Goldman Defends Its "Stocks Are Overvalued" Call From Angry Clients

Tyler Durden's picture

Ten days ago, the few carbon-based habitual gamblers left in the market stopped and read Goldman's report which, as we said, may have 'just killed the music' with its slam of the market saying the "S&P500 is now overvalued by almost any measure." Recall: "The current valuation of the S&P 500 is lofty by almost any measure, both for the aggregate market as well as the median stock: (1) The P/E ratio; (2) the current P/E expansion cycle; (3) EV/Sales; (4) EV/EBITDA; (5) Free Cash Flow yield; (6) Price/Book as well as the ROE and P/B relationship; and compared with the levels of (6) inflation; (7) nominal 10-year Treasury yields; and (8) real interest rates. Furthermore, the cyclically-adjusted P/E ratio suggests the S&P 500 is currently 30% overvalued in terms of (9) Operating EPS and (10) about 45% overvalued using As Reported earnings." Since then, many of Goldman's client must have been displeased that David Kostin refuses to drink from the punchbowl anymore, and sent in their complaints. However, Goldman has refused to budge and issued a follow up defense to its thesis that stocks are overvalued more than at any other time except the tech bubble with "Valuation fact vs. fiction part 2: Responding to common questions about S&P 500 valuation."

From Goldman:

We received a barrage of questions following the publication of last week's Kickstart commentary about the valuation of the US equity market. Last week we argued that the S&P 500 currently trades towards the higher end of a fair value range based on a variety of metrics, and noted in particular that the P/E multiple has rarely been higher than it is now, outside of the Tech bubble.

With the market close to fair value, we believe the forward path of US equities will depend on the trajectory of profits rather than further expansion of the valuation. We forecast S&P 500 EPS will grow at a 7% annualized pace through 2016, driving the S&P 500 to 2100 by the end of 2015 and 2200 by the end of 2016: a gain of 20% over the next three years.

Most client responses attempted to justify personal expectations for continued multiple expansion in 2014. This supports our observation that many on the buy-side expect price gains of 10% to 20% this year, well above the 3% upside to our target of 1900 for year-end 2014. Below we continue the conversation and respond to the most common questions.

1. Many investors asked how our conclusions would change if we used a longer historical valuation series than forward P/E, which starts in 1976. Exhibits 1 and 2 show a historical series of trailing P/E multiples since 1921 and the distribution of these multiples.

The 90-year timeseries of trailing P/E multiples shows a similar picture to the 40-year forward P/E multiple timeseries. At 18x, S&P 500 still trades above average valuation, ranking in the 75th percentile historically. While this is modestly lower than the 83rd percentile ranking of 16x forward P/E, the fact remains that market has rarely traded at a higher P/E outside of the Tech bubble, or coming out of recessions when EPS were extremely low.

2. The low interest rate backdrop was the most common client justification for continued P/E expansion. Exhibit 3 shows average P/E multiples in different nominal and real interest rate environments. Current S&P 500 P/E is either at or above the historical example regardless of the metric used. In addition, the Fed Model suggests S&P 500 upside roughly in line with our 1900 target given the current gap between equity and bond yields. For more on the Fed model, see our US Kickstart from Dec. 6, 2013.

Some clients pointed out that stronger US GDP growth and Fed taper should lead to higher yields, which have implied higher P/E historically. While this view is in line with our 2014 economic and interest rate outlooks, S&P 500 P/E already exceeds the historical average when nominal yields fall between 3-4%. While yields between 4%-6% would imply a modest 1 P/E point increase, we don’t expect rates will reach that level in 2014. More importantly, averages in that bucket are largely biased by the Tech bubble.

It is important to note that long-term histories of the market have inherent limitations. First, because 5% of S&P 500 constituents turn over each year on average, fewer than 20% of current S&P 500 constituents were members of the index 40 years ago. Index composition was also different at a sector level: Energy was 25% of the market in 1980 and traded at a low multiple of 6x, while in 1990 Consumer Staples was ascendant and accounted for 15% of the market at a premium multiple (at the time) of 13x.

3. We also received questions about specific sectors or stocks that may be biasing aggregate S&P 500 valuation. As we have highlighted recently, however, the distribution of S&P 500 P/E multiples is currently historically tight. Even the most expensive sector in the market, Consumer Staples, only trades at a single P/E point premium to the broad market.

4. The most cautious question from clients related to the relationship between rising valuation and margins, which are near record levels. We have not found any reliable historical relationship between S&P 500 margins and P/E multiples. Margins have hovered near their peak of 8.9% for over three years, where we expect them to stay in 2014. However, as the labor market recovers and inflation eventually rises, risks are to margin downside. That said, consensus expects margins will expand to a new high of 9.5%. Flat P/E and margins means sales growth will be the key driver of returns.