Venture capitalist are seen as the wisest investors. But even for them finding the ideas that will return 50 times the initial investment is proving difficult. In simple terms 1/3 of companies fail, 1/3 of companies return capital (or make a small amount of money), 1/3 of companies do really well.
To expand on the last statement here is the overall performance of Correlation Ventures for 2003-2014, a study that shows the distribution of outcomes across over 21,000 financings and spanning the years 2003-2014:
- ~65% of the accounts return in between 0 and 1 time the initial investment (basically the most optimistic scenario for this tier is breakeven)
- ~25% of the companies return between 1 and 5 time the initial investment
- ~6% of the companies return between 5 and 10 times the initial investment
- ~2.5% of the investments result in a return of 10 to 20 time the initial investment
- ~1% return 20 to 50 times of the initial investment
- And only ~0.4% return more than 50 time the initial investment. Using this data as a benchmark we have created a calculator to help corporate leader understand in how many ideas they need to invest in in order to move the growth needle.
The calculator we’ve put together is using the Monte Carlo computational algorithm – random sampling is used in predicting the likelihood of the outcome of investing in ideas.
How it works
To use the calculator you need to know how many ideas are you going to invest in and how much will the initial investment be.
The model will randomly assign the ideas to one of the 6 tiers described above using the probabilities shown. Within each tier the calculator will randomly assign a corresponding return value for each idea falling in that tier.
It makes sense to run the calculator multiple times to see the effects of the Monte Carlo computational algorithm. You can also clearly see the difference between making an initial investment of € 500.000 in 10 startups, or € 50.000 in a 100 startups.
Interpreting the diagram
Since the certain outcomes are difficult to predict we are using random samples. The result of random sampling should be interpreted as follows: the highest likelihood value of the return is given by the tallest bar in the graph.