Tuesday, February 21, 2006

How can the private sector fund more long-term R&D?

The question above uses the word "more" for a reason. Since the 60s, US R&D spending as a % of GDP has stayed between 2-3%, but with one twist - In the 60s, the government spent all of that, now the government spends slight more than 0.5% with the rest of the tab picked by the private sector. A recent Time magazine article laments this trend, and the apparently consequent slowing rate of US patent filings and publications in engineering and science journals. So, the private sector has picked much of the slack, but can it do more?

The problem, of course, is that no one knows if and when R&D will payoff (in most cases). Today, Google is a massive R&D engine and out-innovates most in computer science, but how much of that has been monetized? Turns out targeted advertising and Adsense comprise 90% of Google's revenues, so some great recent innovations in machine learning, operating systems and algorithms may have little or no marginal revenue impact.

A simple insight might suggest one way to get more funding into R&D. The payoff from R&D looks exactly like a call option - essentially, the same insight that has driven recent work in real options. There is much criticism of real options, something I will not delve into as it's less relevant to the question at hand. Consider a project funded entirely by writing options on R&D.

This could be a project in academia, or an entire firm. Let such R&D options be tradeable and liquid in a market-place, and let a new breed of analysts emerge that can value such options. Along with new standards around the disclosure of the progress of R&D projects, such analysts are critical to ensure liquidity in the market, and hence the viability of R&D options.

Assuming R&D options mimic European calls, an interesting corollary is that the price of the option increases with the risk of the underlying asset. This provides a perverse incentive to fund projects that are more risky, exactly the kind of projects that find funding hard in the current system.

As a global leader in science and engineering, the US might well consider investigating this route to draw more funds into risky R&D. Thoughts?


Blogger Aanand said...

Hi Mukul,

I've become a regular reader of your blog now.

As far as private R&D is concerned, the riskiest (in terms of probability of productization) activities are usually done by the "labs" arms of companies - like HP labs, Microsoft research etc. OR by start-ups. The question is then, how does the average investor directly invest in start-ups pre-IPO?

The precedent for funding high risk activities already exists in the movie industry where hedge funds invest in an assortment of future movie releases from say, Disney or Paramount. Interestingly enough, there's even an online game called the Hollywood Stock Exchange where you can "buy" call or put options for movies.

1:27 PM  
Blogger Mukul Chawla said...


Thanks for taking the time to look up my blog.

I think the terms R and D need to be separated. Lets define R (research) as the pursuit of solutions to problems (without specifying a motivation). D (development) is productizing innovation, almost always motivated by economic profit.

In that context, I think D is at much less risk than R. D is well rewarded in the corporate world, and is therefore less of an issue.

I claim that in recent times (based on anecdotal evidence), R has all but vanished from industry. There was a time when freewheeling research was desirable at places like IBM's Zurich Lab and Bell Labs, increasingly, these places have seen their budgets cut and their orientation turned towards D. From friends who work there, I hear that Microsoft Research (MSR) is mostly D, little R.

So, I think what you state may be true for D, but the real issue is who will pick the tab for R?

3:47 PM  
Blogger Ramsu said...


Fascinating post. Concurs with something I've been thinking about as well.

Spent a few minutes thinking about how this would translate into numbers in a real scenario, and something occured to me: given that we're discussing longer term R&D projects, the pieces of the puzzle in such a project would each take some time to assemble. Which means that funding will have to be decided/allocated at discrete time intervals. Also, the bigger the project (in terms of scope and research potential), the tougher it is to get each of these pieces in place, hence the greater the risk in allocating funding.

So here's my question: Do you think this aspect of discrete time intervals would affect the way a real option on R&D is valued? If so, how?

Admittedly, I have not spent much time reading up on real options, so the answer to this question may be obvious or already well-known. Just thought I'd raise the question anyway.

11:44 AM  
Blogger amar rama said...

Hi Mukul,
Your post on R&D reminded me of an article I read recently in Inc. magazine that questions if R&D should be called an "investment". Short read..
article here

11:48 AM  

Post a Comment

<< Home