“The time to buy is when there's blood in the streets."
Conclusion: Equities, commodities, and high yield bonds have seen a dramatic sell off in the last couple of weeks. While they are cheaper, recent history suggests that we are not yet at washed out / “blood in the street” valuation levels.
In the last couple of weeks, we’ve had a massive sell off in global risk assets. One of the better proxies for the sell-off in risk assets in our purview is the SP500, which closed on July 26that 1,332 and has sold off in a straight line to 1,119 (as of yesterday’s close), a decline of -16.0%.
As Keith noted this morning in the Early Look, there is no doubt the market is flashing oversold in our quantitative models. In fact, according to our most recent lines, the SP500 is three deviations oversold at 1,086 on an immediate-term TRADE basis. Therefore, the purpose of this note is not to suggest there is substantial downside from here, at least in the short term, but rather to actually frame up some key valuation metrics on various asset classes that might be a better indicator of truly “cheap”.
Stepping back, one of the key domestic catalysts for the sell-off in equities was both the release of the estimate of Q2 GDP at +1.3% and the downward revision of Q1 GDP to +0.4%, which, assuming the GDP estimate for Q2 is accurate, takes GDP growth for the first half of 2011 to sub 1% in the United States. The key value of GDP growth relating to the broader equity markets is that economic growth ultimately drives earnings growth and, therefore, valuation.
We wanted to reiterate a point we made in a note on Thursday specific to the correlation of economic growth and corporate earnings, which is as follows:
“In fact, slowing or declining GDP growth can lead to dramatically decelerating earnings. In the last decade we have seen this in spades as noted by the 21.5% decline in SP500 TTM earnings from March ‘01 through June ‘02 and 44.0% decline in S&P 500 earnings from September ‘07 to September ‘09. Going back the last thirty years, there have been five periods in which earnings for the S&P 500 broadly have declined.
On average, the decline has been a peak-to-trough decline of -25%. In a scenario analysis where we assume we are entering a period in which earnings are in decline and they decline by the average of the five declining periods over the last thirty years, the implied earnings of the S&P 500 over the next twelve months is ~$74.79. Based on the current price of the S&P 500, this is a ~16.2x earnings multiple. Not exactly cheap.”
Interestingly, since we wrote the note above, the SP500 is now even cheaper trading at closer to a 14.9x multiple on the earnings projected in the scenario above and based on yesterday’s close. As well, those equity investors that continue to support being long of the equity markets on TTM earnings or NTM earnings estimates, can also argue a more compelling valuation case as the multiples of the SP500 are 12.3x and 11.2x, respectively on these metrics. We’ve summarized these valuations in the table below:
In the table, we’ve also incorporated CAPE earnings. Recall, CAPE earnings, or Cyclically Adjusted Price-to-Earnings, are utilized by Professor Shiller of the Yale Economics department to determine the fair value of the SP500. In terms of the numerator, or price, Shiller uses the monthly average of daily closes for the SP500. To derive the earnings data, the denominator, Professor Shiller uses the quarterly earnings data from the SP500’s website and utilizes an interpolation to provide earnings data by month. He then adjusts both the numerator and denominator for inflation using CPI from the Bureau of Labor Statistics. Finally, the inflation-adjusted price is divided by an average of ten years of real monthly earnings to determine the CAPE.
When we looked at this valuation methodology in late March, the market was at a 23.6x earnings on this basis. Since then, valuation has certainly become substantially more compelling with a decline in multiple of roughly 18.2%. That said, the CAPE P/E is still above its long run average of 16.4x. The equity market has become substantially cheaper, but is still marginally pricey on a cyclically adjusted basis.
In the chart below, we’ve looked at equities based on a dividend yield basis, which is typically another metric that is given to validate the valuation call for equities. In the chart, we show the Dow Jones Industrial Index, a better proxy for high dividend companies than the SP500, going back three years. While we could have gone back much further, we purposely wanted to look at the dividend yield based on the most recent cycle. As the chart shows, the dividend yield for the Dow is currently 2.8%. In this cycle, which for purposes of this analysis we will consider the last three years, the Dow has reached a 4.0% yield. On this metric, it would seem there is potentially downside based on yield before dividend yield stocks become “cheap”.
Another asset class that has been very relevant in this most recent global sell off is gold, which has been a haven for safety and protection from runaway Keynesian doctrine. Gold was up yesterday and is up more than +20% in the year-to-date. So now, of course, the question is whether gold is expensive. With gold this is obviously a difficult question to answer since there is no truly conventional valuation metric from which to evaluate its value. Specifically, gold has no earnings power.
In the chart below, we looked at the value of gold compared to the value of oil. We imputed and charted the number of barrels of oil that an ounce of gold would buy. Interestingly, on this metric, gold doesn’t look overly expensive based on its levels over the last three years. Currently, one ounce of gold will buy roughly 22 barrels of West Texas Intermediate oil. This metric peaked at closer to 28 barrels in early 2008. Even more interesting is that if we look at gold versus Brent oil, the metric is closer to 17x. Regardless, the valuation suggests there is reasonable more upside to gold, more downside to oil, or both.
The market that looks truly expensive is the recently downgraded U.S. Treasury market. As outlined in the chart below, the 10-year U.S. Treasury note is currently yielding ~2.39%. In the most recent three year cycle, the 10-year bottomed at yield of ~2.05% on December 30th, 2008. So, we are close to that price, which is the fifty year low for 10-year Treasuries. Given, it is likely fair to say that Treasuries are expensive, but, obviously, yields have the potential to go even lower especially subject to another round of incremental easing.
As it relates to fixed income, the yield on high yield credit has ramped dramatically over the last two weeks, as high yield bonds have sold off in line with equities. In fact, on August 1st 2011 the yield on high yield credit (based on the Bloomberg High Yield Index) was 7.28% and has since ramped to ~8.20%, for an increase of +11.2%. In the same period, as highlighted in the chart above of 10-year Treasuries, the spread between government bonds and high yield has widened, which we have outlined in the table below. Interestingly, this spread remains dramatically off the spreads we saw in late 2008 and early 2009.
This is not to suggest that high yield spreads will reach the parabolic widening that they did in late 2008 and early 2009, but rather just to suggest that while high yield has become less expensive it is not extremely cheap on either a relative or absolute basis.
Tomorrow we will be doing a call to discuss our more discrete thoughts on the various key asset classes at 11am. We will be sending out presentation materials and dial in info tomorrow morning.
Daryl G. Jones
Director of Research