Category: lessons

Lessons from our Anti-Portfolio


This post is by Cathy from Seedcamp


By Carlos Espinal and Reshma Sohoni

Investors are generally judged by the size and volume of their successes, for, after all, it is expected that you will lose some to win some, but rarely do investors talk about their mistakes and misses. There are a few investors out there, however, who have had an investment portfolio they call the ‘anti-portfolio’ which highlights their most regretted ‘passes’. Effectively, which companies would have yielded them amazing outcomes had they invested. Not all anti-portfolios are the same, however, sometimes companies will go up before they ultimately don’t materialize into what was expected of them. So in the short term, some companies might be part of an investor’s anti-portfolio, even though in a longer term, they might ultimately (or not) be satisfied with their decision – just comes with the territory.

It is, therefore, no surprise that in spite of our successes, with 280+ companies backed to-date and $5Bn in follow-on funding raised by our portfolio, here at Seedcamp we’ve had our own anti-portfolio in the making since 2007. This includes companies such as Citymapper, Bulb, Patch, Flux, Strapi, Impala, Colvin, GTMHub, Cognism, Unmind, Patch, Rahko, Pipedrive, Deliveroo, and SimilarWeb to name the most notable ones. Whilst the process of decision-making internally is beyond the scope of this post, what I would like to do is highlight the kinds of issues we felt companies like the above had that, at the time, we felt were sufficiently significant that we chose not to invest. Hopefully, this effort to highlight the investor decision process (Read more...)

COVID-19 Lessons: The Cost of Half Measures



There is so much to be learned from the entirely botched handling of the COVID-19 crisis. There are lessons about markets compared to command-and-control, wartime versus peacetime institutions, decision making under uncertainty, scientific progress and much more. I am likely to write a post about each of these, starting with today’s installment: the cost of half measures.

Let’s start with the observation that most people still don’t seem to understand exponential growth (another lesson right here). Any transmission rate R > 1 results in exponential growth. Sure, R = 1.2 is not as fast as R = 1.8 but it is still exponential which means that eventually infections start to really add up. Here is some simple math: 1.8 ^ 10 = 357 whereas 1.2 ^ 10 = 6. So 1.8 seems to be much much worse. But if you turn the question around for a moment you see the problem quite quickly, how long to get to 357 with R = 1.2? Naively you might approach this as 357/6 = 60x slower, but in reality 1.2 ^ 32 = 341 and 1.2 ^ 33 = 410, so you get to the same level of infection at only 3.2x to 3.3x slower!

This is exactly what we are seeing play out in most countries. We never got R < 1 sustained because we took half measures. Instead we simply reduced R but stayed in exponential growth territory. We were buying time, but because of the exponential nature of viral growth (Read more...)