Monday, August 8, 2011

Momentum versus Contrarian: Two Reads of the ERP

I am not much of a market timer but there is one number I do track on a consistent basis: the equity risk premium. I follow it for two reasons. First, it is a key input in estimating the cost of equity, when valuing individual companies. Second, it offers a window into the market mood, rising during market crises.

For the ERP to play this role, it has to be forward looking and dynamic. The. conventional approach of looking at the past won't accomplish this. You can however use the current level of the index, with expected cashflows, to back out an expected return on stocks. (Think of it as an IRR for equities.) You can check out the spreadsheet that does this, on my website ( on the front page.Here is the link for the July 1 spreadsheet. Just replace the index and T.Bond rate with the current level and use the Goal seek in excel:

A little history on this "implied ERP": it was between 3 and 3.5% through much of the 1960s, rose during the 1970s to peak at 6.5% in 1978 and embarked on a two decade decline to an astoundingly low 2% at the end of 1999 (the peak of the dot com boom). The dot com correction pushed it back to about 4% in 2002, where it stagnated until September 2008. The banking-induced crisis caused it to almost double by late November 2008. As the fear subsided, the premium dropped back to pre-crisis levels by January 2010. I have the month-by-month gyrations on my site, also on the front page.

Now, to the present. The ERP started this year at 5.20% and gradually climbed to 5.92% at the start of August. Today (8/8/11) at 11.15 am, in the midst of market carnage, with the S&P 500 at 1166 and the 10-year T. Bond at 2.41%, the implied ERP stood at 6.62%.

So what? If you are a contrarian, you could view this as an opportunity: a return to past norms (4-5% ERP) would translate into a 30-40% jump in the index). If you are a momentum investor, you see the thundering herd and join in, selling short or buying puts. If you are a fence sitter, you are liquidating your stocks and holding cash, waiting for steady state (which may be a long time coming). My last post should provide enough clues as to where I stand but if you are in one of the other camps, I will not try to convert you. Different strokes for different folks!


Anonymous said...

do you still use 10yr US tbond as a risk free rate?

Sanjay Mistry said...

As of 8/8, I used S&P 1130 and T-bond rate of 2.554% and got ERP of whopping 6.81%. I trust this number more than backward looking downgrade of S&P credit rating.

Unknown said...

Definitely contrarian at this point. Corporate earnings look solid across the board (though if 2H 2011 turns out to sluggish, that can change).

Going 'all in' within this week. Why? Because hysteria is probably highest with the double whammy of the European and US crisis. Both of which are probably overblown.

Jonathan Selsick said...

Wondering how you think about the growth rate assumption and how it affects the implied equity risk premium?

For example, this week, it would seem that a lot of the market's decline can be attributed to expectations of slower growth, which are not immdeiately reflected in analysts estimates. Inserting a lower growth rate into the spreadsheet would give us a lower equity risk premium.

Given that analysts estimates are not updated fluidly, how do you differentiate between lower growth expectations and higher equity risk premiums on a daily basis, especially when the markets go into a panic mode like we have seen this week?

Henry Bee said...

jonathan, as long as the model is using consistent numbers then the ERP estimate should be consistently comparable to its own historical numbers. in this case, the forward earnings estimates are used consistently and not revised. if you are worried about growth rate overestimation, then know that 2008 the model also had grossly overestimated earnings forecast being fed into the same model. giving you a good point of reference.

Jonathan Selsick said...

Henry, I agree that using a consistent methodology is essential, however if you are using the ERP to value equities using a DCF analysis, on a daily basis, then it should reflect conditions in the market on that day, including the growth outlook.

If you look at the other link in the good professor's article, you will note that the growth estimate has been unchanged at 6.95% for the past 4 months, and it will likely take a few months for the analyst's estimates to reflect the recent "gloom".

So, if the ERP is calculated on a date on which all analysts estimates were updated, then that would be an accurate number.

But if you look at what happened in the past week, it is very clear that the indices were down based on concerns about growth. If we change the growth estimate from 6.95% to 4%, then the ERP is 5.04% which is actually a decline in the ERP.

Henry Bee said...

good point. data at the daily frequency is more noisy than data at the monthly frequency. and we probably shouldn't compare them to the monthly data. we'll have to see how much the growth estimate gets revised down. the ultimate answer will be during the Q4 earnings season from october to november.

reflections by cobblestone parapet said...

Dear Professor,
Please share your views on European ERP, given the debt contagion?

BD said...

A recent post by my friend Greg pointed out the following :

Equity Risk Premium - Sound Conceptual Foundation?
Can efficient market theory lead to market inefficiency?

Considering the way EMT is taught, a recent study of a key component shows this may be a real possibility.

A Spanish academic has canvassed 150 authoritative corporate finance textbooks published over the last 30 years. He found their recommendation for the equity risk premium, central to CAPM, has a range from 3% to 10%, with 51 different values cited and its five year moving average declining from 8.4% (1990) to 5.7% (2009).

Further, there are four different ways by which the equity premium concept is explained, with historical, expected, implied and required variants cited.

The author posits the required equity premium should always be positive. For his own calculations, he applies a value of approximately 4% as a suitable compensation for risk.

It seems there should be no wonder that equilibrium price discovery is so challenging if generations of new investors are inculcated with very distinct expressions of a foundation concept. Such variance in views makes a horse race, and looks like a market too.

sap testing said...

You are right that the equity risk premium is the key input for estimating the cost of equity, when valuing companies.I am not much aware of these terms but reading articles related to them.