What are Investors Looking for this Earnings Season?

Earnings season is normally the time when investors use freshly reported results to update their expectations. It is a time to evaluate whether a company is on track to achieve its targets, exceed them, or perhaps fall short. Those results, combined with management’s forward-looking comments, form the basis for updated expectations. However, this earnings season will not fit into that framework. A recent article from Bloomberg highlighted the lack of information currently facing investors.

Operational and financial results from the current earnings season stand to be less reliable, thus not an accurate picture of a company’s normal earnings potential. Additionally, most companies have withdrawn previous guidance and replaced it with some version of “it is too uncertain to provide an accurate forecast at this time.” These two developments change the dynamics of what is relevant this earnings season.

Prior to earnings season, sell-side analysts normally adjust earnings forecasts and tweak financial models based on new information or updated expectations ahead of company announcements. In normal times, there would be a consistent split between positive and negative revisions. The current situation has not created a dearth of revisions but has considerably altered the mix. Unsurprisingly, revision activity over the last few weeks was predominantly negative. Several analysts even commented their revisions with the caveat that it was an initial best guess and further revisions were probable.

For perspective, we compared current revision activity to previous years. Below, we compare year-to-date 2020 estimate volume to the average from 2005 through 2019. The analysis counts the daily EPS estimates for each company’s “next” quarter. The numbers on the x-axis represent business days from January 1st. As of April 21st (No. 75 on the chart), the totals are relatively similar. Charts 1.a and 1.b show the total number of revisions is not that different from the average of the past 15 years. What is different about 2020 is there has been a higher running total instead of the usual lull in between earnings seasons.

Chart 1.a

Chart 1.b

Source: S&P Global Clarifi; 361 Capital.
X-axis represents number of business days from January 1st. For companies on standard March/June/September/December quarter reporting, days 14 – 45 comprise the bulk of December quarter-end earnings announcements. March 31st is day 62 in 2020. Total count of estimates on each individual day for each company’s next quarter.

Estimate revisions have not disappeared, they are just mostly negative. The next two charts plot the positive revision percentage. This shows the positive revision rate is below previous years. For both the S&P 500 Index and Russell 2000 Index, roughly 20% of revisions have been positive this year, compared to the 15-year average of 40%.

Chart 2.a

Chart 2.b

Source: S&P Global Clarifi; 361 Capital.
X-axis represents number of business days from January 1st. For companies on standard March/June/September/December quarter reporting, days 14 – 45 comprise the bulk of December quarter-end earnings announcements. March 31st is day 62 in 2020. Total count of estimates on each individual day for each company’s next quarter.

How do we reconcile the perceived lack of estimate revisions with the fact that they have been trending in line or above average? Let’s look at a few heavy volume days from charts 1.a and 1.b as an example. There are spikes on days 45, 51, and 63, particularly in the Russell 2000 Index, during what is normally a quieter period for estimate changes. Day 51 represents March 16th. It was the Monday after the second emergency rate cut by the Federal Reserve. The S&P 500 Index fell 11.98% and the Russell 2000 Index lost 14.27%. As expected, a sudden 100-basis point decline in the target borrowing rate had an impact on estimate revisions – especially rate-sensitive businesses and those more levered to overall economic activity. Most revisions that day occurred within the Banks and Energy sectors. Those two groups accounted for 40% of the S&P 500’s negative revisions and 56% of the Russell 2000’s negative revisions. If we include Industrials and diversified Financials (ex banks), the percentage of negative revisions attributable to those select companies grows to 60% for the S&P 500 Index and 67% for the Russell 2000 Index. Banks, energy, diversified finance (ex banks), and Industrials account for about one-third of the Russell 2000 constituents and about 28% of the S&P 500. Negative data that affects these economically-sensitive sectors has a disproportionate impact on the total number of revisions.

The point of this exercise is to show that revisions have not completely disappeared. The total number of estimates has not declined because of the large, but concentrated, volume in certain industries. This is more pronounced in the Russell 2000 Index. We use “361 Sectors,” which is a combination of GICS Sectors and Industry Groups.

Table 1

Source: S&P Global Clarifi; 361 Capital.

Another way to analyze this is to look at how “old” estimates are this year compared to previous years. The next two charts plot the “age” of the average negative and positive estimate in the S&P 500 Index (3.a) and Russell 2000 Index (3.b). Age is determined by how many business days have passed since a company last received an up or down estimate revision. Therefore, there are two ages – one for upward and one for downward revisions. Again, we use business days since January 1st and next quarter estimates. Higher up the y-axis indicates an older estimate. The blue line on the right plot of each graph reveals the average age for upward revisions at each day this year has been consistent with prior years. On the left, it shows the average age for negative revisions in 2020 is much “younger” than at similar points over the most recent 35 trading days. For comparison to the Financial Crisis, 2008 and 2009 highlighted in red and orange.

Chart 3.a

Chart 3.b

Source: S&P Global/Clarifi, 361 Capital.

Total revision activity has trended above average during the normally quieter period between the first and second quarter earnings seasons this year. Years 2008 and 2009 experienced quicker downward revisions and slower upward revisions relative to the comparison group, but still followed the same periodic pattern. Current year negative estimates broke trend, occurring quicker than in the past 15 years. What would cause the belief that there is less revision activity? My hypothesis is because revisions have been based primarily on macro data, already known to be negative, and concentrated in certain sectors. Company-specific revisions based on idiosyncratic fundamental or relative performance within peer groups appear to be lacking. In its place, financial health and liquidity concerns have supplanted earnings analysis. That is not to say those measures are otherwise unimportant; those quality metrics have taken on a greater significance.

No matter the estimate revision activity, security analysis will be more challenging in the near term. Standard methods of earnings analysis and judging the accuracy or probability of meeting guidance do not apply in this environment. A possible silver lining is that it will force investors to analyze companies without anchoring to the most recent results and management guidance. Investors need to evaluate how companies will manage through this period and adapt to whatever economic reality awaits on the other side.

Read our recent blog post, Understanding Long/Short Fund Fees >