Full replication

The simplest way to copy an index is to buy everything in it, in the same proportions the index uses. That is full replication — no approximation, no substitution, just the index's own constituent list held directly.

It works cleanly when the index is small and its stocks are easy to trade. It becomes much harder once the index has hundreds of constituents, some of them thinly traded — which is exactly why a second technique, sampling, exists for those cases.

Section 18.1.4.1 defines it directly: when a manager creates a portfolio by holding all the securities in the index with the same or close to the same weightages as the index, it is called Full Replication. Whenever the index is rebalanced, the manager must adjust the portfolio to match — and index rebalancing is publicly announced by the index provider in advance, so the manager can plan for it.

Full replication is advisable when three conditions hold together: (i) the index has a limited number of constituents, (ii) each constituent is liquid, and (iii) the portfolio has the minimum threshold size needed to hold every constituent in the right proportion. The workbook's own examples of indices that suit full replication are Nifty 50 and the S&P BSE Sensex — both have a limited number of stocks, liquid and tradable constituents, and represent the broad market well, which together make it straightforward to build a portfolio that mirrors either index directly.

Illustrative figures. A portfolio manager runs a Rs 50,00,00,000 Nifty 50 index fund using full replication. Nifty 50 has exactly 50 constituents, all large, liquid, actively traded stocks.

The manager buys each of the 50 stocks in exactly the weight it carries in the index — for instance, if a large private bank carries an 11% index weight, the fund holds Rs 5,50,00,000 of that stock. When Nifty 50's quarterly rebalancing (publicly announced by NSE in advance) removes one constituent and adds another, the manager sells the removed stock and buys the added one in the corresponding weight, on the rebalancing date.

Contrast this with an attempt at full replication of the Nifty 500 — 500 constituents, many of them small-cap and thinly traded. Buying all 500 in exact index weights on a Rs 50 crore fund would mean some individual positions are too small to trade efficiently, and several small-cap constituents may not have enough daily volume to be bought or sold without moving their price — exactly the scenario the workbook says makes sampling, not full replication, the more practical choice.

Chapter 18, section 18.1.4, contrasts full replication (18.1.4.1) with sampling (18.1.4.2) as the two indexing portfolio construction techniques, using Nifty 50/Sensex and Nifty 500 respectively as the workbook's own contrasting examples. Expect a question on the three conditions that make full replication advisable, and one asking which technique suits a large, illiquid-tailed index like the Nifty 500.

• Full replication needs all three conditions together — limited constituents, liquid constituents, and sufficient portfolio size — an index that is small but illiquid, or liquid but very large, is still a poor fit for full replication.
• Index rebalancing is publicly announced in advance, so a full-replication manager is not caught by surprise — the adjustment is a scheduled, known event, not a reactive scramble.
• Full replication is a form of passive management, like sampling, but the two techniques solve the same problem (tracking the index) in different ways depending on the index's own characteristics.
• Nifty 50 and Sensex suit full replication; Nifty 500 typically needs sampling — the workbook's own named contrast, and a natural fill-in-the-blank question.