Index funds and ETFs are becoming a popular investment choice. In terms of assets under management, they are way behind actively managed funds but the momentum seems to be building in their favour. We will not talk about pros or cons of index investing; there is enough material out there. This will be a 2-part series. In part 1, we will look at why investing in Nifty 100 is not the same as investing in Nifty 50 and Nifty Next 50.
This is the index construction of NSE India. For our discussion here, we will focus only on the green box:
More than 2000 companies are listed on NSE. Nifty 100 is made up of the top 100 companies by market cap. This group constitutes more than 75% of the total market cap. Likewise, Nifty 50 comprises of the top 50 companies by market cap and Nifty Next 50 is made of companies from 51 to 100. And yet here I am saying, to the contrary! Sure, Nifty 100 is sum total of Nifty 50 and Next 50 but it is not the same thing when it comes to investing. Meaning, a rupee invested in Nifty 50 and Nifty Next 50 each is not the same as investing 2 rupees in Nifty 100. To understand better, let us look at relative performance of these three indices:
We see that Nifty 50 and Nifty 100 are moving in tandem whereas Nifty Next 50 is not. Here one might argue saying this is point to point data, which is prone to narrative bias. So, let us see the rolling returns.
Even here we notice that Nifty 50 and Nifty 100 are moving in tandem but Nifty Next 50 is either over-performing or under-performing the other two. For the sake of brevity, I have put 3-year rolling return data. 1 year, 2 year or 5 year rolling returns have similar pattern.
Traditional wisdom says that, if something is made of two parts, the final product is the average of the two. In this case, Nifty 100 is made up of Nifty 50 and Next 50 and yet it is nowhere in the middle. In fact, it is leaning heavily towards Nifty 50. Why is that? The answer lies in the way the indices are constructed. These are free-float adjusted market-cap weighted indices. To put it simply, first calculate free float for each constituent. Meaning, shareholdings of promoters, group company’s cross holdings, strategic investors, employee welfare trust etc. are taken out. What remains is free-float. Next, determine the market cap for each stock with this free-float. This is done simply by multiplying close price with the free-float. Finally, index is calculated by (Current Value) / (Base Value of Index). Base Value is taken as 1000 as on 3-Nov-1995. Let us see with an example to simplify things a little: Weightage is determined by 3 factors:
But it still does not answer why does Nifty 100 behaves more like Nifty 50 and not as an average. To understand that, let us break down the constituents of Nifty 100 into 5 buckets of 20% each, in descending order of their weights:
Although Nifty 100 has all the stocks of Nifty Next 50, the weight of those dwarfs in comparison to Nifty 50. No wonder Nifty 100 behaves more like Nifty 50 rather than an “average” of its two constituents.
Indexing is the way forward; there is no denying that. In fact, mutual funds have also realized this and been launching ETFs, in many different hues and themes. Not all are investment grade. Some are just decorative and some with hardly any AUM to speak of. We have seen than Nifty 100 behaves like Nifty 50. So, there is no reason why Nifty 500 ETF should be any different. It does not offer any meaningful diversification. Nifty Next 50 on the other hand does provide diversification to an existing Nifty 50 fund.
What can investors do? Most large cap mutual fund portfolios have the Nifty 50 stocks to a large extent. After all fund managers also have to chase alpha. Investors can consider Nifty Next 50 index funds as a diversification strategy to park any incremental investments. As we have seen, a combination of Nifty 50 and Next 50 provides better overall returns than Nifty 100.
In the next part, we will try and figure out if this simple index strategy could be refined further. And whether investors can generate better returns with some minor tweaks. Until then, happy indexing.