Equity Risk Premiums (ERP) and Stocks: Bullish or Bearish Indicator

144 comments:

3 comments

1. Im not sure what valuation model youre using to get Value of S&P 500 in your chart. Normally if no explanation is given I would assume that you are just discounting expected cash flows at the expected rate of return, but that would imply (unless the expected cash flows are different for different ERP/RFR combinations) that the value is independent of the composition of that expected rate of return, so apparently that is not what youre doing here. But what are you doing?

2. The crux of your argument here seems to be the positive historical correlation between movements in the risk free rate and movements in the equity risk premium, which — if it continues to be true in the future — would mean that expected mean reversion in todays risk free rate would be associated with a rise in the equity risk premium (and hence falling stock prices — or at least more-slowly-rising-than-usual stock prices). However, given that the equity risk premium is near a historic high, that scenario is kind of implausible. You have to imagine that, when we get back to normal business cycle conditions (which will be associated with a rising risk free rate), the equity risk premium will move to new historic highs. But why would that be? I think rather that the historical correlation is due to factors that dont apply today.

You seem to be using nominal interest rates, rather than real rates, to measure the risk free rate. So what youre really getting is mostly a measure of expected inflation. The correlation between movements in risk free rates and movements in the ERP is really mostly (or maybe entirely or even more than entirely) a correlation between movements in the expected inflation rate and movements in the ERP. This does not surprise me, in that times of high expected inflation are typically times of high expected risk. However, today, I believe the Fed has developed quite a lot of credibility, and any increases in expected inflation associated with recovery are likely to be temporary and likely, if anything, to be associated with reductions, rather than increases, in expected risk.

Anonymous said.

In the table of market sensitivity to risk free rate/ERP combinations, we can get to the market index overvalued by 8.28% when risk free rate=4% and ERP=5.5%, not 5%.

Shiv said.

I suppose it comes down to trying to figure why interest rates are so low.

Suppose interest rates were 3.5%, what would happen. Most likely a deeply inverted yield curve — i.e. long yields would stay low.

The long end of the yield curve yields low, because there is no private demand for leveraged productive capital. QE has lowered long yields a little bit further, but I suspect most of its being low is because of no competing higher yield demand from private parties requiring productive leveraged capital (& in the case of US Ts the safe haven factor pushes yields even lower).

When will interest rates rise? I think once corporates have returned sufficient cash from their hugely cash overweighted balance sheets. And this would need to be combined with rising demand for productive leveraged capital, which will only come when there is growth. And even when there is growth, the initial capital will likely be raised through equity issuance if the markets are substanitally over-valued.

I suspect once interest rates rise, there will be a decline in stock markets. But assuming that the treasury bubble deflates slowly rather than pops which is what I believe most likely to occur, the slow rise in interest rates will be offset by rising growth expectations. The reversion of ERP to 4-4.2% may occur through rising interest rates, offset by rising growth expectations at a slightly lower rate. Put differently, multiple contraction combined with decent growth may lead to no dramatic falls in markets even while ERP & valuation revert.

In some ways low interest rates at the short end of the yield curve are not an option. They are low because if they were high, the credit markets would totally dysfunctional, because at the longer end there would be no demand for debt. QE on the other hand is another story. Its pushing down the long end of the yield curve because of buying from Central Banks. Its flattening the yield curve & adding liquidity which is leading to some overvaluation in all asset classes (but none more than long duration US Ts).

The benefit is creation of public demand for leveraged capital to replace absent private demand for leveraged capital. And that could help growth. To some extent, the lowering of public demand for leveraged capital which is now occuring is not a bad idea. But if the growth now being seen in US is not self sustaining, it would be a drag. Id be much happier in seeing public demand for leveraged capital drop away after growth remains over 2.5% for 1 to 1.5 years while unemployment falls to 6.5% or even a bit lower.

In the period when US public demand for leveraged capital is absent, the Feds QE is adding risk to other markets where there is stronger private demand for leveraged capital.

Id do believe that SP500 is somewhat over valued at present. But 14% to 15% over valued. That kind of over valuation can sustain for long periods as long as growth expectations provide support.

Shiv said.

In light of this evidence, consider again two periods with high ERPs. In 1981, the ERP was 5.73%, but it was on top of a ten-year US treasury bond rate of 13.98%, yielding an expected return for stocks of 19.71%. On May 1, 2013, the ERP is at 5.70% but it rests on a US treasury bond rate of 1.65%, resulting in an expected return on 7.35%.

In 1981 real interest rates were at 3%. But what were real interest rate expectations? In the absence of hard data I assume expectations of a reversion to a very long term rate of 1% real interest rates would not have been unreasonable. That would take the market real return expectation to 6.73%. Today real interest rates are negative 0.3%. I suspect a reversion of long term real interest rates to 1% is likely over time (as it happens a 10 year median inflation is 2.2% while 10 year nominal rates are 3.34% giving a median 10 year real rate of 1.1%). SO really a real return expectation of 6.7% is not unreasonable.

The big difference between now and 1981 is that in 1981 bonds were priced to win (with falling rates), while now bonds are priced to lose (with rising rates).

Shiv said.

Last note, ERP now 5.7%; 1981 5.73%. Median 10 Y Real Interest Rates now 1.1%; Median 10 Y Real Interest Rates in 1981 0.08%. Assuming reversion of real interest rates to 10Y median is an expectation, return expectation now is 6.7% while in 1981 it wouldve been 5.65%. Now that is consistent with a market where debt is over valued now compared to a market where debt was under valued in 1981. Expectation that money would flow from equity to debt then & from debt to equity now.

Andy,

On (1), I do use a discounted cash flow model and the model is in the spreadsheet linked to in the model. The perpetual growth rate is tied to the risk free rate (as it should be in any internally consistent DCF model). Hence, the link between the composition of the expected returns and stock levels.

On (2), you may be right about historical correlations not applying but you cannot be selective about the mean reversions that you think will apply. The same argument can be made about ERP reverting back to the mean.

On (3), you have a good point about real interest rates, but I dont think that that nominal interest rates can be waved away. There is a big difference between a negative real interest rate at a 10% nominal rate and one at at a 2% nominal rate.

On (4), see (1).

Aswath:

The perpetual growth rate is tied to the risk free rate (as it should be in any internally consistent DCF model)

I think this would be true in a long-run equilibrium, but we are clearly far from that equilibrium now. 10-year TIPS yields are negative. Surely one shouldnt expect real cash flow growth to be negative. The market is clearly in disequilibrium, and besides, the equilibrium conditions dont apply when the Fed is deliberately bidding down the term premium.

Regarding (2), I do think that both ERP and the risk-free rate will revert to the mean. However, the former should happen because of the behavior of rational investors, behavior that will, if anything, be encouraged by policymakers, and therefore should happen relatively quickly. The latter is a business cycle phenomenon that is being deliberately manipulated (with good reason, in my opinion) by policymakers. It will happen at the rate that policymakers allow it to happen, so were talking about business cycle frequencies. Its difficult for me to imagine a scenario where both reversions happen at the same rate, unless we get a whole lot of strong economic news in a short time (but if that happens, the expected cash flows from stocks will also likely rise, so you could get rising prices despite mean-reverting interest rates).

Even though interest rates are unusually low, its possible that they are at the same time artificially high. By that I mean, the rates are being determined by the central bank rate floor rather than by fundamentals. If that is correct, then an improvement in conditions will not translate into higher rates.

Exactly what are Cheap GW2 Gold . Curiosity or POI (sights) are shown using a tiny square in the spotlight. These kind of Diablo 3 Gold Cheap offer you experience points when you have frequented all of them after, plus full the charge card reputation. GW2 Field

had a quick look at the spreadsheet- please check cell references. the Goal Seek refers to blank cells. — not 100% sure what the spreadsheet is supposed to calculate.

It is a well executed post. I like the diagram most. It is a helpful informative post. Thanks for sharing this great information. Stock Tips | Free Stock Tips | Stock Market Tips | share tips | commodity Tips

stock tips said.

Thank you for such a fantastic blog. Where else could anyone get that kind of info written in such a perfect way? I have a presentation that I am presently working on, and I have been on the look out for such information.

Here are many ways to reduce your small-business tax burden.wages and make sure that the work benefits your company to qualify for these exemptions

A number of other corporation tax reliefs can help reduce your corporation tax liability. thetaxexperts relief is available on qualifying research and development.

Anonymous said.

Hi prof

Thanks for this post — particularly relevant in view of whatevers happening in the bond markets. Getting a bit confused, with bond yields now edging higher on the back of improved growth, would u expect equity risk premium to increase (alongside risk-free), or would you hold market return steady and reduce ERP?

If the latter applies, is it a right inference that high-beta stocks would benefit in such an environment. Appreciate your insights

Anonymous said.

Hi, professor. Im a fan of your blog and online classes.

Dear Mr. Damodaran, Thanks for this. I wish to clarify if the DV01 for Rf is so much lesser than DV01 for ERP (assuming the Rf on shorter end of the curve will remain low). Essentially assuming different delta Rf across the curve.