Nasdaq Launches Analytics Hub for Buy Side

Terry Wade, senior vice president and head of product and business development at Nasdaq Global Information Services, says the hub has been developed in response to demand from asset managers and hedge funds who are constantly on the hunt for alternative datasets.

“There is a need in the marketplace for an ever-growing supply of data, but also a need for datasets that have been cleansed, validated and vetted over a long period of time,” Wade says.

Clients of the hub currently have access to core Nasdaq data, as well as datasets from PredictWallStreet, which polls retail investors to produce long and short signals with high levels of alpha; iSentium, which converts unstructured social media content into sentiment indicators; Prattle, which quantifies the language of central bank and corporate communications into directly tradable signals; and Nasdaq Dorsey Wright, which identifies investment ideas by ranking stocks, exchange-traded funds and mutual funds from strongest to weakest based on relative strength analysis.

Each dataset applies a slightly different aspect of machine intelligence, Wade says. The social media sentiment produced by iSentium, which is mainly used in short trading cycles, applies natural-language processing to raw social media feeds to convert those into sentiment and then draw correlations to the financial markets. PredictWallStreet gathers sentiment from retail investors and then analyzes that using a series of algorithms to identify the predictive patterns within that data.

The hub provides investors with access to not only data from individual data vendors, “but some value-added signals derived from machine intelligence that they may have not been able to access on their own,” Wade says. “We bring the data in and use a series of algorithms to analyze that data and produce an event analysis on top of that data to provide, in some cases, different time-series signals or different use cases for the data. Some of the partners have been marketing their data more as a daily signal, and through the use of machine learning we’ve been able to identify those instances when the data is also appropriate for longer holding periods.”

All Nasdaq data customers can use the hub, and can access the data at different levels of processing and sophistication appropriate to their needs. “Very sophisticated funds can buy something much closer to the raw signal and process it themselves,” while others that are newer to using alternative datasets “might decide to use more of the process signal to just get a plus one or minus one, or whatever the score is that’s coming out of the signal and use that in a much more processed way,” Wade says. “You have all sorts of funds that are running all sorts of algorithms today, and they’re going to take this data and see whether it has value on its own, or if it adds value incrementally on top of the algorithms that they have today. There are an infinite amount of strategies that you can run in the marketplace. As companies begin to take action on data, then different signals are going to rise up. There will be a constant need to experiment with different datasets to take on the current conditions of the marketplace.”

Wade says Nasdaq will rapidly add new datasets to the hub, but adds that the exchange is looking to strike a balance between “new and creative” data sources. “Funds are after anything that is new, but there’s a balance there: If something is very new and you can’t create a vast history of it, then the company doesn’t have the ability to back-test it very well,” he says. “There are a number of funds that just will not buy it if they cannot back-test it.”

The hub was developed in partnership with analytics firm Lucena Research, which vets the datasets, while Nasdaq uses Lucena’s QuantDesk platform to create some of the value-added signals.

The Analytics Hub is the second product to launch from Nasdaq’s Innovation Lab, which was launched last year.