Like previous years, I have once again revised the **Nifty PE-Ratio & Return analysis **to include fresh data (up to December 2018). Now, this analysis has data spanning from early-1999 to late-2018, i.e. full 20 years.

This analysis uses one simple valuation metric (P/E Ratio of *Nifty50*) and attempts to correlate it to the returns achieved across various time periods (rolling-periods ranging 3 to 10 years) when investments were made at different PE levels of the index.

The purpose of this analysis is simple.

To arrive at some sort of conclusion and see how well or not-so-well correlated are the two things – the market’s current valuation and its future returns.

This time, I have tried to make this analysis more comprehensive and have included a few additional updates that weren’t part of the **previous years’ PE vs Return analysis**. So even if you have read the earlier ones, I suggest you go through this one again. My guess is that you will find it incrementally useful.

But before we get to the findings, let me say this upfront that this is not a sure-shot method to make money.

**Just because we can find some trends in past data doesn’t mean that the same trends will be replicated in future.** Markets are dynamic and the history is no guarantee of the future. In markets, the history is known to rhyme but it is never exactly the same.

The sole purpose of this analysis is to highlight that there exists a relation between the broader market valuations and returns you can achieve.

If you buy low (valuation wise), chances of earning good returns increase and vice versa.

**How to Analyse PE Vs. Returns?**

What I have done is that I have calculated returns earned on investments made at all Nifty PE levels starting from the year 1999. The time periods of return calculations are 3-year, 5-year, 7-year and 10-years.

Let’s use a simple example to understand this.

Suppose you had invested money in the index *Nifty50* on 24^{th}-February-2004 when its PE was 19.97 (actual data).

Now in the next 3-year, 5-year, 7-year and 10-year period (starting from 24^{th} February 2004), the CAGR returns would have been 29.3%, 8.5%, 17.0% and 13.0% respectively.

Using the above-described approach, I have done the return analysis for each and every trading day since 1^{st} January 1999 (the day from which the Nifty PE data is available publically). Obviously, I had several thousand data points for each of these periods.

The days that do not have forward returns for 3-years have not been considered in the analysis of 3-year returns. Same is the case with 5, 7 and 10-year studies. For example, data for 15-July-2013 is used for calculating 3-year return (as data is available for July-2016) and 5-year return (as data is available for July-2018). But the same is not used for 7-year and 10-year returns as data is naturally not available for July-2020 or July 2023 (at the time of writing of this study).

To simplify the findings, I have been grouping Nifty PE into 5 groups earlier.

But to provide more granular data this time, I have decided to present my finding in a larger 7-grouping set.

**Nifty-50 PE Ratio and Investment Returns**

The findings are based on 4243 data points for 3-year analysis, 3751 data points for 5-year analysis, 3250 for 7-year analysis and 2509 for 10-year analysis.

And here is what has been found after updating the dataset:

Due to the choice of intervals in the PE range, the figures differ slightly if I revert back to the old PE-range choice. Here is the same summary on the basis of the old grouping:

Essentially, both tables tell you the same thing – *when you invest at low PEs, your expected future returns are comparatively higher.*

Remember, these figures are based on past data (of the last 20 years).

The trends no doubt are easily evident here. But they may or may not repeat in future. There are no guarantees that the future returns will follow similar patterns.

**Markets won’t behave as you expect them to behave just because you have found its rhythm. You will not get returns just because you want them.**

But the above data set and relying on a common-sense based approach to investing tells that investing at low PEs is difficult but profitable.

So let’s say if one had the courage to invest in Nifty when PE was less than 12, the average returns over the next 3, 5, 7 and 10 year periods would have been an astonishing 39%, 29%, 22% and 18% respectively! Unfortunately, it’s very rare to find days where Nifty is trading at such low valuations.

On the other hand, if investments were made when *Nifty50* was trading at PE ratios of above 24 and above 27 (which are the supposed overvalued territories), the chances of earning decent returns in (atleast) near term are pretty low. This shows that if you invest in high PE markets, your chances of low (and even negative) returns increase substantially.

Investing at lower PEs can give bumper returns! But it is not easy. It takes a lot of ** c**ourage,

**ash and common sense to invest when everybody else is selling (in times of**

__c__**risis). This is what the**

__c__**3 C’s of proper investing**is all about too. It is very easy to sound smart and quote things like

*‘be greedy when others are fearful’*. Unfortunately, very few are able to be actually greedy when others aren’t.

Before we further slice and dice the dataset to find out more interesting things, let me highlight a few important points about using index data here:

- The Nifty PE data is published by NSE (here) and is currently based on Standalone numbers. This means that actual earnings (that includes those from subsidiaries, etc.) in consolidated figures are higher than what the standalone numbers would suggest. And that also means that current actual PE may not be as high as that suggested by the standalone ones.
- The constituents of
*Nifty50*keep changing. The index management committee that is responsible for maintenance of the index regularly brings in and moves out companies from the index. The Nifty composition of 2008 was quite different from that of 2018. Similarly, the index composition of 1999 might also be very different from that of 2008 and 2018. Try to understand it like this. If the index is made up primarily of companies that are low PE-types, then index at the overall level will tend to have low-PE. Whereas if the index is made up of high-PE companies, it will tend to have a high PE. So actual definition of high and low PE will be different for both type of companies, and so in turn for the index. A PE of 15 for a low-PE company might be very high whereas for a high-PE company might be very low. This is an important factor that should be kept in mind. I have addressed this to some extent in the latter part of this analysis by comparing returns while changing the periods under consideration from 20 years to 15 and 10 years.

Moving on, let’s address another important aspect here:

**Risks when dealing with Average Returns**

The table above gives a very clear relation between P/E and Returns.

But the above numbers are just ‘averages’. And that can be risky if you solely invest on basis of averages.

To explain this more clearly, let’s take an example.

Imagine that your height is 6 feet. Now you don’t know swimming. But you want to cross a river, whose average depth is 5 feet. Will you cross it?

You shouldn’t – because it’s the average depth that is 5 feet. At some places, the river might be 3 feet deep. At others (and unfortunately for you), it might be 10 feet.

That is **how averages work**. Isn’t it?

So this needs to be kept in mind…always.

To counter this, we need to analyze a few more things. So…

**Adding More Data points to PE-Analysis**

A better picture can be painted if in addition to the average returns, we also consider the following:

- Maximum returns during all the periods under evaluation
- Minimum returns
- Standard deviation
- Time spent in at a given valuation band

Have a look at the tables below now:

I want to spend some time here to highlight what all this means. I will also visually depict this using a small graph.

In the graph below, I have plotted PE ratios on X-axis and 3-year CAGR returns on Y-axis for all available data points in the last 20 years. As you can easily see, the trend is clear – returns are higher when investment happens towards lower PEs.

But you don’t always get the average returns that the preceding table shows. You can get anywhere between the MAX figure and MIN figure and the difference is huge.

Sounds confusing?

Let me try to segregate the above chart into PE-bands to highlight this:

As you can see, within each PE band, the actual returns are all over the place.

Focus on the first green block of PE12-15 band.

The red circle is the position of average return (38.7%).

But as the two text bubbles show, the Max and Min returns are very different from the average returns. Refer to the table above and you will find that max is about 58.1% and min is about 5.9%. And all this with a standard deviation of 14.6%.

What this means is that even though the average figure might tell you something, the actual returns can vary a lot.

I hope you get the drift of what I am trying to show you here…

Let’s now see the data for 5, 7 and 10-years:

Spend some time studying the above tables.

By now, you would be clearly noticing that there are big differences between the minimum and maximum returns for almost all periods.

So even though the average return figures will tell you one thing, the fact is that the actual returns that you get will depend a lot on when exactly you enter the markets and what happens afterwards.

Two different investors entering the markets at same valuation levels (let’s say PE=17.5) but at different times would have got 7-year returns ranging from 7.7% to 26.1%. Check the table for 7-year return above and you will know how.

For a lack of better word, the difference is shocking.

What this means is that even though the average return for 7-year period in the example taken above (PE17.5) is 16.2%, the actual returns have ranged from 7.7% to 26.1%.

This is equity investing for you. Welcome!

The statement that I made a couple of paragraphs before, *‘returns that you get will depend a lot on when exactly you enter the markets’*, does sound like trying to time the markets. But this is a reality. I cannot deny it. For those who can, timing the market works beautifully.

Hence even though the average returns give a good picture for long-term investors (look at the table for 10-year analysis), its still possible that you end up getting returns that are closer to the ones that are shown in minimum (10Y Returns) column and not the Avg. Returns. 🙂

This is another reason why I introduced the column for standard deviation in all tables above (see the last column of the table above).

Analyzing standard deviation tells you – how much the actual return can vary from the average returns. So higher the deviation, higher will be the variation in actual returns. For completing the scatter analysis, here are the 3 remaining scatter plots for 5, 7 and 10-year returns:

**Can You Catch the Markets at Right Investible PE?**

Ideally, and armed with the above insights, it makes sense to buy more when valuations are low. Isn’t it? Buy Low. That is the whole idea of investing.

But real life is not that simple.

It is very difficult to catch markets on extremes. It’s like a pendulum – it keeps oscillating between overvaluation and undervaluation. It is almost never perfectly valued.

So should you wait to only invest at low PEs? Though it might make theoretical sense to do so, it is still very difficult to wait for low PE markets. And extremely low PEs are extremely rare.

Just have a look at this 5-year table I shared earlier in this post:

Look at the column *‘Time spent by the Nifty in PE band’ *at Below-PE12 levels.

It is just about 1.5% of the time since 1999 that markets spent below PE12. This is extremely rare.

For common investors, it’s almost impossible to wait for such days. In fact, such days might be spaced several years apart!

So the best bet for common people is to keep investing as much as possible, via disciplined investing (like **SIP in equity mutual funds**). It is not perfect (like buy low sell high) but it is your best bet given all the constraints. And once your portfolio grows in size, make sure you rebalance it periodically to adhere to proper asset allocation and manage risks appropriately.

**Long-Term Investors have Better Chance of Doing Well**

Another insight that this study gives is that as your investment horizon increases, the expected returns more or less are reasonably OK-ish to good enough, even when one invests at high PEs.

Have a look:

So, even if an investor puts his money in the index at PE above 24, the historical average returns are more than 8 to 10%. That’s quite ok I guess. Atleast it’s much better than being in losses.

The longer you stay invested, higher are the chances of not losing money in stock markets…even if you have entered at comparatively higher levels.

*Caution – I know that’s a dangerous statement to make but to keep things simple, please read it in the right spirit and get the drift.*

Now let’s compare this with someone who is thinking to invest at high PEs (above 24) for less than 3 years. Have a look at the table below:

There are un-ignorable chances that the person will not do well. Chances of losing money are fairly high if you see the average figures and Minimum returns for PE24-27 and PE>27 bands.

But let me touch upon an important point now.

**Will this Result change if only the more Recent Data (and not last 20 years’) is considered?**

That is no doubt an interesting and fairly valid point.

In last few years, it does seem (and I repeat ‘seem’ – may be due to our recency bias) that average PE levels are much higher than what they used to be in earlier years. The Reason for this may be many:

- One of them can be that unlike earlier years, the constituents of Nifty50 are companies which in general have high PEs. So obviously there is a case for slightly higher average PE figures as the new normal.
- The PE figures are based on standalone numbers of constituent companies of the index. If we consider the consolidated numbers, then chances are that the earning would be higher and lead to lower (calculated) PE figures. Earlier, the difference between standalone and consolidated figures wasn’t much as Indian companies did not have large subsidiaries that would distort the figures. But as Indian companies grow and so do their subsidiaries, the consolidated figures will be incrementally bigger than standalone ones.

To accommodate these facts, it makes sense to give more weight to consolidated figures. But NSE publishes data on the basis of standalone numbers till now. So we live with it for time being.

Another option is to reduce our period in consideration from 20 years to a more recent one – like last 10 or 15 years – where the comparative difference in the nature of index constituents in not as stark as what might be (like) 20 years ago.

Ofcourse the number of data points would reduce. But we can atleast have a slightly more relevant comparison if not a perfect one. So I wanted to try out the above analysis with different (but more recent time periods).

As a first case, let’s consider the **last 10 years data, i.e. Jan-2009 to Dec-2018**.

Do note that the beginning of this period coincides with the depth of the last severe bear market. So numbers will change. Let’s see:

You would agree that the trend remains the same. At higher valuations (PE), the future returns reduce.

I have __not__ done the analysis for 7- and 10-year period as the period under consideration (i.e. 10 years) is too small to have any sufficient number of data points for the 7- and 10-year analysis.

But how does this compare with our previous analysis where period under consideration was 20 years? Let’s see:

Spend some time comparing the above 2 tables. The left one is based on the analysis of the last 10 years and the right one is based on the analysis of the last 20 years.

The basic conclusion once again is the same.

But the extent of these returns changes when we the period under consideration is changed.

How?

Focus on the red arrows for now. The returns for PE12 case moderate from 38.7% to 23.0% (for 3-year) and from 29.2% to 18.4% (for 5-year) when we change the analysis period from 20-year to 10-year.

Now focus on the green arrows. The returns for PE18 and above (i.e. PE18 to 21 to 24 to 27 and beyond) increase somewhat for both 3-year and 5-year analysis when we change the analysis period from 20-year to 10-year.

This might be getting slightly number heavy but what I want to highlight here is that depending on the data set I chose, the return figures change to some extent (increase and reduce for different PE bands).

The overall conclusion still remains the same. But the expectations need to be revised when we give more weight to the recent past (last 10 years) than what we gave to the full last 20-year period.

You might feel that by choosing the last 10-year period, I am missing out on the Bull Run between 2004 and 2007. And rightly so.

So let’s consider a different period now:

The last **15 years (from Jan 2004 to Dec-2018)**.

This period takes into account the great bull run of 2004-07, the big crash of 2008-09 and the upmove since then 2009-2018.

Let’s see what the data tells here:

I may sound repetitive here – but the overall trend is still the same.

But once again, what changes though is the extent of returns in each case.

Let’s now compare 3-year and 5-year investment returns for each of the PE bands when different analysis periods of 20, 15 and 10 years are considered:

As is clearly evident, the figures moderate if we reduce the analysis period to data of the last 10 years instead of 20 years.

To be fair, there is no perfect answer as to which one should be used or which shouldn’t be. But this moderation analysis proves that we should revise our expectations to more rational levels. Due to various factors (like increasing difference between standalone and consolidated earnings and PE figures, change in index constituents and their normal valuations), we may have to keep a range of outcomes in our mind when making investment decisions. It was never black and white. It was always grey. Now it has even more shades of grey!

For the sake of completeness, I am tabulating the full findings for everything below. This includes 3-year, 5-year, 7-year and 10-year Nifty return analysis compared across various PE levels and after considering data for different sample space of 10, 15 and 20 years:

Also, note one more thing – this analysis is based on one time investing at specified PE levels.

If you are a SIP investor, then returns will obviously vary as your investment would be spread out across time and PE ranges. You cannot and should not expect to receive lumpsum-like returns on your SIP investments. The mathematical concept of average doesn’t allow that to happen.

**Asking Again – Whether This Approach works in all kinds of investing?**

My answer is that no one strategy can work in all conditions.

Knowing the broader market PE gives a fair idea about the valuations of the overall markets. It tells you when the market is overheating and that you should take cover (reduce equity in line with asset allocation). This, in turn, helps reduce the chances of making mistakes when investing.

Similarly, this knowledge of PE-Return Relationship also helps in identifying when markets are unnecessarily pessimist. If you are brave at such times, you can make some serious money.

**And please don’t think about investing in individual stocks just because Nifty PE is low. Individual stocks have their own stories and need more in-depth analysis.**

I have been doing this analysis now for several years. If you wish to access old ones done in the year 2017, 2016, 2015, 2014, 2013 and 2012, then please access the **archives**. But I think its best to stick with this current analysis as it’s the latest and most comprehensive one till date.

I regularly update PE and other ratios of *Nifty50* and *Nifty500* on **State of the Indian Market****s** page.

**What Should You Do as a Common Investor?**

I am assuming that you are not Warren Buffett. 🙂 But jokes apart, the fact is that most people do not have the skill or time to get into deep investing.

So what should such people do?

First thing is to simply stick with a regular disciplined way of investing (easily achievable through **MF SIPs**). A proper way to do it is to first **find your real life goals** that require money (use this **free excel**) and then stick to **Goal-based Investing**. This is more than enough to begin with.

And once your corpus size grows, you should regularly rebalance your portfolio to de-risk it when needed and to position it for better risk-adjusted returns in future.

With that taken care off, you should try to **invest more when market valuations are low**. This will help increase your overall returns in the long term.

This is easier said than done but this is what really works in the market.

To sum it up…

**There is a reasonable (but not guaranteed) correlation between the trailing PE and Nifty returns. And this study proves it and provides some useful insights. If we were to go by the historical data, the Nifty delivers higher return (in long-term) whenever investment is made at low PE ratios. On the other hand, it tends to deliver low to negative returns whenever investment is made at high PEs and when the investment horizon is short. You as an investor can use this insight as a backdrop to take your investments decisions.**

I hope you found this detailed and comprehensive PE-Return analysis useful.

I will try to revise this study with new data points and other insights as and when practically feasible.