This is a guest post. The views expressed here are solely those of the author and do not represent positions of IEEE Spectrum or the IEEE.
Two years ago, New York Times columnist Paul Krugman wrote a much-discussed piece on the discrepancy between corporate profits and labor compensation. The column sparked a huge debate, and on the Times website alone readers left more than 1,300 comments. The part that interested me, however, wasn't the discussion on capital versus labor. What caught my attention was a comment Krugman made about robotics. He referred to it as a capital-intensive technology. It may not sound like a big deal. The problem is, it isn't true.
The claim that robotics is capital intensive is a myth that needs to be debunked. This isn't just an academic curiosity. Capital intensive means that a lot of money is required to make money, usually because you have to own a lot of assets-cash, buildings, equipment-to be in the industry. Examples of capital-intensive industries are utilities, railroads, and telecom. Venture capitalists steer away from such industries, seeking instead emerging markets that promise rapid growth-and potentially handsome profits. In fact, VCs often allude to "capital intensity" as a reason for not investing in robotics. And some observers have used the same argument to explain why robotics is growing so slowly and commercial success is elusive. They seem to be making the case that robotics is somehow fundamentally different from other high-tech industries like web software, e-commerce, or biotechnology.
To make things worse, Krugman is not alone in his claims about robotics. Other influential people and organizations have referred to robotics in the same way, with the myth being repeated by business schools, venture capitalists, startup incubators, robotics associations, and others. The troubling thing is that these are entities that influence the creation of robot businesses and the future of robotics. So we need to get this right. It is vital for the robotics community to know that successful robot companies are not capital intensive: they are actually capital efficient.
Bear with me. This is an important distinction. Capital efficient means a relatively small amount of assets is needed to make money. Which is to say, the robotics industry has no excuses not to attract more investments and grow at a faster clip. The technology is there and cases of success show that robotics companies can have the characteristics that venture capitalist seek. Indeed, many opportunities could be describe as a VC's dream. If we are not commercializing, it is because we are doing something wrong.
Following a suggestion from Michael Ewens, a professor of finance and entrepreneurship at CMU's Tepper School of Business, and under the guidance of Richard Green, a professor of financial economics at Tepper, I decided to examine this issue in detail. Our idea was to compare robotics to another industry-one that is known to be capital efficient enough for VC investment. We would compare them based on some key criteria and, if there are similarities, this would be an indication that robotics behaves in a similar way. For this reference industry we chose biotechnology. It is not quite as capital efficient as software, but it is widely accepted by investors and venture capitalists as having a capital efficient "hits" model similar to software. By a "hits" model I mean that a few companies are expected to get really big and really profitable. So profitable, in fact, that they pay back the majority of investments that fail or underperform. This model comes from the movie industry and the oil exploration business, but has been applied to technology investment.
In our analysis we used financial data from the top 10 public robotics companies over the last economic cycle. Among these were Intuitive Surgical, iRobot, AeroVironment, Adept Technology, Cognex, and the recently acquired MAKO Surgical. (We didn't include private companies because we wanted to focus on relatively large and well established companies.) The criteria we examined from these companies focused on three things: the kinds of assets listed in the company's balance sheet; how much revenue a company generates from its assets; and margin of profits on that revenue.
And what did our analysis show? Drum roll, please . . . It turns out robotics companies clear the biotech reference mark in style! In all three criteria robotics performs almost as well, or better, than biotechnology. In fact, consider for a moment that Krugman and others who said robotics is capital intensive were right. If that was the case-if robotics were an inherently capital intensive industry-then we would see three things in their public financial data: 1. A balance sheet full of long term assets and long-term liabilities; 2. Very little revenue created by the capital employed in the company; 3.Low returns (profits) from assets the company uses.
But we don't see any of that! Our examination shows that the opposite is true. What we see is: 1. A balance sheet with a large proportion of short term investments and cash; 2. Significant revenue generated from assets; and 3. Sizable profits, especially in comparison to the amount of assets the companies have. The three graphs below are rather technical, but they illustrate well our findings. So let's go into some more details for each criteria.
1. The Balance Sheet: Capital intensive industries typically acquire a lot of debt and hold many long term assets. Robotics companies do the opposite, even in comparison to biotech companies. From a pure balance sheet perspective, it looks like robotics companies are the nimble risk-takers with little in the way of cumbersome long term assets and more cash than debt.
2. Revenue: Robotics companies are more efficient at generating revenue from their assets. On average, robotics companies generate twice as much revenue, per dollar of book assets, as biotechnology companies. Moreover, their assets tend to have a greater portion of current assets like receivables and inventory as opposed to equipment and factories.
3. Returns (Profits): Although robotics companies generate more revenue than biotech companies, they do not get the same margin on that revenue. Our research found the amount of profit to be marginally higher for a biotech firm, but certainly not of the magnitude that would make robotics unattractive for investors. The upshot is both industries appear to be capable of producing excellent and extremely similar returns, so we would call this one a tie.
So what do our findings mean? Our analysis shows that robotics can generate dramatically more revenue from a given set of assets than biotechnology, but that it falls a bit short in delivering this revenue as net income. Given that the biggest expense for robotics firms is always personnel, it would appear that rather than having a capital intensity problem, if anything, robotics has a labor intensity problem. This might be a clue to where further research should focus.
Still, everywhere we look we see that robotics exhibits the same or better traits in regards to capital intensity as the biotech industry. So complaints and excuses that investing in robotics costs too much and returns take too long simply don't add up. It's time to get back to the business of building this amazing industry-and stop perpetuating the myth that capital is the problem. I'm looking at you, Krugman.
