Please contact firstname.lastname@example.org or call (212) 634-9085 and we’ll be glad to answer any questions you might have about our data, or schedule a time to give you a deeper run through via a screen share.
Yes, you can retrieve and download our informational deck here
Our internal quantitative research team has developed various alpha models that we are happy to share with potential clients in order to help them in their research process. We have also worked closely with academics and sell side quantitative research groups to develop other models. These models can be found throughout the various papers that are posted to the bottom of the main data page. We also makes available several Jupyter Python event study notebooks for potential clients to leverage. Our post earnings drift event study and through earnings drift event study can be accessed via these links.
Our testing files for each data set provide a full point-in-time look at all of the data that would have been available to you at that time. The US Equity EPS/Revenue, US Equity KPIs, and Global Economic Indicator data sets each have consensus time series files and detailed individual estimate files. For a full list of variables included in the files please see the comprehensive data dictionary.
Our API is real time, meaning you can pull date from us throughout the day at whatever interval you’d like and you will receive all additional date in our system (estimates, actuals, report dates, etc.). Estimize also provides two solutions for daily delivery. Our FTP access provides files generated nightly and our email solution provides Excel files each morning before the open with data for names on your watchlist.
There are various direct and indirect incentive structures built into the the Estimize platform. The most direct is the give-to-get model we employ where a contributor is shown estimate data from other Estimize contributors for the specific release and specific metric they are estimating on, only after submitting their estimate. This also produces a data set where contributors are not able to anchor their first estimate to a distribution of estimates from their peers, increasing the independence of the data set as a whole. You can think of Estimize as a kind of dark-pool for the trading of fundamental estimate information. Soft incentives include the ability to track all of your estimates in one place, receive feedback regarding your estimate accuracy, both independent and in relation to your peers, and the inward and outward egotistical aspects of our leaderboards. Based on our polling of the community, we also believe that one of the most significant reasons analysts contribute to Estimize is their desire to see a better benchmark of true expectations after decades of the Wall Street consensus being a poor biased measure.
Yes, the Estimize team is working hard to lay the groundwork for the collection of estimates in international equity markets. If you are specifically interested in licensing this data in the future or contributing to these data sets, please feel free to contact us so that we can put a note in our CRM to contact you when they are available.
We use a range of methods to govern the platform and protect against bad date that may be contributed in both innocent and nefarious ways. Our estimate flagging system (see the next question down) does an excellent job of making sure bad estimates do not get into the consensus figures. We also employ IP address checks for multiple accounts by the same user, as well as a manual review process of accounts and the data set on a regular basis. While we don’t make public all of the specific methods by which we protect the platform, we can say that we have never seen a successful attempt to systematically skew our consensus values at scale.
Estimize uses a flagging system for estimates submitted to the platform in order to make sure that the consensus values are clean and do not include estimates that may be unreasonable. The algorithms that Estimize uses to flag estimates at the time of their creation are not meant as perfect assessments of the reliability of an estimate, they are designed as a gate. Estimates that are flagged are put into a queue for a member of our data team to review using a human brain to determine whether the estimate was made in error, is malicious, or just acceptably aggressive. Estimates that are deemed to be reliable after the review process are unflagged. All flagged estimates are reviewed within 8 hours of creation (max), and are never unflagged after 4PM Eastern time for companies that report between the close of trading that day and the open of trading the next. Estimize also uses a flagging system for releases where there is no Wall Street consensus, mostly impacting the Equity KPI data set. In these cases, the first three estimates submitted to the platform for a given release are automatically flagged. When there are three estimates, a member of our data team will review the estimates to make sure they are reasonable, at which point they will be unflagged and the Estimize Consensus will be generated for that release. All estimates contributed after that point in time for that release will be judged by the normal flagging algorithms. In both each of the API, FTP, and data testing files you will see that all individual estimates are available on a point-in-time basis as they are created, whether or not they were flagged. This allows firms that wish to judge all of the individual estimates for reliability themselves in order to create a consensus the ability to do so.
All datetime formats are given is ISO-8601 (see Wikipedia). Time zones are included in the datetime format and are UTC unless specified otherwise.
Yes, we keep point-in-time data for all estimates contributed to the platform. We never delete estimates or change estimates. The sanctity of this data is extremely important to us because our clients need to be able to trust that the historical data they are backtesting is what they would have seen at the time in production.
Estimize is a crowdsourced dataset contributed to by tens of thousands of individuals. It is impossible for us to guarantee the provenance of information contributed to the platform is not MNPI. With that said, we see zero incentive for any MNPI to be contributed to the platform at any time for any reason. As well, Estimize is an open dissemination platform, similar to a newspaper. The moment an estimate is published on the platform it becomes publicly available information regardless of the provenance of information used to make the estimate. Our clients, especially our systematic quantitative clients who use our API, rely on this fact as protection against ever being in the situation where for some odd reason we can’t think of, an estimate does contain MNPI.
Compliance has become an even more important aspect of what buy side and sell side firms do as the use of alternative data and crowdsourcing has exploded. We are always happy to chat with compliance departments. Please email email@example.com or call (212) 634-9085 and we will set up a time to answer any questions you have.
Members of the Estimize community have the option to provide their structured biographical information upon signup or afterwards. This data is optional and self reported. Community members can choose to identify themselves as either “professional” or “non-professional”. Professionals then have the choice of “buy side”, “sell side”, or “independent”. Each of those variables has one level deeper as well, with buy side including “hedge fund”, “mutual fund”, “pension fund”, “venture capital”, etc. Non-professionals are given the option of identifying which GICS sector and industry they are affiliated with professionally, or whether they are a student or member of academia. This structured biographical data is provided in both the testing files and live data.
We include all releases for instruments irrespective of Estimize coverage to provide historical context (for actuals and Wall Street estimates) as well as continuity across quarters for lower coverage stocks and economic indicators.
Please contact firstname.lastname@example.org or call us at (212) 634-9085. If you have filled out the data request form already we will be in touch shortly.
Estimize provides a real time API, daily FTP, and daily emails with attached Excel files to clients.
We have a long history of licensing data to academics which has resulted not only in more than a dozen published academic papers but also significant improvements to the Estimize platform itself. Our full historical data set is available to academics to download and conduct research. If and when you choose to make your research public, in any way, we require you to purchase an academic license. You can also choose to purchase a license up front, but either way Estimize will not impede or make any judgement of your research priorities or conclusions. As a platform and a company we have benefited significantly over the years from the independent nature of research conducted by academics. Please contact us directly if you would like to discuss our growing list of research ideas or any potential experiments we may be able to run together on the Estimize platform.
Absolutely not! You are welcome to become a client and never contribute a single estimate. Though we have seen that for discretionary buy side firms it is common that analysts and PMs end up contributing anyway in order to leverage the analytic tools we provide which measure their own estimates.
Estimize is an OpenFactset partner and delivers its US Equity EPS/Revenue feed through the OpenFactset platform. If you are already a client of Factset data feeds and would like to trial or purchase a license to receive our data through their pipes, please head over to this link and get started.
Yes! We think of Estimize as the iOS of estimates, we don’t strive to build every application or derivative end user use case for the data. Our third party clients include platforms such as ETrade and ChartIQ which visualize data and build derivative features for their clients.
Yes. We are proud to partner with several leading brokerages who conduct both quantitative and fundamental research using Estimize data in support of their clients. Twenty leading brokerages such as Jefferies, Susquehanna, Telsey and others also provide all of their estimates to the Estimize platform in order to be included in the Estimize Consensus. If your firm is interested in participating in the platform as well, please contact us.
Yes. We delineate between individual journalists and public media platforms. We are glad to provide journalists with free access to the Estimize data set and “unblinded” accounts so that they may utilize our data within their stories. Media organizations who wish to visualize Estimize data systematically on their platforms must acquire a data license.
Data licensing to buy side firms most often takes place on a PM by PM or book by book basis. Three main variables associated with each book are taken into account, including the AUM of the book, whether the strategy is systematic or discretionary, and whether it is long/short or long only. We are also happy to discuss data licenses that include full access for entire firms across many groups. Pricing for each of our individual data sets is independent, but we are happy to offer discounts to firms who purchase multiple data sets.
Two main variables are used to build a confidence score applied to each individual estimate within the consensus. Those variables are, the historical accuracy of the analyst, and the amount of time between when the estimate was made and the company’s report date. Estimates from analysts who have a strong track record within that given sector who revise their estimates closer to the report date will receive higher confidence scores and be more heavily weighted in the Estimize Weighted Consensus. For a more detailed explanation of how our model works, please see the document at this link.
Yes. We can also provide unadjusted testing files if you would like.
Estimize uses Factset as its primary provider of these variables, but has access to others as well. On top of the Factset data, Estimize uses humans on our team to double check potential issues where there may be ambiguity regarding the nature of items in Non-GAAP financial reports and other irregular issues that may arise from time to time.
There are a few rare occasions when there are estimates after the report due to either companies accidentally reporting their numbers early or inaccurate report dates from our data provider. We have a policy of not altering historical data so these anomalies remain in our archive.
All estimates are made in the Non-GAAP accounting standard (except in rare cases where companies no longer provide Non-GAAP financials such as Facebook) and exclude options expense. On occasion, revenue estimates may exclude certain figures (as is the case for Google regarding traffic acquisition costs).
Members of the Estimize team are in close contact with both our buy side and sell side contributors for feedback on which variables are most important to investors. With our KPI product, we do not strive to deliver a complete list of variables on the income statement for each company, only the few that truly matter each quarter. We are consistently building our a broader coverage universe of names by focusing on industry groups where KPIs are consistent across companies. If you would like to see specific KPIs added for specific companies, please let us know!
When we initially implemented coverage of our Economic Indicators, we selected those indicators and countries from the G10 that we felt best both provided a useful snapshot of popular indicators and would attract estimates from our users. To that end, we have, in the past, removed indicators that have received little to no attention and are currently looking to expand our own coverage based on feedback from our users and clients.Back to Top
A detailed explanation of how the factor model is constructed can be found at this link.