geolocation_banner.jpg

Long term strategy

On September 26th, 2019 we were pleased to be able to host a discussion of long-term strategy in investing featuring:

  • Valerie Valtz, COO of Fundamental Equities at D.E. Shaw

  • Lisa Schirf, former COO of Data Strategies Group and AI Research at Citadel

  • Angela McNab, former COO of Systematic Strategies at BlueMountain

Our panelists spoke about what it takes for investment firms to remain competitive as well as how they can best integrate new resources and tools into their established processes.

Summary

Future of investing

  • In the old days, any PM was able to hang a shingle and have a decent chance of success; now larger investments are needed for firms to be successful - larger, established players are at an advantage

  • At the same time talent and resources in the market are much richer than they were just a few years ago; even a small player can cobble together competitive capabilities - but they need to think like a product manager rather than just an investor / quant / CFO

  • Larger firms are also shifting from having siloed teams to having more of a systems approach

  • Diversity has historically been seen as a nice-to-have, but is now often thought of as absolutely essential to maintaining edge

Limits of alt data

  • In talking about data science everyone likes to discuss the end-state - few really go in the weeds on how to take the first steps in incorporating a more systematic thought process in traditional investing

  • Though there are certainly exceptions, at most firms there is a culture gap between data scientists and traditional investors; unfortunately close collaboration is necessary for a data effort to be successful

  • Confirmation bias is very common among investors just beginning to use data

Role of diversity

  • Diversity of thought comes down to identifying blind spots; both fundamental analysts and technical specialists have a lot to teach each other

  • Highlights from relevant literature:

    • According to Alex Pentland's Social Physics, a tight-knit network tends to become an echo chamber; adding an unconnected node adds information while connected nodes add more of the same

    • Diverse groups tend to generate more thoughtful and more accurate conclusions; this has been observed specifically with juries (source)

    • Promoting one kind of diversity makes organisations more welcoming of diversity of thought in general; for example, more racially diverse groups exhibit less sexism (source)

Getting credit

  • It can be difficult to course-correct once technical hires have been made - so need to try and figure out how to onboard, integrate, and ultimately assess them from the get-go

  • The cultural shift necessary for successful integration of technical hires into discretionary investment firms needs to be driven from the top; management has to make a commitment not only to back an initial launch but also to stick with it through the inevitable trial and error

  • Data scientists within discretionary funds seldom have a clear P&L impact, but there are multiple other ways to quantify the value being created - for example you can look at:

    • How many projects have been executed

    • How often each project gets accessed

    • How many "repeat customers" each data scientist has

  • To keep data scientists and quants from being seen as outsiders, it is helpful to have them sit alongside analysts; it's also good to have dual reporting, both to data leads and PMs

On talent

  • Silicon Valley talent tends to be put up on a pedestal but is often a less than ideal fit for asset management; cultural differences aside, they tend to want to work on products

  • Understanding motivations is key to identifying great talent; look for people who want a challenge - the signal to noise ratio is lower in investing than in other fields

  • Don't overhire; building out a team of rockstars will inevitably lead to discontent as some of them will end up having to do work that doesn't begin to challenge them

  • When onboarding novel resources, you can't just plop them in the way you would a banker from Goldman or a devops hire; management needs to prioritise cross-team integration, and in some cases this may necessitate making incremental investments in HR

  • Anticipating a culture clash can be a self-fulfilling prophecy; for example, establishing a permanent translation layer between data scientists and fundamental analysts often limits direct communication instead of improving it

Blind spots

  • Identifying blind spots often just comes down to company culture; some firms are more welcoming of change, others less so - and it is important to be honest about where you stand

  • Instead of initially focusing on shortcomings, in internal messaging around new initiatives firms should focus on opportunities for growth

  • The whole point of having a culture welcoming of diversity is to make it easier to identify and address blind spots