Sign up for free newsletter


Emmett Kilduff, Eagle Alpha

Flying high

As alternative and traditional fund managers look to enhance their trading strategies and move away from crowded trades, James Williams discovers that Eagle Alpha is leveraging the explosion in `big data' to offer buy-side firms alternative data sets to harness alpha.

One of the biggest challenges for any hedge fund manager is finding novel data sets that they can utilise to develop new investment ideas for the portfolio and steal an edge on the competition. The problem is that traditional data sets are used by all and sundry, and while some brokerages excel at producing niche sector reports in small cap equities, the majority uses the same data. 

In that regard, it is a problem for the big mutual fund houses as well. Eagle Alpha is a pioneer in the alternative data sector and is working to meet this challenge head-on. In brief, the objective at Eagle Alpha is to provide investment managers with an array of ‘alternative data’ tools that have the potential to generate new insights, and, by default, market alpha. 

“I left Morgan Stanley in 2012 to set up Eagle Alpha on the basic premise that the amount of ‘alternative’ data is growing exponentially today but a lot of buy-side firms aren’t taking advantage of it,” explains Emmett Kilduff (pictured), CEO and Founder of Eagle Alpha. “At a high level, we define alternative data as ‘non-traditional data’ that can be incorporated into the investment process. More specifically, we define it across 20 different categories such as online search data, trade data, satellite and weather data.”

The three most common categories used by the buy-side, says Kilduff, are consumer transaction data, application usage and web traffic data, and thirdly geo-location data. 

Traditional data sets are available to the market from the usual sources. It is data that is easy to source.

At the other end of the spectrum one might have consumer transaction data, which doesn’t have a long history, isn’t available on well-known platforms and is therefore non-traditional. Social media is still a relatively new concept. All the data generated by Facebook, Twitter, LinkedIn is non-traditional data. 

“However, in five years’ time it will all just be data, regardless of whether it’s traditional or non-traditional,” says Kilduff.

“This is the next major change in the investment process; it’s something that hasn’t happened since 1998 with the introduction of expert networks. We see a lot of fundamental asset management firms moving to adopt quantitative strategies. One of the ways to achieve this is by implementing alternative data into their investment process to gain better insights.” 

There are five parts to Eagle Alpha’s offering: thought leadership; data insight; analytical tools; raw data, and bespoke projects. 

Thought leadership is offered because most buy-side firms haven’t yet integrated alternative data into their environment. There is a large focus educating investment managers on the best processes to use to leverage alternative data and to develop ongoing best practices, says Kilduff. As for the raw data component, this is offered by way of a data directory that keep track of the best emerging alternative data sets worldwide. 

“Our data procurement specialists are constantly sourcing interesting new data sets; to date, we have handpicked 490 data sets and we can help clients to navigate the best data sets for them to try. With regards to bespoke projects, we just launched this in January 2017. Some firms don’t have the capacity to crunch through lots of data and can now use us to do bespoke work for them,” confirms Kilduff.

Ronan Crosson is Director of Data Insights at Eagle Alpha. He explains that the firm’s in-house team of data scientists and engineers are constantly finding interesting data sets, which then get cleaned and structured. “Then, the step we are most heavily involved in is alpha extraction by way of data insights. The reports that we produce look and feel similar to traditional investment banking reports, but the data underlying them is very different, as are the techniques that we employ. 

“We have skill sets within the team that are not available to traditional analysts,” explains Crosson. 

The reports focus largely on macro developments in the markets and equities; specifically US and UK consumer discretionary and technology stocks.

To illustrate the efficacy of these data insights reports, in a report on 15 December 2016 Eagle Alpha (using Google Trends data) wrote that its Search Signals index on Finish Line pointed towards a weakness in the company’s same store sales. 

“The sharp downtrend observed in the index in recent months indicates that expectations for SSS growth at FINL are at risk, or that management outlook for the February quarter may disappoint”, said the report. The Search Signals indicator was prescient in providing directionality on this key company metric. 

The company’s stock reacted to the change in fundamental outlook and is down significantly since the day of the announcement; currently -24.7 per cent since they reported in December.

“We do a lot of work to back test data sets to test correlations, to build models and perform out-of-sample testing; testing the predictive power of the data sets. That gives clients confidence that the data and the insights we are extracting are valuable,” says Crosson.

One significant development in recent weeks is the announcement by Eagle Alpha that it has added a China Auto Insight alternative data set to its stable of partner data sets, which buy-side clients can choose to access; other examples are UK housing data, US job postings data and Global EPOS data that provides insights into consumer trends.

This is the first time such granular data has been collected on China’s automotive industry, cleaned and made available to the finance industry. 

“Our partner in China is the leader in providing data on China’s auto industry. They are covering 80 per cent of the market in terms of the manufacturers they work with. And these manufacturers are their own clients so that gives you a sense of how valuable their data is,” says Kilduff, adding that the data is also unique in terms of how it is collected. 

“It is collected using a panel of over 1,300 Chinese dealerships, combining other data sources such as web data and more traditional data sets to create a large and well-structured database. 

This data will be useful for those looking to invest in China’s auto industry as well as global car manufacturers that have large exposure to China. China is now the number one car market globally and also one of the fastest growing. A lot of Western brands have dependence on China and it’s therefore important for them to have access to this data.” 

The auto sector is also an important economic indicator of consumer health, consumer spending, which can be useful to macro investors who are looking to get early signs and leads on China’s wider economy. Crosson says that there are four main components to the data set: volume sales, transaction price, rebate (insights into manufacturers’ promotional activity), and showroom indicators (real inventory, real data on enquiries).

“The CAI data set will provide investment managers actionable insights on the Chinese auto industry, well in advance of industry participants reporting their results. This supports Eagle Alpha’s objective which is to enable asset managers to obtain alpha from alternative data,” adds Kilduff.

“It is critical nowadays to analyse the performance of dealers at the local level. The data allows investment managers to assess opportunities in a competitive market with many participants,” opines Crosson. “The findings will better inform numerous strategies including equity long only and long/short, credit and macro.”

In Kilduff’s view, data beats opinion. Investment managers are always looking for a new edge. “Corporates want to know what consumers think when the latest iPhone is released, for example. They are analysing consumer behaviour data so the buy-side really ought to be too. 

Most of the hedge funds we speak to want to think about next quarter’s earnings for consumer stocks and a lot of our quant fund clients want to make sure they are not missing out on the next best alternative data sets.”

The amount of data being created is growing exponentially. In a few years’ time, says Kilduff, there are going to be thousands of alternative data sets that will create opportunities for investment strategies to generate alpha.

“In the last two quarters of 2016 we saw a significant uptick in interest in alternative data sets among traditional mutual funds as well as hedge funds. It’s not just the most innovative active managers using these data sets; the level of adoption across the asset management as a whole will, I believe, continue to spread in 2017,” concludes Kilduff.

other gfm publications