INTRODUCTION

Why do we take risk? Barring some adrenaline junkies1, most of us take risk in the search of rewards that we otherwise wouldn’t gain. Over the Christmas holidays, I read Nigel Cliff’s The Last Crusade, The Epic Voyages Of Vasco da Gama. Da Gama and his crew took great personal risk to find a direct sea route to India in search of the most profitable trade routes in the world at the time. For instance, these voyages were punctuated with brutal wars and looting and it is estimated that on his first successful voyage to India, Vasco da Gamma returned with less than half the crew he started with. Many romanticise this period as an age of “discovery”, however, there is little doubt as to what motivated these men and their kings to take these risks – the possibility of untold riches which would go on to forever change the fortunes of Europe.

This got me thinking about equity risk premiums and the risks we take today in search of riches. Even though global trade has become much more sophisticated (and thankfully much less gruesome), I find it fascinating how little has changed in the fundamentals of risk and reward. In this post, I’ll try and gain some insights from the last 60 years of equity risk premiums. But first, what is an equity risk premium?

EQUITY RISK PREMIUM: THE PRICE OF RISK

It is intuitive that investors should demand a premium as compensation for investing in riskier assets. An equity risk premium, therefore, is the return that investors demand over and above the risk-free rate 2 for investing in equities as a class. It is the one number that captures all the emotions of equity investors – fear and greed, hopes and dreams – at any given point of time.

But why should investors demand an extra return for investing in equities? If we’re happy with, say a 5% return on government bonds, shouldn’t we be equally happy with a 5% return on equities. Well, the chart below shows us why. The green bars represent market’s expectations of earnings growth for the 500 largest companies in the US over the next 5 years whereas the blue line represents the actual earnings growth over that period.

It should come as no surprise that reality is more volatile than expectations. After all, equity is a residual claim on the profits of a business as opposed to a contractual claim like debt. By definition, therefore, it will be more volatile than a contractual claim. It carries potential upside through earnings growth but it also comes with the risk that these earnings and growth are not guaranteed. This is the reason why investors should and, in fact, do demand a premium for investing in equities.

Types of equity risk premiums

There are two types of equity risk premiums – historical and implied. As the name suggests, a historical equity risk premium looks at the past returns that equity markets have delivered over the risk-free assets like government bonds. Its estimation requires judgments on the time frame over which you compare equity market returns relative to return on risk free assets like government bonds. One of the drawbacks of a historical risk premium is that it is backward looking and the standard errors on historical premiums are so large that they render any kind of statistical analysis meaningless.

Unlike the historical equity risk premium, an implied equity risk premium is computed in real time and is forward looking. It incorporates investors’ expectation of future cash flows as well as the price investors are currently paying for those expectations. I won’t bore you with the details of how an equity risk premium is computed but suffice to say that it borrows from the concept of Yield to Maturity on a bond. However, unlike the bond which has fixed cash flows (coupons on the bond), cash flows on equities (dividends and buybacks) are not fixed and, therefore, estimating them requires assumptions around growth3. We then discount these future cash flows at a rate that equates the present value of these cash flows to the current level of the index. People who’ve studied finance will realise that this is similar to the internal rate of return used to analyse projects in corporate finance. When we subtract the risk free rate from this discount rate, we’re left with the implied equity risk premium. For anyone who is interested, I have previously done a detailed post on computing an Implied ERP in the context of the Indian market here.

Insights into investor psychology

A lower ERP suggests that investors are feeling confident about the future and are therefore, willing to take on more risk and pay a higher price for future expectations. This is a euphemistic way of saying that investors are feeling greedy. Similarly, a higher ERP suggests that investors are fearful and uncertain about the future. Hence, they are collectively demanding a higher return from equities than they otherwise would.

So with that basic understanding of equity risk premiums, let’s analyse these premiums in a bit more detail.

IMPLIED EQUITY RISK PREMIUM: SETTING THE STAGE

Prof. Damodaran’s dataset on implied equity risk premiums goes back to 1961 and since the 2008 crisis, he has been computing this number on a monthly basis. The graph below plots the equity risk premium at the end of each year beginning with 1961.

Looking at the historical data, we can make the following observations about ERPs in general:

  • Stating the obvious, implied equity risk premiums are dynamic. There are times when investors get carried away and demand too low a risk premium for investing in equities and others when investors shun equities and demand a much higher premium than warranted4.
  • There’s a tendency for the implied ERP to be mean reverting and it has averaged 4.25% during this period. However, its range over this period has been wide. It’s been as low as 2.1% at the end of 1999 during the famous dot com bubble and reached a high of 6.5% in 1979 on the back of high inflation and the oil shocks of the 1970s. It again tested those high rates during the housing market meltdown of the GFC in 20085.
  • Lastly, even though the ERP has been mean reverting, there have been prolonged periods when it has deviated from its average. It was consistently below the average for much of the 1960s and has been consistently above the long term average post GFC. It is only very recently that ERPs have trended down toward their long term average of 4.3%

PREDICTIVE ANALYSIS OF ERPS

One of the benefits of having a rich database that’s been created in real time is that we can assess its usefulness by running predictive analysis with little hindsight bias. Recall the logic of the ERP – a high ERP suggests that investors are uncertain of the future and are demanding a higher expected return for investing in equities and vice-versa. Therefore, implied equity risk premiums should have a positive correlation with future equity returns, i.e. a high equity risk premium at a given point in time should lead to high returns in the future and vice-versa. In the next section, I am going to test this hypothesis.

ERP and shorter time periods

I am going to first test it over shorter time periods, say 1 year. I ran a regression with equity risk premiums as the independent variable and the return on the S&P 500, the following year.

1 year forward return=−0.06+4.18×ERP  where n=63; t Stat=2.27; R sqaured=7.8%

Firstly, notice the positive sign (+) with the coefficient of the ERP (+3.88). This means that next year’s return is positively correlated with the ERP – a higher ERP today predicts a higher return 1 year from now and vice-versa. This is good since this flows from the logic of the implied ERP that we discussed earlier. The t-Stat > 2 is also good since it suggests that the correlation is statistically significant. However, if you think of equity risk premiums as the magic bullet that will hep you predict what the market will do next year, I am afraid you’re in for disappointment. Notice the R squared at just 8%. This means that the implied ERP only explains 8% of the variation in next year’s returns or to put it differently, about 92% of the variation in 1 year return on the S&P 500 is explained by factors other than the implied equity risk premium.

We can see this with the help of a scatter plot which brings the regression equation to life. This chart plots the implied equity risk premiums on the X-axis and the returns on the S&P 500 in the following year on the Y-axis.

As you can see, even though there is a positive correlation between ERPs and the following year’s equity return, it is weak at best. If we analyse longer time periods, however, it is a slightly different story.

ERP and longer time periods

Now instead of using 1 year returns on the index, we’ll use returns on the S&P 500 over the following 5 years.

5 year forward return=−0.03+3.22×ERP  where n=59; t Stat=4.34; R sqaure=24.8%

In this instance, not only is the correlation statistically significant (t Stat > 2), the ERP explains more of the variation in long term stock market returns relative to returns over shorter periods, almost 25% in this case6. This is even clearer in the scatter plot below, which is more tightly spread compared to the chart we looked at previously for 1-year equity returns.

The odds of success at various levels of the ERP

Rather than think about ERPs in precise terms, it is more helpful to frame the discussion in terms of probabilities. Notice that whenever the equity risk premium has dropped below 3%, it is almost always followed by poor equity market returns over the next 5 years. Thus, as per historical data, the odds of positive returns over the next 5 years are greatly diminished whenever the equity risk premium has fallen below the 3% level. This makes sense because an ERP of less than 3% leaves little margin of safety and even the slightest hiccup causes the markets to correct.

On the other hand, whenever the equity risk premium has crossed the 5% hurdle, the market has never delivered a negative 5-year return and the odds of success are tilted in favour of the equity investor. Does this mean that the index will never deliver a negative return when the ERP crosses this threshold? Of course not. All it tells you is that the odds of that happening are really low.

As you would expect, anything within these two extremes is not as clear. These levels would suggest that markets are somewhat in balance and not being driven by extreme sentiments, either positive or negative.

One more thing…

There’s one final way I want to slice and dice this data. The following chart shows the number of years the equity market had a more than negative 10% real return over a 5-year period at various levels of the ERP. For example, the left-most data point means that when the ERP was close to 2% (1999), out of the 5 years that followed, the equity market had 3 years with more than negative 10% real returns. Why did I choose a 10% threshold as opposed to any other? Well, it’s an arbitrary choice to be honest but I did want to exclude years that were only slightly negative and a 10% decline in the index is a meaningful correction.

Again, it’s not a surprise that difficult years are on the horizon when investors collectively demand too low a risk premium for investing in equities.

A DETOUR: ARE AVERAGE ERPS STRUCTURALLY HIGHER IN A POST GFC WORLD?

Before we get to the conclusion, I want to briefly acknowledge the possibility that equity risk premiums may be structurally higher in a post GFC world. The difficulty in finance is that all of our data comes from the past but what we really care about is where we are heading in the future. Normally, the past can be a useful guide of the future but, there are times when a structural shift can make conclusions drawn from past data irrelevant, especially in finance. Therefore, any insights drawn from past data should be taken with a grain of salt and shouldn’t be upheld as gospel.

Remember when we discussed earlier that the average ERP since 1961 has been 4.3% and that ERPs since 2008 have been consistently higher than this average. The average is closer to 5.5% instead of 4.3% if we only consider the last 16 years of data (post GFC).

Nobody can say for sure whether the “higher than average” ERP experienced since 2008 have been excessive. One explanation for these higher ERPs is that they reflect a structural shift in the developed world. The events of 2008 woke investors up to economic risks which were previously considered unthinkable in an economy like the US and other western economies. With the benefit of hindsight, we now know that outside of the big tech, these countries have struggled for economic growth in the period post the GFC. Hence, we need to be careful to hang onto long term averages when making decisions about the future.

SO WHERE DO WE STAND TODAY?

The equity risk premium as of December 2024 is 4.1%. This is not bubble territory yet which I define as anything approaching an ERP of 3% or less and we might not get there anytime soon. However, if you were to put a gun to my head, I’d say that the market is probably overvalued at this stage and the one thing that makes me take a bit of a step back, is that this ERP is built on earnings growth expectations reaching close to 10% at an index level over the next few years.

If we have an earnings hiccup and the lofty growth expectations don’t materialise, this ERP would prove to be woefully inadequate. Of course, we will know this only in hindsight and I am not in the game of predicting the next economic recession. Remember at the end of 2022 when most experts were confident of a pending recession on the back of higher interest rates and high inflation but what followed was 2 years of a resilient economy, low unemployment in the US and strong equity market performance.  Maybe the economists were just early and not wrong and we do end up getting a recession after all but, I don’t know. What I can be more confident of is that after two years of greater than 20% returns on the index, finding attractively priced ideas has become harder and that returns from here are going to be driven by good stock picking rather than a general advance in the market.

Happy investing!

Footnotes

  1. You could argue that the reward for an adrenaline junkie is the rush that he/ she gets doing something risky. ↩︎
  2. A risk free asset is one that fulfils the following two conditions a) there is no default risk; and b) there is no reinvestment risk. ↩︎
  3. We use earnings growth forecasts by analysts since those are the expectations being built by market participants. ↩︎
  4. The adequacy or not of an equity risk premium is known only in hindsight. ↩︎
  5. There have been intra year periods where the ERP has crossed well above 7% as well. ↩︎
  6. If we look at 10-year forward returns, the relations becomes even stronger with equity risk premiums explaining close to 40% of the variation in those returns. ↩︎