It should come as no surprise to ValueWalk readers that alternative data has an obsolescence value that is negatively correlated to adoption rates. While we have noted data value dropping as usage and knowledge rise, a known concept in the hedge fund industry. What has been unknown until now is the levels to which fund managers must go to ensure the value of their alternative data does not get lost to the dulling mainstream consensus.
A recent Greenwich Associates report, pointing to hedge funds allocating nearly $900,000 per year to alternative data, confirms the trend of alpha is slowly heading into the beta mainstream. The deepening reach for alpha can be found in hedge funds, on average, devoting an increasing amount to obtaining generally unproven data sources that even when acquired can be a challenge to measure effectiveness.
In their report “A Buyer’s Guide to Alternative Data,” Greenwich Vice President Richard Johnson reveals other issues that those wading into the alternative data waters must navigate. Not only are the data sources becoming more popular, reducing strategy effectiveness, but there is also no consistent mechanism for testing or modeling results.
“Standardizing the data was one of the biggest frustrations we heard,” Johnson said of the 40 mostly quantitative investment managers surveyed between March and May of 2018.
Standardization has long been a critical underlying issue in a battle of sorts occurring on Wall Street. Many Wall Street products have been designed without conforming to some level of standardization – OTC derivatives, for example. While the lack of uniformity can lead to increasing the value of the product – customized derivatives insurance can sell at a higher cost than exchange-traded products – standardization can make the product more commonplace as well. This is why those deeply intertwined with alternative data, and quantitative strategies are employing extensive methods to measure the usage of alternative data.
Quandl CEO Tammer Kamel, whose firm was the most recognizable alternative data provider in the Greenwich study, told ValueWalk hedge funds are employing “data hunters” who scour the alternative investment landscape to measure just how unique data has become. But the search for data alpha is just “scratching the surface,” with “way more gold to be discovered.”
The problem is not just an influx of gold miners competing for a limited supply, but the Greenwich report also pointed out that even the most fundamental task of evaluating the effectiveness of alternative data is a challenge. Johnson noted the importance of “unique and uncorrelated” strategies, which do not have an identifiable benchmark, while at the same time pointing to a difficulty measuring the effectiveness of datasets.
“The process for evaluating new data sources is not straightforward,” Johnson wrote in the report, pointing out “frustration.”
Many of the larger data providers are working to provide multi-channel data streams, as the trend of combining alternative data sets with different strategy types has proven successful. The challenge becomes assigning an alpha value to relative to the data or existing strategy, which can be like assigning value differentials to a chicken and an egg.
Measuring the effectiveness of the approach is time-consuming and costly much as is evaluating such strategies. The Greenwich report noted that it could take two highly-paid quantitative engineers on average 85 person-hours to assess a new alternative data source.
While no one source provides an easy and effective method of data modeling sold alongside the alternative data, Johnson sees a need. “There is likely a larger audience who would like to take a pre-packaged signal without having to do all that heavy lifting,” he told ValueWalk. “For these firms, the large market data vendors are integrating alternative data in a more easily consumable way. The downside of that approach is that as the data becomes more homogenized, it can become less valuable.”
In the world of alternative data, unlike free markets, the value of data goes down when adoption rises. But unlike free markets, understanding performance attribution and evaluating strategy effectiveness is increasingly a challenge.