Positive Surprises Leading to Upward Estimate Revisions

We are more than three-quarters done with earnings season and earnings continue to surprise to the upside. As things stand now, it’s about an even money bet that the median S&P firm will post double-digit year-over-year earnings growth. That would make it the 19th straight quarter. While the 9.8% growth is still lower than previous quarters it is much better than expected. The surprise ratio now stands at 3.5:1, down from 3.8:1 a week ago, and 5:1 two weeks ago but still very strong. Positive surprises have been widespread, with every sector showing more positive surprises than disappointments with the exception of Utilities.

Health Care and Materials have been battling it out for first place in the growth derby. However, Health Care appears to be the real standout so far this quarter, posting 15.0% year-over-year growth with 33 positive surprises against only five disappointments. The median surprise is a very strong 5.4%. The Materials sector is also having a very robust quarter, with 14.3% year-over-year growth and positive surprises outnumbering disappointments by more than 3:1. However, its median surprise is somewhat below average at 2.44%. Telecom is the only sector showing negative year-over-year growth (-6.7%) and its surprise ratio is just 4:3, but it does have a large 5.33% median surprise. Funny things happen when you have a small sample size. Utilities just don’t have much juice this quarter either with only 2.5% growth and almost twice as many disappointments as positive surprises. Tech also appears to be the laggard this quarter, with median earnings of only 3.1%. However even here, positive surprises are far exceeding disappointments by more than 3:1, including large positive surprises by many high profile firms.

It is worth noting that the results so far have been much stronger for the S&P 500 (large caps) than the S&P MidCap 400 and the S&P SmallCap 600. Large-caps have been much more active in shrinking their share bases through stock buybacks than have their little brothers, and they also tend to have more foreign exposure and thus benefit from the very weak dollar.

The Zacks “Revisions Ratio”

To help gauge the direction of the market, we take note of what analysts are thinking. By tallying their EPS changes, we can determine our “revisions ratio.” This ratio simply divides the total number of positive estimate revisions by the total number of estimate cuts. Thus, a high ratio is a bullish indicator and a low ratio is bearish. For the S&P 500 as a whole, a number below 0.80 or above 1.25 is generally significant. For individual sectors, the distance from 1.0 should be greater for the numbers to be significant.

The much better than expected first quarter earnings are being reflected in upward estimate revisions for 2007. For fiscal 2007, the revisions ratio stands at 1.59, up significantly from last week’s 1.24. Over the last four weeks, there were 2,006 upward EPS revisions and 1,259 downward, for 3,265 revisions total. This is up 42% from a week ago and is probably close to the peak for this earnings season. Remember that these are moving four week totals, so last week’s reading is not totally independent from this weeks (three weeks of overlap).

Upward revisions are widespread with every sector enjoying a Revisions Ratio above 1.0. The strong showings for individual sectors in terms of both growth and positive surprises are being reflected in positive estimate revisions. Health Care leads the pack with more than three times as many upwards revisions as cuts. Industrials, Consumer Staples and Materials (all sectors which have done very well in the first quarter) all have revisions ratios above one. However, perhaps the most remarkable thing is that no sector has had more cuts than increases. Even Tech and Financial Services which have been the laggards in recent weeks, are seeing net upwards revisions. The ratio of firms with rising estimates to falling mean estimates moved up to 1.36 from 1.30 last week.

The 2008 revisions ratio now stands at 1.64 up from 1.35 last week, and it remains higher than the 2007 revisions ratio. The total number of FY2 revisions over the last month was 2,687, up 39% from last week. As with the 2007 revisions we are probably close to the peak for this earnings season. The big surprise for 2008 is the extremely strong revisions showing for the Utilities sector, with a ratio of 4.0. The sector is a relative laggard for 2007, and has had a weak first quarter earnings season (relative to the rest of the market). As with 2007, Health Care, Staples, Materials and Industrials all saw more than twice as many estimate increases as cuts. Unlike 2007, positive first quarter surprises have no direct impact on 2008 earnings. The strong preponderance of upward revisions therefore indicates better expected fundamentals going forward. Financial Services was the only sector with more cuts than increases, but statistically the ratio was hard to distinguish fro being even. The overall ratio of increasing to decreasing firms was a very strong 1.80, indicating widespread upward revisions.

Full Year Growth On a full-year¹ basis, median earnings growth for the S&P 500 is expected to remain healthy. On a total basis however, growth is expected to slump into the high single-digits for 2007 and then rebound in 2008.

There are a couple of technical points which are affecting the expected growth rates. Recently firms have been pumping enormous, and unprecedented, sums into share repurchases. In 2006, over $432 billion was spent by S&P 500 firms to buy back their own stock. These buybacks amount to 3.4% of the S&P 500’s year-end 2006 value and have the effect of increasing ‘‘06/’’05 growth rates by approximately 3.5%. All indications are that buybacks will have a similar effect on 2007 earnings growth.

The 2006 adoption of options expensing offset this effect somewhat. While the expensing of options provides a level of earnings that is much closer to economic reality than not expensing them, doing so in the numerator (2006) and not in the denominator (2005) depresses the growth rate. For 2007, all firms should be on an apples-to-apples basis.

On a total earnings basis, growth is expected to show a far sharper deceleration in 2007 than on a median earnings basis. All three of the sectors which posted greater than 20% growth in 2006 on a total earnings basis are expected to see growth fall to the single digits. Two sectors, Consumer Staples and Technology, are expected to see noteworthy earnings acceleration in 2007 that is expected to carry over into 2008. Of the two, the Staples sector is probably the better bet since it is seeing its estimates rise at a more rapid pace. The Staples also have much more reasonable P/E’s. In the Materials and Energy sectors, growth measured on a median basis is far healthier than if measured on a total earnings basis, which is in large part due to the concentration of a high percentage of the total earnings for the sector in a few firms in those sectors.

Market Cap versus Total Earnings

When making investment decisions, growth should always be looked at in conjunction with how much you are paying for a stock. Thus, it makes sense to look at the total earnings expected for a sector, relative to that sector’s total market capitalization. This is basically a variation on looking at the P/E.

The chart below shows the share of total earnings for 2006, 2007 and 2008, as well as the share of total market capitalization for each sector (the final bar shown). Since the S&P 500 is a market cap weighted index, this is the same as its index weight. On the chart below, the difference between the sizes of the first three bars shows if a sector is gaining or losing “earnings share”. The difference between the final bar and the first three bars shows if the sector is selling for an above or below market P/E. If the final bar is smaller than the other bars, the sector is selling for a below market P/E. However, as opposed to just showing the sector P/Es, it also shows the relative importance of the sectors to the overall index.

Clearly the Financials are the biggest influence on the market. However, that sector is also relatively cheap on an earnings basis, and thus serves to hold the overall P/E of the index down. In other words, Financials deserve to be the biggest influence on the market, since they contribute approximately 27% of the total expected earnings for 2007.

Energy looks to be under represented in its influence on the market since it is expected to provide 13.6% of the total earnings for 2007, but represents only approximately 10% of the total market capitalization of the index. While its earnings share is currently expected to decline to approximately 12.7% in 2008, even on this basis it is still much cheaper than the overall market.

In more conventional terms, the Energy sector is trading for 11.8x 2007 and 11.3 x 2008 earnings. The S&P is trading for 16.1x and 14.5x, respectively. The second chart below shows the sector P/Es ranked from lowest to highest based on estimated 2007 P/E ratios.

Matt Thurmond contributed significantly to this report.

¹ We follow the convention of referring to the most recently completed full fiscal year as “2006”, and the next full fiscal year to be reported as “2007”.