>In nature stability

or at least in somewhat understandable models beyond simple linear ones:

http://mathworld.wolfram.com/StrangeAttractor.html