More Ways To Maximize Returns Using Estimize Data, As Highlighted In The Latest Wolfe Research Paper

Updated on

Today we’re excited to share the findings of a recently published research paper from independent quantitative research firm Wolfe Research. Entitled “More Accurate and Timely Estimates Lead to Better Investment Strategies,” the piece not only backs up what we always knew about Estimize data, but explores ways in which short and long term investors can leverage Estimize data to produce alpha and manage risk around earnings.

Let’s first recap some of the findings that previous papers have reported, and that Wolfe’s research reaffirms using the latest cut of Estimize data:

  • Estimize earnings estimates are “significantly” more accurate than the sell-side (IBES and CIQ combined), and as the number of analysts increases, the accuracy improves.
  • For large cap names with over 30 analysts, Estimize is almost twice as accurate as the sell-side.
  • As an earnings release date approaches, the Estimate data becomes more accurate as estimates are timelier due to less regulation impediments for crowdsourced estimates.
  • The lack of regulation also allows for fresher estimates that incorporate new information as it is available, therefore leading to more accuracy.
  • Professionals within the Estimize data set are no better than non-professionals when making estimates. In fact, non-professionals tend to be more accurate, although the best consensus comes from a combination of the two.

Alpha-generating Investment Strategies Using Estimize Data

For the second half of their paper, Wolfe shares new ways to use Estimize data in the investment process.

PPre-earnings Announcement Strategy

The white paper found that EPS revisions made in close proximity to the announcement date lead announcement day return. They believe this is because estimates revised ahead of a report are made with higher conviction. In their findings, an upward EPS revision in the week ahead of an earnings announcement lead to a more positive announcement day return, just as a negative revision leads to a less positive announcement day return. The larger the revisions, the larger the impact on the return. When both Wall Street and Estimize revise in the same direction ahead of a report, the announcement day excess return becomes even more prominent.

We started to follow these trends this earnings season, and found, indeed, had you followed the revisions activity for the likes of Home Depot, Snap Inc., Disney and others, you could have made additional gains.

As previously mentioned, and highlighted in Figure 23, the ease of revising estimates in Estimize leads to the estimize database having much larger revisions than the sell side.


Wolfe found a greater consistency in excess earnings announcement day return with Estimize data, and a significantly larger alpha, but combining the two data sets still produces the best result. You’ll see in the chart below for revisions of >5%, the alpha on the short side is substantially higher than the long side. “The short-side alpha based on the sell-side is minimal, which reflects sell-side analysts’ inability to identify short opportunities ahead of earnings releases.”


By buying those stocks with the highest positive revisions by both Estimize and the sell-side consensus right before the earnings announcement day, Wolfe found the average alpha increases monotonically as the conviction threshold was raised and that the excess return is skewed on the short side, especially at the extreme levels.


Post Earnings Announcement Drift (PEAD)

The Post Earnings Announcement Drift (PEAD) is another strategy that benefits from the use of Estimize data. Research has show that investors tend to underreact to earnings news on the day a company reports, the stock price then moves in the same direction as the announcement day return, however, PEAD disappears in a matter of days due to arbitrage.

Figure 26 shows the first day PEAD, the maximum possible return that investors can expect, after a company reports earnings. Estimize data shows larger alpha and a more monotonic pattern.


Low Risk Strategies

It’s long been believed that investors are rewarded for a willingness to hold risky stocks. Flipping this notion on its head, Wolfe presents a low risk strategy which overtime delivers higher returns than risky securities. They define risk in a number of ways: beta, realized volatility, options-implied volatility, idiosyncratic volatility, informed trading and earnings dispersion.

As shown in figure 28, lower dispersion stocks tend to garner higher returns on the announcement date.

To highlight the benefit of adding a low risk overlay to a traditional long-only portfolio, Wolfe compares three strategies:

  1. Benchmark – a market cap weighted long-only portfolio of all stocks in the Estimize universe.
  2. Low risk portfolio avoiding earnings risk – This portfolio excludes companies that have a pending earnings announcement one day before the announcement date, from the benchmark.
  3. Low risk portfolio avoiding earnings uncertainty – This portfolio excludes stocks in the top half dispersion by either the Estimize or the sell-side.

As shown in Figure 29, both low risk portfolios beat the benchmark, while removing high earnings uncertainty produces an even higher return. In addition, as shown in Figure 30, steering clear of high dispersion names increases Sharpe ratio and lowers portfolio volatility.


Long Term Investment Strategies

Given its overall superiority when it comes to accuracy and timeliness, Wolfe applies Estimize data to traditional earnings yield, earnings growth and Return on Equity (ROE) strategies.

Enhanced Earnings Yield

This strategy measures the contribution of Estimize estimates by comparing three earnings yield factors:

  1. Benchmark earnings yield – consensus sell-side EPS (average of CIQ and IBES)
  2. Standard Estimize earnings yield – Here the sell side consensus is replaced with the Estimize consensus (at least 3 estimates to qualify)
  3. Estimize FES (Freshness, Experience, Skill) earnings yield – Here the FES model which reflects the freshness of each estimate, analyst experience, and analyst skill, is used.

Figure 31 shows how using Estimize data improves the traditional earnings yield performance “Accounting for the freshness of each estimate, analyst experience and skill further boosts returns.”


When rebalanced daily, and even after the transaction cost, Estimize still outperforms the benchmark sell-side consensus.


Even with Estimize’s large cap tilt, the performance of the Estimize FES earnings yield factor also adds alpha in the Russell 3000.


To highlight the benefit to long-only strategies, Wolfe looks at the performance of the top value stocks (i.e., the top quintile stocks with the highest Estimize FES earnings yield) with an equally weighted Russell 3000 index. Figure 34 shows this strategy delivers consistent alpha and increases the Sharpe ratio by almost 45%.


Low Risk Value Strategy

Often value stocks tend to be cyclical names that have been underperforming, and the reason they deliver high returns is to compensate investors for the potential risk.

As discussed above, this paper shows that avoiding earnings risk and uncertainty can lead to higher returns. As such, the combination of Wolfe’s enhanced value factor with a low risk overlay delivers a Share of 1.25x with lower volatility than traditional value signals, as shown in Figure 36.


Once again Estimize data is proven to be superior to the existing sell-side consensus, and when used in certain investment strategies shows the ability to generate higher returns. For more details on the research detailed in this article, read the entirety of Wolfe’s paper here.

If you are interested in discovering more alpha using the Estimize data set, please contact us today!

Article by Estimize

Leave a Comment