Index Investing and Systematic Investment Plans (SIPs) are becoming increasingly popular forms of investing. Putting aside some money every month in a low cost index fund is a great way to get exposure to equities and build wealth over the long term. However, I have always been curious about rules based passive investing i.e. making index investing decisions based on simple valuation metrics (buying more when certain valuation indicators indicate undervaluation and buying less when they indicate overvaluation).
In this post I try to see whether the Nifty50 Price to Earnings (P/E) ratio, a commonly used valuation metric, can be used to formulate a simple rules-based passive investment strategy.
The first thing we notice is that the Nifty50 P/E ratio has been quite volatile, ranging from 11 to 30 and averaging 20 over the last 2 decades. We also notice that when the ratio goes to either extremes, it promptly comes back to its average or normal level. On the lower side, when the P/E drops below 15, you see a steep rise in the subsequent months and on the higher end if it crosses 25, you tend to see a steep correction.
There are two ways for the P/E to correct. Either earnings can go up or the price can go down. For investors buying at high P/E ratios, the justification often given is that earnings are likely to go up and hence the P/E will come down (“I am not paying for current earnings but paying for future earnings”). So in trying to formulate a rules based index investment strategy we need to ask that when the P/E ratio reaches extremely high levels, do earnings catch up or does the price correct? After some trial and error, I settled on the P/E levels of 19 and 25, which seem to indicate levels of undervaluation and overvaluation on the Nifty50. The below graph explains why.
The graph shows the movement in the Nifty50 over the last 20 years, highlighting points in green when it traded at a P/E of less than 19, points in orange when it traded at a P/E between 23 and 25 and points in red when the P/E was over 25. If you invested at the points in green, you were almost certain to catch an upcoming bull market rally. Sometimes the P/E remained below 19 for a long time but if you continued investing through this period and were patient, you were likely to be rewarded handsomely.
A P/E of >25 was a sure sign of danger. Before 2017, the Nifty50 P/E ratio crossed 25 only twice (in Oct 2007-Jan 2008 and Sept-Nov 2010). Both times we saw a subsequent correction of over 25% in the next year. In fact, before 2017, a P/E of 23 was also a pretty reliable indicator of an upcoming correction. It would have predicted the correction between Feb 2015 and Feb 2016 and also the one between July and December 2016. So before 2017, if you saw the Nifty50 P/E crossing 23, you were sure to see lower levels on the Nifty50 at some point in the next 3-6 months. Sometimes it took longer and you could see the Nifty50 rally even more and the P/E being stretched to 26-27-28 levels but if you were patient enough you would most definitely be able to buy the Nifty50 at lower levels.
We see a peculiar trend emerging after 2017. The P/E crossed 23 in February 2017 and 25 in July 2017 and did not come down for the next 3 years. It rose all the way to 29.9 in June 2019. It stayed at these elevated levels till the crash of March 2020. People blame the March 2020 crash on the Covid-19 pandemic but we should be open to the possibility that the market was just correcting to reasonable valuations after trading at expensive valuations for a very long time. After a brief correction, the Nifty50 has risen sharply and as of 17 July 2020 is trading at a P/E of 28.55.
Earnings Yield – Inverse of the P/E ratio
Another metric I like to use is the earnings yield (E/P), which is the inverse of the P/E ratio. If the P/E ratio is 25, the earnings yield is 4% (1/25). It basically means that if I shell out INR 100 to buy the Nifty50, I get INR 4 in profits every year. I like to compare this earnings yield to the risk free rate or what I can earn from owning a government bond. In India, even today it is possible to get 6-6.5% on a government bond. Taking the risk of holding equities and earning a 4% yield versus owning relatively risk-free government bonds that give you 6% does not sound very attractive to me.
One may argue that it may still be worth owning equities because these earnings grow where as the bond coupon payments don’t grow. My question to that is “How much will earnings grow and how fast will they grow?”. Nifty50 earnings need to grow by 50% (from INR 4 to INR 6) just for the Nifty50 to give you the same yield as a government bond. How much time will it take for earnings to grow by 50%? It may be difficult for the largest 50 companies in India to grow their earnings by 50% over a short period of time.
On the other hand if you buy the Nifty50 at a P/E of 19-20, you are purchasing at an earnings yield of 5%-5.3%. This is closer to the risk free rate and even moderate growth in earnings will help push the yield to above the risk free rate.
At the current P/E of 28.55, investors seem to be happy with a 3.5% earnings yield. Nifty50 earnings need to double for the yield to be higher than the risk free rate. With the ongoing pandemic expected to dampen earnings at least for another year (maybe more), investors need to question where the earnings growth will come from. Will we see another correction that will bring the P/E to a more normal level? Or will markets continue to trade at steep valuations (like it has for the last 3 years) till we are out of the crisis and earnings growth comes back? Only time will tell.
Key Questions for Investors
- Is this the new normal? Have the overvaluation and undervaluation indicators changed?
Before 2017, levels of 19 and 25 worked very well as indicators of the Nifty50 being relatively cheap and expensive. But in the last 6 years the Nifty50 has traded below a P/E of 19 only for two brief periods (February 2016 and March 2020). It is possible that in the sustained low global interest rate environment that we have seen over the last decade (and are likely to see at least for the next couple of years) the “normal” level of the P/E has moved higher and the cheap and expensive levels going ahead could be 21 and 26 (or some other numbers).
- How relevant is the P/E ratio?
The P/E ratio is just one of a plethora of valuation metrics that are used. The P/E has its own drawbacks as a valuation metric. It may be worth replicating this study with other metrics such as Price to Book or Dividend Yield (both easily available from the NSE website)
- Is this too short a history to form conclusions?
The Nifty50 index is a little over 2 decades old, which is short compared to indices in the US which have a much longer history. Perhaps our data set is too short to come up with reliable levels to indicate a cheap or an expensive market.
Here is the excel with the Nifty50 historical data downloaded from the NSE website that I used to do the analysis. If you are interested then download it, play around with the levels and let me know if you find any other levels that are more suitable for a rules based investing approach.
Please note that this is not investment advice. I am not a financial advisor. Please do your own research or consult a financial advisor before making any investment decision. Please assume that I have a position in any stock that I write about.