Thinking “out of the box” has taken on new meaning as asset managers mine previously overlooked or insignificant data to ascertain which topics are trending and which can be ignored.
For example, one of the world’s largest private equity companies recently used employee-administered drug test data gleaned from potential hires to project whether employment rates in the U.S. are trending up or down. The quirky approach isn’t unique as private equity companies adopt unusual methods of gathering and analyzing data, discovering that creative methods can produce results that are, in their opinions, as insightful and accurate as traditional methods.
Boston Consulting Group partner Craig Hapelt is an advocate of unorthodox research methodology to mine big data. He agrees that contemporary times require contemporary approaches. He’s not the only fund manager willing to blend disparate resources like Twitter, Starbucks and Google Flu Trends to forecast based on internet search terms.
Hapelt is joined by forward-thinkers like Nick Thomas, a Baillie Gifford partner, whose Edinburgh-based enterprise is populated with researchers who aren’t shy about mixing and matching Google-generated data with facts and figures extrapolated from Baidu and Amazon to assess financial directions internet businesses are taking. Thomas admits that this is a slow process requiring a long-term approach.
Is taking an unconventional route to data gathering for the faint of heart? According to Peter Harrison, head of Schroders’, not every asset manager is comfortable with it, but he believes that if asset managers don’t work smarter, they can be left behind. He points to US pension and sovereign wealth funds driven by untraditional data collection methods that improved internal investment strategies and produced positive results within a year.
Taking so revolutionary an approach to data mining has introduced a new concern: entities like Amazon and Facebook could conceivably decide to enter the asset management business. After all, power players have a better understanding of data analytics than most. One savvy prognosticator wishing to remain anonymous joked that Google’s mix of talent and wherewithal could conceivably “clone an asset management stalwart before breakfast.”
RavenPack’s Chief Executive Armando Gonzalez works with what he refers to as “half of the world’s biggest quantitative hedge funds.” His customer list has grown large thanks to RavenPack’s ability to morph solid, market-moving data compiled from news feeds and social media into nuggets of rich data, but all is not blissful in this universe where forward thinking can be impacted by social media fickle enough to give those making investment decisions pause.
Gonzalez warns that public sentiment isn’t yet refined enough to recognize insider trading detection and monetary policy changes that can mask red flags and optimism. He cites a low score given UK supermarket chain Tesco following his company’s analysis that was impacted by social media sentiment that could have swayed the likes of Warren Buffett. The investor extraordinaire regrets investing in Tesco based on the company’s £250m overstatement of forecast profits.
Michael Hintze, CQS chief executive, cautions asset managers to err on the side of caution when entering this parallel universe of data-gathering. He advises filtering and harnessing unstructured data, unearthing context and appropriately timing trade executions. Professor David Hand of the UK hedge fund group Winton agrees with Hintze, warning that relying on data alone can be dangerous. Asking the right questions and threading through noise offers a clearer picture of market behavior.
Toward that end, Hand is hiring 140 researchers to focus on trend-based data analysis. This huge investment in personnel is critical, he feels, noting that traditional asset managers must come to grips with the fact that this is the way of the future. Those ignoring this reality could conceivably be left behind.