Mca 6 2007 def pdf
Download 192.28 Kb. Pdf ko'rish
|
MCA200706121
Could you translate this line of thinking
to management? I wrote about what I call stochastic tinkering in the Black Swan and elsewhere. It is exactly the same concept. I discovered that effectively returns in mo- dern environments are much more linear to the number of investments than to the quantities in- vested. In other words, it is a lot better to do a lot of small speculative investments than a few large di- rected ones. Why? For a fundamental reason, and this is what I call the envelope of serendipity. In The Black Swan I look at the history of technolo- gy, in two fields, engineering and medicine. It be- comes clear firstly that most technology is unpre- dictable, whether in the invention or the implementation. Only after the fact do we think that there is a pattern, a design. Effectively, our predictive powers are horrendous. We just do not know what will be the next breakthrough. Our forecasts for five years are going to be largely ineffectual. But again it is much better to invest in a lot of small things, for the very simple reason that the small things and the big things tend to have the same payoff. MCA: oktober 2007, nummer 6 14 Geoff Delderfield: beeld MCA: oktober 2007, nummer 6 15 In my book I took the most influential technologies of the last 50 years. The top three are the laser (the basis of the digital revolution), the computer, and the internet. Now, the internet was not designed for e-mail or as a place for teenagers to hang around in chat rooms. The computer was designed as a cal- culating aid, not as a word processor, which turned it into a useful machine, or as a communication tool. The laser was designed by somebody who wan- ted to show off by splitting light beams! For years we did not even know what to do with it. We never could imagine that it would lead to eye surgery, you see. What is your message to business managers? Firstly, if you are shooting for returns, try to get as high an exposure to serendipity as possible. Once you find something that looks like it might work, invest in it for the max. Like with the 100 books: if one starts selling, go for it and do everything you can to get more sales. The same happened with my book: at first nobody took much notice, but when it hit the NYT bestseller list the whole com- pany, Random House, began to get involved. At a higher aggregate: look at Bertelsmann. They own so many publishing houses, they publish thou- sands of books. They really took my idea to the max. How about budgeting where one tries to grasp the future? I was talking about statistics, not budgets. Budgets are a necessity. Of course a flexible budget is better than a fixed one. I have not looked at budgeting as such, at the risk of budgeting. But I have looked at projections of future cash flows. People who do cash flow analyses do cash flow projections, without in- cluding an error rate of past cash flow predictions. If you were to do this, and indeed included past error rates, you can not imagine how shocked you will be. The best example is the billiard ball. What hap- pens to the trajectory of a billiard ball? The first bounce is predictable. At the ninth bounce you have to accumulate the computational powers of every single person in the room. At the 53d bounce you must include every single particle in the universe. The error rate multiplicate. Forecasts of oil prices are another example. The US government produced a fo- recast in January 2004, of 27 dollar per barrel for the next 25 years. Six months later they doubled the fo- recast. And still they did not realize that when you have to double your forecast in 6 months there is so- mething seriously wrong with your forecast as such. I came up with a very simple rule of thumb: a 25 year forecast has in the best of circumstances a bil- lion times bigger error rate than a one-year fore- cast. The same problem affects climate predictions. We may be very good at predicting the weather on Tuesday. Even in Holland. But predicting climate change. …Come on! This is how I talk about risk. Domain separa- tion, based on sensitivity to fat tails. And I have been doing this for twenty years, even as a trader I did. Download 192.28 Kb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling