Each element of a company such as its CEO’s resume, the scientific validity of its idea and the track record of its backers, are given a numerical value. These data are then are fed into the algorithm and the computer gives it’s suggestion on whether the company should invest or not, in less than two weeks.
“It’s kind of like counting cards in blackjack,” Correlation co-founder David Coats said. “You may lose any individual hand, but if you play enough hands, you should win. The odds are truly in your favor.”
But can you purely model success or failure?
How are the founders so confident?
The algorithm is based on history. The business, which has running since 2012, has created a database of more than 80,000 American venture capital deals from as far back as 1997. Every aspect of these deals is also assigned a numerical value. Coalesced, these figures create patterns, which show which deals and which aspects of those deals worked and which did not.
Using this information, the firm’s algorithm predicts which future deals will pan out and which will not. Ultimately though it is up to the discretion of the firm’s directors, which of the computer’s suggestion they decide to invest in although they will not invest in any the computer does not recommend.
Only 10% of the all potential investments pass this test. Some of the companies they have invested in include a drug developer looking to treat cancer, an IT company hoping to make doctor referrals more efficient and a start-up that has created a device that can ablate uterine fibroids and alleviate heavy menstrual bleeding.
Healthcare in the hands of venture capitalists
What capital investors identify as valuable may not align with important health needs and this is primarily because that which will brings the highest returns might not be what the healthcare industry needs. If they do prove to be the same, then this is more likely to be a coincidence as several survey findings show that 85% of capital investors consider public health ‘not at all or somewhat important’.
This is expounded in the case of Correlation Ventures since none of its decision-makers see all the potential investments so there is no moral or emotional pull to any. Although it is true that deals which succeeded in the past, may not in the future, having 20 years worth of data means the algorithm accounts for both booms and busts and this gives the company confidence. Additionally, the firm only invests in those start-ups that have already received funding from other venture capitalists.
But this proves troublesome because it is not just financial gain that influences the decision. Investors also look the number of regulatory requirements in the area. Ideas with few regulatory barriers are more likely to be accepted and invariably these fall within areas that already have a high level of investment and similar technology is already implemented and marketed, so new technologies may find it harder to gain investment.
Venture capitalists can also affect the kinds of technologies available to patients, clinicians and healthcare systems by defining the milestones such as clinical trials and regulatory approval and this can affect the design process, even going so far as altering the primary use for a technology.
Does the algorithm really work?
Since Correlation Ventures does not disclose its returns it is difficult to discern if its methodology is working but it is believed to be outperforming its contemporaries. Despite this, there have been few significant scientific breakthroughs.
It invested in Flex Pharma to produce a drug that prevents muscle cramps but this failed its first major clinical test. It also invested in Mirna Therapeutics for a cancer drug but this failed because of its innumerable and severe side effects. Its deal with Aldea Pharmaceuticals also failed because the start-up’s plan of a drug to reverse alcohol intoxication did not work.
Despite these failures, the diversity of Correlation Ventures’ current investments spark hope for the future of invested healthcare; from biopharmaceuticals that target oxygen metabolic pathways to patient controlled breast tissue that inflates after mastectomies. MIMS
Hospitals may be investing too much in technology and IT systems
Big data analytics may lead to improvements in healthcare
Reality TV might be the new way for the pharma industry to advertise