I'm excited to announce that the FactSet Estimates – Guidance dataset is now available via the Pipeline API.
From the docs:
FactSet Estimates – Guidance is a global dataset that displays an indication or range of a company’s future metrics as reported by the publicly traded firm. Corporate Guidance data is reported by firms via press releases or transcriptions that are formatted by FactSet into a low/high range. By leveraging corporate guidance data, you can analyze a firms forward-looking statements based on key reported financial metrics. The metrics covered by this dataset match the metrics covered by the FactSet Estimates - Consensus dataset.
As a Quantopian engineer, I always enjoy delivering a new dataset to the community; but I'm especially interested in this one, because of an eye-opening personal experience.
I used to work at a large, publicly traded tech company, which published sales and EPS guidance at the beginning of each fiscal period. At the end of one quarter, when we published our results, I was thrilled to see that we had once again exceeded the guidance we'd issued—only to watch in dismay as our stock price immediately declined.
Many of the engineers were surprised, but our finance team was not: they explained that while we had outperformed our own expectations, we still had still fallen short of the market's. Analysts had observed our steady streak of successes, and had become more optimistic about our future than we were!
Ever since then, I've wondered: in what circumstances do analysts' estimates diverge from a company's guidance? Can a difference between them indicate a market inefficiency, or a source of alpha? Now that we've added Guidance to the platform, we can use Quantopian to start exploring these ideas.
Here are a few more questions that the Guidance dataset could help answer:
- How accurate do various companies tend to be in their guidance?
- Which companies are more conservative in their guidance, and when?
- How do these trends correlate with other data? How do they change over time?
- When Guidance and Consensus diverge, when does each one better match Actuals?
- Does Consensus display more or less inertia than Guidance?
- Does it respond to surprise more or less severely?
- How do Consensus values tend to respond to changes in Guidance?
- What happens when a company changes their guidance before they release their actuals? What happens when the change occurs closer to the beginning of the period, or closer to the end?
- Is there a "surprise" effect between Guidance and Consensus? Between Guidance and Pricing?
- How do earnings surprises affect companies that don't issue guidance, compared with those that do?
I'm looking forward to continue exploring this dataset, along with our community. As a Quantopian employee, I can't personally participate in our contest; but I hope that many of you will discover trends worth trading on.
Good luck, and happy hacking!