Can investors' decisions be influenced by the weather?

Imagine you are a venture capitalist looking to invest in an early-stage startup. There are myriad such businesses out there (there were close to 5,000 seed-stage investments globally in 2019), all competing to attract your attention and part with your doubtless hard-earned cash. How do you decide who you are going to fund?

In theory, you are going to use the most relevant and up-to-date research you can lay your hands on, carry out exhaustive due diligence, interrogate all aspects of the business proposition and the founders, and draw up the most rigorous set of criteria on which to base your totally objective decision.

The entrepreneurship literature is rich with studies of investment drivers, which broadly use two distinct approaches. One approach focuses on startup-specific characteristics; while the other is concerned with investors’ attributes. Both reveal behavioural factors, such as social and cognitive influences, that shape investment track records, and suggest that variation in investment decisions is a function of the characteristics of the startup-investor pair – hardly a surprising finding, given that the literature has concentrated on situations where information about startup business traction and/or investment track record is readily available.

Lack of information

But, as Gary Dushnitsky, Associate Professor of Strategy and Entrepreneurship at London Business School (LBS) and Sayan Sarkar, PhD student at LBS, detail in their new and highly imaginative research study, “Seed-stage ventures have no sales traction and face uncertainty on multiple fronts, including the viability of the technology, the ability to attain product-market fit, business-model feasibility, and so on”. This means that seed-stage investment is actually characterised by scarce information, great uncertainty and – given the many competing sources of funding available today – significant competition to find a ‘good bet’.

The research, entitled ‘Here Comes the Sun: The Impact of Contextual Factors on Entrepreneurial Resource Acquisition’, published in the Academy of Management Journal in December, draws on affect-as-information theory to consider the impact of incidental contextual factors – such as changes in the physical environment – on investor mood and resulting likelihood of investment.

The effect of such factors may seem impossible to capture, yet alone measure in any meaningful way, but the researchers hit on the ingenious idea of using information from startup ‘Demo Days’ to capture and quantify just such an impact. (Every startup graduating from an accelerator has the opportunity to pitch to potential investors during a graduation day – known as Demo Day – which is scheduled by the accelerator well in advance).

Using proprietary data on nearly 1,400 startups graduating from European accelerators, the authors found that those who pitched when the Demo Day fell on a day that was sunnier than the previous one experienced an increase in the likelihood of actual investment in the pitching startup. Moreover, this positive sunnier effect was stronger for nascent ventures and those where the founders had limited human capital. Critically, the study not only found a positive sunshine-investment association, but also established that the association was, indeed, mediated by investor mood.

 

Tipping the balance

Establishing that incidental factors (such as a change in the physical environment) can “tip the balance towards specific decisions” has important implications for investors.

To see this, recall that investors need to expend great time and effort in collating data in order to inform their investment decisions and that, broadly speaking, the data relates to two distinct ‘pools’: startup characteristics (is the founder experienced? Do they have a computer science degree? Does the business own the IP? Has it gained traction? and so on), and investors’ attributes (cognitive style, social influences, ethnic and gender preferences, etc.)

The research shows that, in fact, investors’ decisions should take account of a third ‘pool’: data about contextual factors, because they move the needle as well. The reason this is increasing in practical importance relates to one of the most exciting recent developments in the startup ecosystem: quantitative venture capital, whereby investors are developing and deploying AI tools to make investments.

Now, quantitative VC analyses run on algorithms, and the algorithms look at historical data (what entrepreneurs and investors have done so far). To make them more accurate, quantitative VCs are dedicating resources to collecting yet more data about startups and founders (including trying to better understand investors’ past investment decisions).

The critical contribution of the paper is to show that, to develop more effective algorithms, investors may be better advised to collect data on contextual factors, rather than pour more resources into collecting and interrogating the ‘traditional’ pools.

As Gary Dushnitsky says, “That’s the big picture here – we have lots of data on contextual factors, such as sunshine, nowadays. There’s lots of readily available data that is actually more cost-effective to collect and which might increase the efficacy of your algorithms.”

“Ultimately, when it comes to the use of algorithms, it’s all about marginal differences. Take two similar startup-investor pairs. One startup gets funded and the other one doesn’t – the literature says it’s because one startup is better than the other, or one investor was more biased than the other. What we are saying is, that’s not necessarily true – it’s possible the two are exactly the same. Why did one get funded when the other didn’t? Because of the sun! It illustrates that the algorithm would fall short of explaining what is actually happening, and therefore the investor would not be as effective in making their investment decision; whereas if you incorporate contextual cues (such as sunshine), you can explain the difference. We believe this will allow for an improvement in VC algorithms.”

 

Europe: an entrepreneurial hotbed

The location of the study is also highly significant, in terms of both the academic literature and real-world implications for VC algorithms. There are thriving accelerator ecosystems in many parts of the world but, as far as the authors are aware, all the academic research into them to date has been done either in the US or in emerging markets, which have a very different feel and purpose. This appears to be a gap in the research – not least because Europe prides itself on being an entrepreneurial hotbed – so it is significant that this is the first study of a European accelerator on such a large scale, looking at startups graduating from many different accelerators over a decade.

Given the authors’ immersion in the world of rigorous academic research, the weather also had an instrumental role in the genesis of the idea for the paper. Dushnitsky happened to glance at the LinkedIn post of a London-based VC investor (whom he knew to be actively investing in seed-stage startups) that featured the screenshot of a London street bathed in sunshine and a caption that read: “It’s hard to beat London in the sunshine. Even the most mundane scenes look beautiful J”.

For Dushnitsky, it was almost a ‘eureka’ moment: “I thought, ‘Oh my God – this person has lived in London for over a decade and, after being here that long, he’s just walking down the street when he stops, takes a picture, and posts it on LinkedIn.” It suddenly made me realise just how much sunshine makes you feel better.

“Then I thought, ‘Here’s this person who makes investment decisions on an individual, basically just using a seven-slide PowerPoint presentation. The kind of feelgood factor that would move them to post on LinkedIn would surely – at the margin and when no other information is available to them – sway them when it comes to their investment decisions.’