Basics

What should a stock market index be?
A stock market index should capture the behaviour of the overall equity market. Movements of the index should represent the returns obtained by "typical" portfolios in the country.


What do the ups and downs of an index mean?
They reflect the changing expectations of the stock market about future dividends of India's corporate sector. When the index goes up, it is because the stock market thinks that the prospective dividends in the future will be better than previously thought. When prospects of dividends in the future become pessimistic, the index drops. The ideal index gives us instant-to-instant readings about how the stock market perceives the future of India's corporate sector.


What is the basic idea in an index?
Every stock price moves for two possible reasons: news about the company (e.g. a product launch, or the closure of a factory, etc.) or news about the country (e.g. nuclear bombs, or a budget announcement, etc.). The job of an index is to purely capture the second part, the movements of the stock market as a whole (i.e. news about the country). This is achieved by averaging. Each stock contains a mixture of these two elements - stock news and index news. When we take an average of returns on many stocks, the individual stock news tends to cancel out. On any one day, there would be good stock-specific news for a few companies and bad stock-specific news for others. In a good index, these will cancel out, and the only thing left will be news that is common to all stocks. The news that is common to all stocks is news about India. That is what the index will capture.


What kind of averaging is done?
For technical reasons, it turns out that the correct method of averaging is to take a weighted average, and give each stock a weight proportional to its market capitalisation. Suppose an index contains two stocks A and B. A has a market capitalisation of Rs.1000 crore and B has a market capitalisation of Rs.3000 crore. Then we attach a weight of 1/4 to movements in A and 3/4 to movements in B.


What is the portfolio interpretation of index movements?
It is easy to create a portfolio, which will reliably get the same returns as the index. i.e. if the index goes up by 4%, this portfolio will also go up by 4%. Suppose an index is made of two stocks, one with a market cap of Rs.1000 crore and another with a market cap of Rs.3000 crore. Then the index portfolio will assign a weight of 25% to the first and 75% weight to the second. If we form a portfolio of the two stocks, with a weight of 25% on the first and 75% on the second, then the portfolio returns will equal the index returns. So if you want to buy Rs.1 lakh of this two-stock index, you would buy Rs.25,000 of the first and Rs.75,000 of the second; this portfolio would exactly mimic the two-stock index. A stock market index is hence just like other price indices in showing what is happening on the overall indices -- the wholesale price index is a comparable example. In addition, the stock market index is attainable as a portfolio.


Why are indices important?
Traditionally, indices have been used as information sources. By looking at an index we know how the market is faring. This information aspect also figures in myriad applications of stock market indices in economic research. This is particularly valuable when an index reflects highly uptodate information (a central issue which is discussed in detail ahead) and the portfolio of an investor contains illiquid securities - in this case, the index is a lead indicator of how the overall portfolio will fare. In recent years, indices have come to the fore owing to direct applications in finance, in the form of index funds and index derivatives. Index funds are funds which passively `invest in the index'. Index derivatives allow people to cheaply alter their risk exposure to an index (this is called hedging) and to implement forecasts about index movements (this is called speculation). Hedging using index derivatives has become a central part of risk management in the modern economy. These applications are now a multi-trillion dollar industry worldwide, and they are critically linked up to market indices. Finally, indices serve as a benchmark for measuring the performance of fund managers. An all-equity fund should obtain returns like the overall stock market index. A 50:50 debt:equity fund should obtain returns close to those obtained by an investment of 50% in the index and 50% in fixed income. A well-specified relationship between an investor and a fund manager should explicitly define the benchmark against which the fund manager will be compared, and in what fashion.


What kinds of indices exist?
The most important type of market index is the broad-market index, consisting of the large, liquid stocks of the country. In most countries, a single major index dominates benchmarking, index funds, index derivatives and research applications. In addition, more specialised indices often find interesting applications. In India, we have seen situations where a dedicated industry fund uses an industry index as a benchmark. In India, where clear categories of ownership groups exist, it becomes interesting to examine the performance of classes of companies sorted by ownership group.

Isn't averaging like diversification; cancelling out vulnerability to one stock?
Yes, the averaging that takes place in an index is equivalent to diversification. Diversification cancels out individual stock fluctuations. From an investment perspective, diversification reduces risk. From an information perspective, diversification cancels out stock noise; the only thing left after good diversification is the common factor -- news such as nuclear bombs -- which hits all stocks and cannot possibly be removed by diversification.


Then a larger number of stocks in an index will give more diversification -- isn't that a good thing? Why don't we put all the stocks of the country into the index?
It is, indeed, the case that putting more stocks into an index yields more diversification. However, two things go wrong when we do this too much: First, there are diminishing returns to diversification. Going from 10 stocks to 20 stocks gives a sharp reduction in risk. Going from 50 stocks to 100 stocks gives very little reduction in risk. Going beyond 100 stocks gives almost zero reduction in risk. Hence, there is little to gain by diversifying, beyond a point. The more serious problem lies in the stocks that we take into an index when it is broadened. If the stock is illiquid, the observed prices yield contaminated information and actually worsen an index.


What is wrong with the price information for illiquid stocks?
There are three problems: `stale prices', `bid-ask bounce' and vulnerability to manipulation. Through these problems, an index is actually worsened when illiquid stocks are put into it.


A stock may be liquid on one exchange and illiquid on another -- what price do you take when calculating the index?
Illiquid stocks yield bad price data; so the best quality data will come from the most liquid exchange. In India, that is NSE. The S&P CNX Nifty uses price data from NSE for calculations.


What is `stale prices'?
Suppose we look at the closing price of an index. It is supposed to reflect the state of the stock market at 3:30 PM on NSE. Suppose an illiquid stock is in the index. The last traded price (LTP) of the stock might be an hour, or a day, or a week old! The index is supposed to show how the stock market perceives the future of the corporate sector at 3:30 PM. When an illiquid stock injects these `stale prices' into the calculation of an index, it makes the index more stale. It reduces the accuracy with which the index reflects information.


What is `bid-ask bounce'?
Suppose a stock trades at bid 1440 ask 1490. Suppose no news appears for ten minutes. But, over this period, suppose that a buy order first comes in (at Rs.1490) followed by a sell order (at Rs. 1440). This sequence of events makes it seem that the stock price has dropped by Rs.50. This is a totally spurious price movement! Even when no news is breaking, when a stock price is not changing, the `bid-ask bounce' is about prices bouncing up and down between bid and ask. These changes are spurious. This problem is the greatest with illiquid stocks where the bid-ask spread is wide. When an index component shows such price changes it contaminates the index.


What about market manipulation - how would manipulation of an index take place, and how would an index be made less vulnerable to manipulation?
The index is a large entity and is intrinsically harder to manipulate when compared to individual stocks. Obviously, larger indices are harder to manipulate than smaller indices. The weak links in an index are the large, illiquid stocks. These are the achilles heel where a manipulator obtains maximum impact upon the index at minimum cost. Optimal index manipulation consists of attacking these stocks. This is one more reason why illiquid stocks should be excluded from a market index; indeed this aspect requires that the liquidity of a stock in an index should be proportional to its market capitalisation.


So diversification yields diminishing returns, and illiquid stocks are best kept out of an index.... what is the ideal middle road?
S&P CNX Nifty is a well diversified 50 stock index accounting for 22 sectors of the economy. It is used for a variety of purposes such as benchmarking fund portfolios, index based derivatives and index funds.

S&P CNX Nifty is owned and managed by India Index Services and Products Ltd. (IISL), which is a joint venture between NSE and CRISIL. IISL is India's first specialised company focused upon the index as a core product. IISL have a consulting and licensing agreement with Standard & Poor's (S&P), who are world leaders in index services.
 

	The traded value for the last six months of all Nifty stocks is approximately 43.72% of the traded value of all stocks on the NSE
	Nifty stocks represent about 55.78% of the total market capitalization as on June 29, 2007.
	Impact cost of the S&P CNX Nifty for a portfolio size of Rs.5 million is 0.08%
	S&P CNX Nifty is professionally maintained and is ideal for derivatives trading List of S&P CNX Nifty stocks

How does the S&P CNX Nifty work?
S&P CNX Nifty is based upon solid economic research. A trillion calculations were expended to evolve the rules inside the S&P CNX Nifty index. The results of this work are remarkably simple: (a) the correct size to use is 50, (b) stocks considered for the S&P CNX Nifty must be liquid by the `impact cost' criterion, (c) the largest 50 stocks that meet the criterion go into the index. S&P CNX Nifty is a contrast to the adhoc methods that have gone into index construction in the preceding years, where indices were made out of intuition and lacked a scientific basis. The research that led up to S&P CNX Nifty is well-respected internationally as a pioneering effort in better understanding how to make a stock market index. See "Market microstructure considerations in index construction" by Ajay Shah and Susan Thomas, CBOT Research Symposium Proceedings, Summer 1998, page 173-193.


What is `impact cost'?
Suppose a stock trades at bid 99 and ask 101. We say the "ideal" price is Rs. 100. Now, suppose a buy order for 1000 shares goes through at Rs.102. Then we say the market impact cost at 1000 shares is 2%. If a buy order for 2000 shares goes through at Rs.104, we say the market impact cost at 2000 shares is 4%. Market impact cost is the best measure of the liquidity of a stock. It accurately reflects the costs faced when actually trading an index. For a stock to qualify for possible inclusion into the S&P CNX Nifty, it has to reliably have market impact cost of below 0.75 % when doing S&P CNX Nifty trades of half a crore rupees.


What do you mean by `an S&P CNX Nifty trade'?
Earlier, we said that the index assigns weightages to index components, and the weight of a stock is proportional to its market capitalisation. This idea can be applied to buying the S&P CNX Nifty. If you buy all 50 stocks in the S&P CNX Nifty, in correct proportions, that would be called "an index trade".


What's the impact cost on a trade for Rs.5 million of the full S&P CNX Nifty?
It is safe to think that the impact cost is 0.2% or so. This means that if S&P CNX Nifty is at 1000, a buy order goes through at 1002 and a sell order gets 998. NSE's NEAT software has special facilities to enable buying or selling the entire the S&P CNX Nifty at one shot. The impact cost is not something fixed. It changes, depending upon the liquidity of the market. Indeed, the time-series of the S&P CNX Nifty impact cost is one of the best measures of changes in market liquidity over the years.


Why does the index keep changing from time to time?
Think of a liquid stock as a good thermometer, one which gives accurate data about the true price of the stock, because it trades actively with a tight spread. The prices observed for an illiquid stock are like readings from a low quality thermometer, which reports noisy data about the phenomenon of interest (the true price of the security). We try to find the fifty best thermometers in the country and average their values to make the S&P CNX Nifty. As time passes, better thermometers become available (in the form of large, liquid stocks that are not in the S&P CNX Nifty). We would like that S&P CNX Nifty always uses the best thermometers possible. So we remove the weakest thermometer from inside the S&P CNX Nifty and accept the new stock into it. The world changes, so the index should change. Yet, the change should not be sudden - for that would disrupt the character of the index. S&P CNX Nifty uses clear, researched and publicly documented rules for index revision. These rules are applied regularly, to obtain changes to the index set. Index reviews are carried out every quarter to ensure that each security in the index fulfills all the laid down criteria. IDBI was once not listed; SBI was once illiquid; Infosys was once an obscure software startup. The world changes, and one by one, these stocks have come into the S&P CNX Nifty. Each change in the S&P CNX Nifty is small, so the continuity of the index is maintained. Yet, at all times, S&P CNX Nifty represents the 50 most important liquid stocks in the country, the best thermometers to build an index out of.


When a stock goes out and a new stock comes in, doesn't that make index levels non-comparable?
No. There are mathematical formulas, which ensure that yesterday's value and today's are comparable, even if a change in composition takes place in-between. Think of an index as a portfolio. The composition of the portfolio changes, but it is still meaningful to keep measuring the overnight returns on the portfolio from day to day. These returns, cumulated up, are the index level.


Index revision sounds dangerous in terms of political pressures. Won't speculators try to push a stock they have purchased into S&P CNX Nifty? Or remove a stock from the index when they are shorting it?
Of course they will. Hence there are no speculators on the internal committee of IISL, which manages the index revisions. Further, there are objective, publicly defined rules which determine when stocks come in and go out of the index. There isn't much room for personal judgement here.
 

Index Funds
What index should be used for index funds?

From a mutual fund investors point of view, the fund manager should accurately replicate returns on the index. The liquidity filtering in S&P CNX Nifty and numerous operational details about index management, help ensure accurate tracking. When investors see that an index fund is unable to replicate the returns on an index, they would have dark fears and would abandon the product. S&P CNX Nifty is the best index in India in terms of the accuracy of tracking possible.

Index Funds
 
Index Funds today are a source of investment for investors looking at a long term, less risky form of investment. The success of index funds depend on their low volatility and therefore the choice of the index.

S&P CNX Nifty is used by a number of well know mutual funds in India for promoting Index Funds. These funds are:

(A) Index Funds :

1. IDBI Index I-NIT’99, an index fund scheme on S&P CNX Nifty launched by IDBI - Principal Mutual Fund in July 1999.

2. UTI Nifty Fund launched by Unit Trust of India in March 2000.

3. Franklin India Index Fund launched by Franklin Templeton Mutual Fund in June 2000.

4. Franklin India Index Tax Fund launched by Franklin Templeton Mutual Fund in February 2001.

5. Magnum Index Fund launched by SBI Mutual Fund in December 2001.

6. Prudential ICICI Index Fund launched by Prudential ICICI Mutual Fund in February 2002.

7. HDFC Index Fund – Nifty Plan launched by HDFC Mutual Fund in July 2002.

8. Birla Index Fund launched by Birla Sun Life Mutual Fund in September 2002.

9. LIC Index Fund – Nifty Plan launched by LIC Mutual Fund in November 2002.

10. Tata Index Fund launched by Tata TD Waterhouse Mutual Fund in February 2003.

11. ING Vysya Nifty Plus Fund launched by ING Vysya Mutual Fund in January 2004.

12. Canindex Fund launched by Canbank Mutual Fund in September 2004

13. Reliance Index Fund launched by Reliance Mutual Fund on Jan 2005

14. Principal Junior Cap fund launched by Principal PNB on May 2005

(B) Exchange Traded Fund :

1. NIFTY BeES an Exchange Traded Fund launched by Benchmark Mutual Fund in January 2002.

2. Junior BeES an Exchange Traded Fund on CNX Nifty Junior, launched by Benchmark Mutual Fund in February 2003.

3. SUNDER an Exchange Traded Fund launched by UTI in July 2003.

4. Liquid BeES an Exchange Traded Fund launched by Benchmark Mutual Fund in July 2003.

5. Bank BeES an Exchange Traded Fund (ETF) launched by Benchmark Mutual Fund in May 2004.

6. Benchmark Split Capital launched by Benchmark Mutual Fund on August 2005

What about index futures?
NSE commenced trading in Index Futures on June 12, 2000. The Nifty futures contracts are based on the popular market benchmark S&P CNX Nifty Index. S&P CNX Nifty is uniquely equipped as an index for the index futures market owing to (a) low market impact cost and (b) high hedging effectiveness. The good diversification of S&P CNX Nifty will generate low initial margin requirements. Finally, S&P CNX Nifty is calculated using NSE prices, and NSE is the most liquid exchange in India, thus making it easier to do arbitrage for S&P CNX Nifty index futures.


What is hedging effectiveness?
Suppose you have some portfolio, and you use index futures for hedging. A good index is one, which gives high hedging effectiveness, i.e. the index should correlate well with your portfolio -- whatever it may be. A good index would give a very high risk reduction when a portfolio owner short sells the index futures. S&P CNX Nifty correlates better with all kinds of portfolios in India as compared with other indices. This holds for all kinds of portfolios, not just those that contain index stocks.


Why not form a small portfolio of the ten most liquid stocks, and work to ensure that the small portfolio is maximally correlated with the S&P CNX Nifty?
This can, indeed, be done. Is it worth doing? That depends upon the cost and benefit. Calculating the weights, in the ten stocks with the lowest market impact cost, so that the correlation with S&P CNX Nifty is maximised, is not easy to do. (See Risk structure of Indian stocks by Dr. John Blin of APT for a calculation of a 10-stock portfolio which is maximally correlated with S&P CNX Nifty. This is found in the book The future of fund management in India, edited by Dr. Tushar Waghmare, Invest India - Tata McGraw Hill Series, 1997.) The gains from such an activity are not large. S&P CNX Nifty is explicitly designed to make it convenient to trade complete index portfolios. This is in contrast with other markets, where indices have arisen before index futures came about, and ways had to be found to trade them. For example, the S&P 500 index was there before index futures came about. When index futures started trading, arbitrageurs had to find ways to trade the index - trading 500 stocks on the floor-based New York Stock Exchange was highly cumbersome. This led to great creativity in finding 250-stock portfolios which correlate well with the S&P 500. In India, there is no need to undergo these kinds of problems. S&P CNX Nifty is the base of the index futures, and S&P CNX Nifty is designed to be convenient to trade directly.

Alternatives to the S&P CNX Nifty

How does S&P CNX Nifty compare with other indices?
Every technical reason favours the S&P CNX Nifty.

Diversification. S&P CNX Nifty is a more diversified index, accurately reflecting overall market conditions. The reward-to-risk ratio of S&P CNX Nifty is higher than other leading indices, making it a more attractive portfolio hence offering similar returns, but at lesser risk.

Liquidity. Over one year (October 1998 to October 1999), the trading volume on NSE for Nifty stocks was Rs.3.5 trillion, giving a liquidity ratio of 105%. The `liquidity ratio' is defined as trading volume over one year divided by market capitalisation today.

Hedging effectiveness. The basic risk of Nifty futures will be lower owing to the superior liquidity of Nifty stocks and of NSE. Nifty has higher correlations with typical portfolios in India as compared to any other index. These two factors imply that hedging using Nifty futures will be superior.

Governance. Nifty is managed by a professional team at IISL, a company setup by NSE and CRISIL with technical assistance from Standard & Poor's. There is a three-tier governance structure comprising the board of directors of IISL, the Index Policy Committee, and the Index Maintenance Subcommittee. S&P CNX Nifty has fully articulated and professionally implemented rules governing index revision, corporate actions, etc. These rules are carefully thought out, under Indian conditions, to dovetail with operational problems of index funds and index arbitrageurs. S&P CNX Nifty is relatively free of manipulation, for three reasons: (a) the index levels are calculated from a highly liquid exchange with superior surveillance procedures (b) S&P CNX Nifty has a large market capitalisation so the consequence (upon the index) of a given move in an individual stock price is smaller and (c) S&P CNX Nifty calculation intrinsically requires liquidity in proportion to market capitalisation, thus avoiding weak links which a manipulator can attack. Users of the S&P CNX Nifty benefit from the research that is possible owing to the long time-series available: both S&P CNX Nifty and S&P CNX Nifty Total Returns Index series are observed from July 1990 onwards. S&P CNX Nifty is backed by solid economic research and three most respected institutions: NSE, CRISIL and S&P.

How did the S&P CNX Nifty come about?
Equities trading at NSE began in November 1994. By late 1995, NSE became India's largest equity market and was looking for a market index to utilise this unique information source. NSE also wanted to have a vehicle for the futures and options market. NSE approached the economists Dr. Ajay Shah and Dr. Susan Thomas, then at CMIE (and now at IGIDR), to do research on methods in index construction. This