Neural Alpha | Avoiding overcrowded investment strategies
portfolio_page-template-default,single,single-portfolio_page,postid-71,ajax_fade,page_not_loaded,,qode_grid_1300,qode_popup_menu_push_text_top,qode-content-sidebar-responsive,qode-theme-ver-10.0,wpb-js-composer js-comp-ver-4.12,vc_responsive

Avoiding overcrowded investment strategies

Overcrowding has been widely cited as one of the most significant reasons for poor hedge fund performance in recent times. Indeed some analysts such as Morgan Stanley’s Adam Parker cite it as the most significant inhibitor of hedge fund returns. Whilst the underlying reasons for this herd behaviour are debatable one strong hypothesis is that too many funds are chasing the same sources of alpha as a result of using the same data sources. This has led to the rise of the Alternative Data movement that seeks to avoid overcrowding by utilising under-used data sets for insights. Examples cited in the press have included satellite photos of parking lots to develop predictive models for sales volumes as well as weather data to predict crop yields. However much of this data is sourced from third parties that conduct acquisition, standardising and anonymization before making it available on public data supermarkets. This misses a significant opportunity to apply a manager’s unique investment insights to truly unique and less commonly analysed datasets.


Alpha rich ‘Alternative Data’ sources


Investment managers have good knowledge of the types of challenges facing their particular strategy, sector or company questions. They are familiar with the lingua franca used to describe these challenges but not how to source the necessary data, prepare it and analyse it. Often this terminology differs across industries and what may represent a challenge for one industry may be an opportunity for another.


For example within the Renewable Energy sector different measures are used as headline statistics (MegaWatts, Gigawatt hours etc) to those of other industries such as marketing where the cost per eyeball may be the most significant metric affecting share prices. Academic research papers similarly will often use industry specific terms to describe discoveries, patent filings or challenges. Whilst a Manager may be familiar with these terms and concepts they typically do not possess the means to search for relevant information and analyse it.


Neural Alpha


We leverage your detailed sector knowledge in conjunction with cutting edge machine learning technologies in order to create customisable knowledge graphs that provide cutting edge investment research capabilities. This allows clients to significantly extend their existing research capability to tap into alpha rich, under-used datasets that are unique to their investment focus. This differs from traditional approaches to Machine Learning which focus on standardised models that are compromise on accuracy for ease of use.

We work closely with your investment teams to create a bespoke approach leveraging the most appropriate datasets to your style of investing. Contact us to find out more.


Things we like

Twitter feed

Neural Alpha
@NeuralAlphaJul 02
Microsoft Concept Graph and Concept Tagging Release
Neural Alpha
@NeuralAlphaJul 02
Webinar: Driving Real Results from Alternative Data [SPONSORED] | Auto Finance News
Neural Alpha
@NeuralAlphaJul 01
6 recommendations to CFPB regarding alternative data