Earnings expectations have remained resilient across U.S., Europe, and Canada as we enter the second half of the year with most global benchmarks down 15-20%.
For most markets, year-to-date performance has been entirely driven by a decline in the P/E multiple as opposed to any deterioration in earnings growth expectations. For example, the S&P 500 has declined 19.7% YTD in comparison to an increase in 2022 earnings growth expectations from analysts (8.4% on Jan 1st to 9.5% on July 1st). This has led the forward P/E multiple to decline from 22.1x at the beginning of the year to a current reading of 16.1x as shown in Exhibit 1.
As a result, there appears to be a significant disconnect between market pricing and analyst estimate revisions. Readers can view our earnings report to track weekly changes in growth expectations.
Exhibit 1: S&P 500 P vs. E
Given the considerable macro uncertainty globally, we look at earnings dispersion to determine earnings stability and if there is strong agreement (or disagreement) between analyst forecasts for the upcoming earnings season quarter.
To calculate earnings dispersion, we use the following formula: = (highest earnings estimate – lowest earnings estimate) / consensus mean estimate. For example, using 22Q2 data for Apple Inc (AAPL.O), we have an earnings dispersion measure of 22% (= ($1.32 – 1.07) / 1.16).
While we don’t have a historical benchmark to reference earnings dispersion, we assume that a high level of earnings dispersion (i.e. greater than 50%) means there is significant disagreement amongst analysts.
Conversely, a low level of earnings dispersion (i.e. less than 20%) means there is significant agreement amongst analysts. This can be of interest in the current environment – for example, does this mean that analysts have a high degree of confidence that companies will deliver during earnings season? Or is this a sign of analysts being hesitant in making material revisions to their current forecast?
We define earnings as ‘Preferred Earnings’ which is typically Earnings Per Share (EPS) but can vary by region/country. In the U.S., preferred earnings are typically EPS except for Real Estate where it can be either EPS or Funds from Operations per share (FFOPS). In Europe, EPS is most common and in Canada we see a wider range of preferred measures including EPS, FFOPS, Cash Flow Per Share (CFPS), and EBITDA.
Using the Refinitiv Workspace excel add-in, we aggregate earnings dispersion for the S&P 500 in Exhibit 2 by counting the number of companies that fall into specific buckets and repeat this process for the STOXX 600 and S&P/TSX in following exhibits.
Exhibit 2: S&P 500 22Q2 Earnings Dispersion
Source: Refinitiv Workspace (data as of July 4th 2022).
At a high level, earnings dispersion is classified as ‘low’ with over half of the index having a value less than 20%. In comparison, 13.7% of constituents have ‘high’ dispersion (above 50%) while 27.3% of the index has average dispersion (20-50%).
At a sector level, Financials has most of its earnings dispersion in the low to average bucket (10-30%) with only 9 constituents being in the 0-10% category.
Consumer Discretionary appears to have high dispersion across the board and has the largest count in the > 100% bucket. This would coincide with recession worries and the health of the consumer which will be harder to predict and can lead to greater earnings surprises in the quarter.
Surprisingly, Information Technology has most of its earnings dispersion (47 constituents) in the lowest category (0-10%). This may suggest that the market sell-off YTD in technology is driven primarily by rising interest rates rather than worries about earnings expectations.
We display the same data as above in Exhibit 3 for the STOXX 600 (Europe) and S&P/TSX (Canada) to provide a global perspective.
Exhibit 3: STOXX 600 and S&P/TSX 22Q2 Earnings Dispersion
Source: Refinitiv Workspace (data as of July 4th 2022).
Source: Refinitiv Workspace (data as of July 4th 2022).
Like the S&P 500, earnings dispersion is classified as ‘low’ for both the STOXX 600 and S&P/TSX with 48.7% and 42.3% of constituents respectively having a value less than 20%.
However, unlike the S&P 500, there appears to be higher dispersion in both indices on a relative basis as there less constituents in the S&P/TSX and not every constituent in the STOXX 600 has quarterly estimate data.
More specifically, 16.7% of STOXX 600 constituents having high earnings dispersion and for the S&P/TSX this rises to 26.5%.
Financials has higher earnings dispersion in the STOXX 600 compared to the S&P 500 and S&P/TSX. Industrials has high earnings dispersion across the board in the STOXX 600.
There is also a high level of earnings dispersion in the Materials sector for both the STOXX 600 and S&P/TSX which is likely impacted by demand for materials such as copper in the face of recession fears.
To conclude, we again reiterate that a low level of earnings dispersion (which is what we see across all three indices) indicates that either a) analysts have a high degree of confidence in management, or b) analysts are hesitant in making material revisions to their current forecast. We note this of importance given the highly uncertain macro-outlook and may lead to unexpected surprises during earnings season.
Refinitiv Workspace is a complete solution for research and analytics. It places the most comprehensive market information, news, analytics and trading tools available into a desktop.
