Remote working can be your investment edge. Here’s a quick summary how

David Ardagh
auquan
Published in
3 min readMay 21, 2020

20 years ago we had mainframe computers. To solve a harder problem, you had to physically add more memory, then more computers and finally get a bigger office. This made sense. Computer memory is a physical thing, so the more you need, the more space you need too.

For thousands of years, this has been true of any problem. If you want a bigger building, a larger empire or even just more sales, you need more resources. In other words, more people. In today’s world, this means getting more desks and then a bigger office.

Nowadays this seems crazy.

On-prem servers and computing resources are mainly reserved for specialist use cases. Computing resources can be scaled instantaneously, thanks to cloud computing services.

However, very few companies to date have taken the same approach for their people. Arguably some of the slowest to move have been investment firms. This was until CoVid-19.

Now, government-mandated social distancing has forced firms to scramble into action as not to lose out against their more agile peers. For most firms, the amount and importance of their work is increasing. Yet, the amount of people, and their productivity, is decreasing. For some of us, this limited access to talent and resources will cost us our companies. For others, it might be millions of dollars wiped from our portfolios, our reputations in ruin and our careers in jeopardy. But, is there something we can do about this?

I believe there is.

Remote work may have been pushed on us as a necessity, but that doesn’t mean we should accept it as a weakness. Instead, by adapting our business processes there are many situations where it can be our strength. Here are some examples:

Discrete, low-skill time consuming tasks

For example, labelling a dataset to see if China is mentioned in a political, economic or business context across 100,000 news articles. Crowdsourcing this work on micro work platforms can be an extremely cost effective way of completing this task. With smart incentive structures you can collect high quality data at a fraction of the cost to do so in-house. (e.g. Mechanical Turk, CrowdFlower)

Short-term, one-off projects (well defined)

For example, running a quick analysis on a new transaction dataset to check if there is any signal in predicting company revenues. These types of tasks often are accompanied by a lot of opportunity cost, with an unknown reward. By outsourcing the initial analysis or MVP you can keep your team focused until you’re sure it’s worth investing in. (e.g. UpWork, Fiverr)

High skill or longer term projects (well defined)

For example, a team of quants looking to add a new, semantic search feature to an existing research platform. The specialised NLP knowledge here is not needed full time so getting an expert will be faster, cheaper and result in a better outcome. There are a few more options at this level, with some platforms taking a more crowdsourced consulting approach and others leaning towards open-source software development. (e.g. Bounties, Talmix, TopTal)

Open-ended and non-deterministic projects

For example, creating an algorithm to predict the change in market direction based on news sentiment with maximum possible accuracy. Multiple users compete to find the best solution and just by virtue of throwing so many things at the problem, the problem is likely to be solved and solved well. Having many solutions may allow you to combine approaches, with every incremental idea increases the accuracy by the little extra bit. Combining multiple such ideas can boost performance from 52–53% to 60% or more. (e.g. Kaggle, QuantQuest)

The QuantQuest platform, which we run, builds on the strengths of open-source communities and project platforms to give you access to 15,000 world-leading data scientists. These experts compete to give the best solution and earn a prize. Our community of experts have previously helped with problems as diverse as understanding which macro features most affected their portfolio and building NLP tools to extract text information to build out knowledge graphs. Whatever your need, QuantQuest can help you excel.

To learn more visit us at: quant-quest.com or auquan.com

--

--

David Ardagh
auquan
Editor for

Cornish born and working in a Fintech in London (how original). I try to make big things simple.