By Jeremy Davies, Co-Founder of RSRCHXchange
Last week in the middle of the Hudson river, on board the USS Intrepid, Battlefin and Jefferies hosted the Battlefin Discovery Day. The conference was catchily strap-lined as ‘where alternative data companies come to meet one-one with hedge funds and investment firms that are looking to use alternative data in their investment process’.
As the quest for alpha has hotted up in the last few years alongside the rise of big data, it is no surprise that investment managers have been led towards new datasets to incorporate into their investment process. Originally the preserve of quant funds with banks of data scientists digesting petabytes of data, this is starting to become more mainstream. Whether it is Alt-data sources being used by banks and brokers to augment their research notes, or the data providers themselves shifting from unstructured data to semi-structured – and even producing research product – it is clear a shift is beginning to occur.
We see this too in the number of new entrants and data sources joining the fray. The investment community has become increasingly aware of geolocation data, web scraping, jobs data, social media scraping, satellite data etc. and these sources have started to become more structured, so that in many cases they now relate back to individual companies/tickers or to industries.
For those operating at the more macro level these micro level sources are often difficult to aggregate up. However, one dataset we at RSRCHXchange have been familiar with for some time does have an easier read for macro investors. That dataset is polling.
In the past few days there has been a lot of coverage in the media of private polls conducted by a number of hedge funds around the Brexit referendum and the profits those funds made from correctly predicting the outcome.
I think it is worth stepping back from the detail of who did what exactly and accepting that some level of truth exists in the fact that a number of hedge funds, and macro funds in particular, do use private polling as an input in its own right and as a data input to augment their own internal models. By all accounts they have done so with tremendous success.
Polling is not ‘precise’ in nature and it is ‘generally available’ to market participants who wish to conduct polls and therefore it is not even close to being inside information (but don’t let that get in the way of a good story). Polling is a gauge, a canvassing of opinion, just like market research. As we have seen in recent years polling can often be wrong, especially the small polls conducted for the press. But polling is a very useful input into decision making processes and models.
As political instability and uncertainty seems to be a truth of our modern age then polling should benefit – both from the regularity with which political events rock markets and the phenomenon which Battlefin and Jefferies pick up on in their strapline, with more investment firms looking for alternative data to use in their investment process.
Putting Brexit aside, in the last twelve months in Europe alone we have seen surprise results in both a General Election in the UK and an election in Italy – with potentially more to come – which have impacted markets. In the coming months we have midterms in the US to look forward to and pollsters and their new found clients are likely to take their lessons in Europe to the US. It seems likely that pollsters and hedge funds will continue to engage wherever potential volatility lurks.