 from the tail, 0.0380 in all. One divided by 0.0380 is 26.3. Fifty tickers, the diversification of about 26 equal positions.

And that is the generous version. The real top 10 is not flat: 12.8, 9.4 and 8.1 at the top pull the squares up hard, so the true effective count of a cap weighted index sits well below the simplified figure.

Push the top 10 share to 70 percent and the count falls to about 20. Drop it to 20 percent and every name sits at 2 percent, so the count is the full 50. The same 50 tickers, a very different amount of diversification.

Two rules for the same 50 names

Cap weight and equal weight start from the same list. They part ways on one decision: what happens to a stock that has just run up.

Neither rule is free. Cap weight trades almost nothing and rides the leaders, which works beautifully in a market led by a few giants and hurts when those giants stall. Equal weight trades every period, carries more of the smaller large caps, and pays costs and short term tax for the privilege. The choice is about which risk you want to carry, not about which rule is clever.

What the equal weight rule did in the model

RupeeCase runs the second rule as a strategy: all 50 Nifty names at equal weight, rebalanced every four weeks. In the model it compounded at 16.35 percent a year with a maximum drawdown of 15.71 percent (Model, back tested, as of 27 Apr 2026). The line below starts at Rs 100 in Jul 2021.

Read that chart with two caveats. The index line here is a smoothed 9.8 percent a year reference for the Nifty 50 TRI, not the daily index. And five years is one stretch of market history, in which mid sized large caps did well. Equal weight has had long spells behind cap weight too, usually when the very biggest companies lead. The optimizer post explains why a rule that makes no forecasts is hard to beat over longer runs.

Myths about the index you already own

The free Learn module on indices and benchmarks walks through how NSE builds the Nifty 50, why it is free float cap weighted, and how a benchmark should be read against any strategy. The tracking error calculator shows how far a different rule on the same universe tends to drift from the index in a normal year.