Nifty, Sensex PEs calculated on standalone basis pulling down EPS.
The most common valuation ratio is the PE (Price-earnings). The most common indicator of stock market health is a broad index reflecting the combined effect of price changes in many stocks.
It is quite logical to use an averaged PE for any such index, to benchmark overall market valuation. Most major indices come with pre-calculated PE values, along with other common ratios like dividend yields and Price-Book Value (PBV).
Investors and traders often use these numbers without delving into the nuts-and -bolts of methodology. However, the methodology is worth knowing because sometimes those numbers can be misleading, or investors can end up talking at cross-purposes because they are unaware of the underlying statistical logic.
I’m wondering if this is true for the Nifty (and the Sensex) with respect to valuation calculations. The Nifty is a normalised weighted average of the top 50 listed stocks in terms of market capitalisation. (The following numbers are illustrative and may bear no relation to the actual values on a given date).
Thus, if Reliance Industries (RIL) market cap comprises say, 12 per cent of the total market cap of the 50 Nifty stocks on a given date, RIL’s price changes will contribute 12 per cent to the Nifty’s price changes on that date. Similarly, if Hindalco contributes 1.5 per cent of market cap, Hindalco’s price changes will contribute 1.5 per cent. This is recalculated dynamically.
The same market cap weights are also used for calculating PE, PBV, dividend Yield and so on. The Nifty uses the last four trailing quarters of declared results for these. There is constant updating involved here as well, since different companies have different financial year-endings and display different degrees of promptness in reporting results.
So far, the logic is clear enough. As investors, we can slice and dice the historically reported ratios in different ways to get a view on how over-valued or under-valued the index may be at a given time. For example, we may calculate what the average market PE has been over a long period, the median PE, the standard deviations and so on.
Analysts sometimes deliberately fudge, when the historical data presents an inconvenient picture, by referring to forward earnings, instead of the trailing earnings reported with the indices. Thus, an analyst will say “The Nifty is trading at a one-year Forward PE of let’s say 15, and that’s attractive”.
Forward earnings are projections and hence, subjective. They may paint a rosier, or gloomier picture than the historic reality. However, once we know the historic pattern, we also know how optimistic or pessimistic the forward projections that are being presented are, by simply calculating the difference between trailing actuals and forward estimates.
The really large problem lies in the rationale of the results considered when calculating earnings. Companies report “Standalone” and “Consolidated” results. Consolidated results take the performance of subsidiaries into account. If a company has many subsidiaries, or its subsidiaries have large turnovers, there will be a significant difference.
Most Indian companies have multiple subsidiaries. First, Indian tax law and company law encourage the creating of complex structures (so do laws everywhere else). Second, many Indian firms have overseas businesses. Third, mergers, including overseas mergers, occur often enough. Fourth, it’s common in infrastructure industries to work via SPVs (special purpose vehicles) and JVs (Joint ventures).
This is normal to any aspiring transnational business environment. As an investor, one studies consolidated results with far more diligence. It would be ridiculous to invest in Tata Steel, while ignoring Corus. Similarly, ONGC Videsh’s performance (OVL) has a large bearing on ONGC, and Jaguar’s performance impacts Tata Motors.
If the subsidiaries are unprofitable, the consolidated EPS may be lower. Vice-versa if they are profitable. A back-of-the-envelope calculation suggests the differences between standalone EPS versus consolidated EPS for the Nifty basket are often around 15-20 per cent. The trend is towards further exposures overseas and more infrastructure plays. So the number of subsidiaries will increase, and consolidated numbers will become even more crucial.
Unfortunately, both the Nifty and Sensex PE are calculated on the basis of standalone results. Adjusting for the differences on a daily basis would be a tedious task. Nor is it possible to easily adjust for the historical differences, which is critical for any statistical analysis of this index valuation data.
In 2010-11, the consolidated results showed much higher EPS – most subsidiaries were profitable. So the current official trailing Index PEs should be adjusted down. But this may not be the case in FY 2012 or FY 2013. All one can do is hope the authorities will change methodology and back-calculate as well.
