I have read some conflicting information from two reliable sources regarding the scaling exponent in multifractal systems - tau.
On the Yale website devoted to fractals, they say "Tau is a decreasing function of q and is concave up."
http://classes.yale.edu/fractals/MultiFractals/IFSMF/IFSMF.html
But in Calvet and Fisher's (who themselves worked with Mandelbrot) paper about multifractals in finance, they say that tau(q) is concave down and increasing (they even show a picture) on p.395
I find this confusing, can anybody explain?
