Gavekal Capital: Digging Deeper into Deteriorating Sales and EPS Estimates

Wednesday, January 28, 2015

Digging Deeper into Deteriorating Sales and EPS Estimates

Analyst estimates of future corporate performance can be some of the most confusing financial data items out there in no small part due to the multitude of different ways of measuring them. Does one look at estimates in local currency or translated to USD, should one look at fiscal year, calendar year, or forward twelve months estimates, how about rolling vs non-rolling, should EPS be adjusted for "one-time" items or not, etc, etc, etc.

As US-based global stock investors we assume currency risk and so it makes sense to take into account currency swings when analyzing estimates data. For this reason we translate all of our estimate data back into USD. After all, if EPS for a Japanese holding increases by 10%, but the yen falls by 15%, EPS translated back into USD is lower than it was previously and this is important. Additionally, due to the lack of standardized data for international companies, we measure estimates based on a rolling fiscal year basis and we do not make any "one-time" adjustments to bottom line earnings.

So how have analyst estimates been acting recently after we make our translation to USD? As we will show exhaustively below, they are telling us that a lot of developed market companies have caught a cold.

In the first table we show the current fiscal year sales and EPS estimated growth rate for all developed market industry groups. We point out that sales growth in aggregate is expected to be negative for the current fiscal year. EPS are expected to grow by a slow 4.5%. Surprising as it may seem, energy is still expected to post positive fiscal year sales and EPS growth. The drag is coming from banks, consumer staples, telecoms, materials and things that weigh a lot like capital goods, autos and transport equipment.

Estimated growth for fiscal year 1 sales and EPS - MSIC World Index Constituents:
image

In the next table instead of showing the level of estimates we are showing the change in the level over the last three months. This helps us get a sense of how estimates are evolving (i.e. are they going up or down or sideways). We observe that sales estimates have over the last three moths deteriorated for every industry except semiconductors and EPS estimates have deteriorated for all industries. Remember, we are not looking at levels here, but at the change in the level. In aggregate we show that over the last three months the expected sales growth has declined by 4.6% and EPS growth has declined by 6.5%. These are large moves in a short period of time!

3-Month Change in Estimated growth for fiscal year 1 sales and EPS - MSIC World Index Constituents:
image

In the next table we show a measure of estimate revisions breadth. Specifically, we are trying to get a sense for just how widespread this decline in estimates has been. Is the decline due to a few companies having their estimates knocked off a lot, or are analysts taking down their estimates across the board? To do this we calculate the percent of companies that have seen analysts increase estimates vs three months ago. Analysts have increased their sales estimates for 48% of semiconductor companies, but only 2% of auto companies. Surprisingly, 20% of energy companies have had their estimates improve versus three months ago. There are fully thirteen industries that have had a lower percentage of companies with improving estimates compared to energy. Only 26% of companies have experienced positive EPS revisions compared to three months ago.

Percent of companies whose estimates for fiscal year 1 sales and EPS have increased vs three months ago- MSIC World Index Constituents:
image

So to conclude, we've shown that the level of estimates is low (negative for the top line), estimated growth has been revised down significantly over the last three months, and the decline in aggregate estimates has been caused by widespread rerating of expectations.