To Beat or Not to Beat: The Shifting Fundamentals of Profiting from Earnings Estimates
No concept stands more fundamental to investing in stocks than earnings, which are, ultimately, the reason we own stocks in the first place. A company’s ability to earn (now and in the future) directly determines its ability to reward its shareholders, whether through dividends or stock price appreciation.
Today, a huge volume of stock trading centers on attempting to profit from earnings estimates. In general, like any other financial asset, we formally expect a stock price to reflect the present value of its future cash streams, discounted for the projected interest rate environment. But this “valuation” is full of uncertainty to the point that it’s almost a total abstraction!
Knowing the “true” value of a stock would mean knowing decades of future interest rates policy, future corporate returns, and even future cost structures (remember, earnings is typically calculated as after-tax profits, not gross revenues). But the uncertainty surrounding a given stock’s valuation isn’t just confined to the distant future: investors are forced to value a company according to the best publicly available information regarding the company’s performance. Practically, this means investors are left to value a firm based on their most recently available regulatory filings.
Indeed, recent earnings statements will likely form an anchor for a given stock’s overall valuation. But, as always, markets are perpetually trying to stay one step ahead: and that’s where earnings estimates come in to play.
To calculate estimates for a firm’s future earnings, financial analysts will draw upon factors including:
- Macro Forecasting Models: these statistical computations may take into account factors including broader industry performance, macroeconomic indicators, prices of relevant input commodities, and more.
- Management Guidance: a company often gives its own projections for future earnings in filings like the 10-Q. While they must be taken with a grain of salt, these projections are still valuable evidence for forecasters (who can discount a firm’s projections if they’ve proven unreliable in the past).
- Micro “Fundamental” Modeling: analysts are also likely to draw upon the firm’s own detailed financial data (eg. Looking at how assets are shifting on the balance sheet) to enhance their own projections.
- Novel Methods: we’re also seeing interesting work on the crowd-sourcing of estimates by platforms like Estimize.
It’s important to understand that some of the bests mathematical minds in the world work on intensively on these sorts of modeling efforts—yet they’re still full of uncertainty. Finance simply isn’t an exact science, which is precisely why unfolding new events can so dramatically alter the valuation of a stock. This means that virtually every financial analyst will have their own projection for a given stock’s future performance.
The projections of recognized financial analysts who publish their work are averaged together into the “consensus projections.” In the press, these are the numbers that are directly compared to actual results to generate a Beat or Miss for a stock when it announces earnings news.
Investors should understand that there’s no set list of analysts providing projections for every stock. Large-cap stocks will likely be covered by far more analysts, while micro-caps may be covered by a handful if any.
Research has shown a strong correlation between whether a stock “Beats” or “Misses” and its subsequent performance—leading to earnings-announcement trading becoming one of the most popular trading strategies in the world. Like any other hot stock trading strategy, however, the stampede of investors toward these apparently easy profits had rendered them far more difficult to find.
As firms become savvy to the vast sea of investors trading based on the spread between forecasted and actual earnings, they’re behaving accordingly, carefully curating a stream of “Beats” by understating their projections and then consistently exceeding them. But the stock market is always a strategic give and take: if companies are consistently outperforming estimates, analysts will begin raising their own estimates even more.
These dynamics make successfully executing projection-based strategies more challenging. Some commentators have gone so far as to proclaim the “death of the earnings beat.” We think the demise of the earnings beat (and miss, for the short play) is greatly exaggerated. It’s simply a more competitive playing field than before.
And the key to soaring past the competition to find consistent earnings-based profits lies in data. It’s true that we’re seeing more companies post earnings beats, briefly surge, and then get attacked in a sharp selloff. The evidence suggests a growing pool of money targeting overreactions to earnings announcements. In this context, investors who are trying to trade an earnings announcement without good historical data are likely to lose their shirt to more data-savvy players.
Our platform features historical data on markets’ reactions to earnings announcements for a huge of array of equities. Vitally, this information is calculated at various hold times for each individual stock. Some stocks, for instance, tend to peak after just one trading day and fall back by Day 5: a great target for a 1-day hold time play on the stock. Or, if you see that a stock gains 5% on average from Beats but often surges 7-9% before falling back, you can buy the stock with an order to cash in at the 5% gain mark (or even short the stock if it shoots too high). Our data can be viewed for both individual stocks, and aggregated according to various parameters.
A glimpse from inside the News Quantified interface.
Every stock is unique, but, as ever, the right data is the key to getting out ahead. If you’d like to learn more about using real-time news and financial data synthesized with historical analysis to find market-beating profits, we highly recommend one of our totally free virtual training seminars. You can sign up for a spot in our next session using the button below: