We believe AI will shortly cease to exist as a segment, shifting M&A activity from current ‘capability buys’ to much higher value ‘platform buys’
There have been 140 Artificial Intelligence (AI) M&A deals so far, with $5B of funding into AI companies in just 2016, according to industry analysts Cognition X (http://cognitionx.com). What didn’t exist only a few years ago is now ubiquitous; Microsoft has bet its future on AI http://bit.ly/29orpKV and VC’s hardly see a plan today that doesn’t say AI somewhere in the pitch deck.
Nearly all AI acquisitions so far, however, have been ‘capability buys,’ meaning buyers acquire teams and some IP to augment their own efforts. Prices paid are usually $50m or less. Mind you some capability deals are significant; Apple reportedly paid $200m for Turi, and Twitter famously paid $150m for 15-person Magic Pony. Capability is so highly valued because demand far exceeds supply for good AI specialists, so much so that companies are resorting to high cost team poaching.
We think these deals are great for the eco-system. The majority involve companies we believe could never have scaled to any enduring size. Having recently interviewed dozens of international AI companies, we see quite a few run by technical, not commercial, leaders, and which rely on relatively low-value contracts with large early customers, many of which are rapidly buying orbuilding their AI capability in-house to transition away from such reliance.
However, we believe this is only the first wave of M&A in the space. Here is what we are seeing in the market today, and the pointers for the next 5-10 years:
- AI/ML is a segment today like ‘internet’ was a segment in 1995. Very shortly it will cease to exist stand-alone and instead broaden to become ubiquitous. Just like the internet, 10 years from now a company won’t be able to be in business without a quality data science team.
- The $5B of funding last year, while up 60% from 2015, is a fraction of what will be invested as AI goes mainstream. This capital won’t actually go to pure AI companies per se, most will go to what we call ‘platform companies’; product companies relying fundamentally on AI to accelerate disruption of broader markets
- Looked at this way we believe AI-based funding could increase 500-1000%, to $25-50B annually within the next 5-10 years. According to KPMG, total global VC funding was $127B last year, down from $140B in 2015. AI is so fundamental to the future of business we can easily see 20% or more of all VC money funding these AI-driven platform companies.
Which brings us to the acceleration in AI M&A values. We believe this will start to happen when larger ‘industrial’ companies acquire platform companies which leverage AI at their core to drive businesses with $ hundreds of million in revenue. These companies will be acquired for very high multiples of revenue and profits, meaning the buyer is paying for the core technology and the future opportunity more than performance or scale.
In fact this future has just arrived. A few weeks ago Intel announced it was acquiring Mobileye for $14.7B. Mobileye posted Q4 revenue of just over $100m, meaning annualised revenue of $400m. That prices the deal at a huge 35x run-rate revenue. A buyer might pay 3-6x revenue for a well performing business bought on performance. At 35x, this implies over $10B of cash, and the vast majority of total deal value, is being paid for Mobileye’s capability and future potential, which the company describes this way:
Our technology is one of the first embedded versions of AI, which means the technology does not live in the cloud but instead lives in the vehicle, on the chip.
As other segments such as the Internet of Things, financial processing and communications begin to leverage AI we believe companies as diverse as GE, First Data Corp and Vodafone will have to do their own version of Mobileye, potentially driving AI-driven M&A values to over $100B annually. That is 7 Mobileye transactions in a 12 month period, and there are literally hundreds of large-cap buyers whose businesses can be transformed by the right Ai-driven platform buy.