In a hypothetical world:-
* I bought a thousand $1 lottery tickets and won $1
* My friend bought one and made $1000
* He made a return about a million times mine
So I've come to the conclusion that next week my friend will do about a million times better than I did.
So I'll invest $1 this time, he will invest $1000 - which I will give him - and we will split the million dollar winnings - cool.
Wrong.
There is no relation between my lottery winnings and my friends.
The problem of this argument is of course that post event analysis of a singular event through a small sample will tell you very little about the next event.
There is no linked process of $1 in, returns $0.90 every time - it's probabilistic.
So with this in mind, I read Booz Allen Hamiliton's report on "Money isn't Everything", something which I've been promising to do for ages.
They took 1000 companies and looked at R&D spending in 2004 and related this to profitability in 2004. They concluded that there is no relation (see above lottery example).
That higher R&D spending translates into competitive advantage is a myth - sure post analysis, but my chances of winning the lottery are directly related to how much I invest pre-event. Doesn't mean I will win or won't, or that my friend will or won't do better than me.
They concluded that spending too little will hurt (i.e. not competing in the lottery means I won't win it and if my livelihood depends upon competing in the lottery - then I'd better compete).
They also note that average R&D spending varies between sectors. The two biggest being internet and health (I assume bio-tech), the smallest being consumer and energy - in otherwise markets which tend towards utility or commodity basis tend to spend less on R&D, though this is not stated.
Certainly some markets are less of a lottery and have become stable utility or commodity environments. Technological innovations in these markets may in fact depress margins (as per the Buffet example and loom machinery).
Their argument goes that innovation is seen as a "black box" - a process presumed to transform R&D spending into results without anyone fully understanding how - sounds like a lottery to me. I agree however that this is not an excuse for reckless funding.
“It’s absolutely a myth that money alone will solve vexing technical problems. Rather, reckless funding largesse is actually a barrier to transformative innovation as it turns scientists into constituents with an incentive to maintain the status quo. Reasonable constraints are a spur to progress." Dr. Allan O. Steinhardt.
Welcome to the world of chaotic processes as opposed to defined. I'm not against managing investment, quite the opposite. It is important to understand though that you can't plan and map out research - unless you know what you are about to invent prior to undertaking the work. In the later case you're not innovating you're implementing (a defined process).
They do note that if you look at a wide sample or group of companies - then the top 500 R&D spenders per industry section outperformed on a gross margin level the bottom 500 R&D spenders and this is particularly pronounced in "new" markets i.e. those who spend more on lottery tickets do better on average than those who don't.
Overall the conclusions of the report seemed to be
* If you don't spend enough, you suffer.
* If you spend too much, you suffer.
* No-one knows what enough is.
* How much isn't as important as on what
* Managing your spending is a good idea.
So in conclusion, spend just about the right amount of money on all the right projects and you will succeed.
In otherwords, when you're next at the store - just spend $1 only and buy the winning lottery ticket. Try not to waste cash on cigarettes and beer.
Easy.