Dealing with autocorrelation.

#1
Every method I came accross so far uses some form of first-differencing, which kills my results.

Is there any other way to deal with autocorrelation?