data x;                                                              
    retain _seed_ 0;                                                          
    do _i_ = 1 to 1000;                                                        
       Normal1 = 1 + 0.2 * rannor(_seed_);                                   
       output;                                                                
       end;                                                                   
    drop _seed_ _i_;                                                          
 run; 

学习了，多谢楼上的指导~！

Thanks a lot, but what I need is the realization from geometric random walk, not from normal ditribution.

。。。。。。。。。。。。。。。。。。

非常感谢啊。。。。。。。。。。。。。