A good answer might be:

  1
0001 0000     1610
1001 1000    -2410
---------    ----
1010 1000    -4010 (wrong!)

Patterns that Add Up to Zero

The "binary addition algorithm" does NOT work with sign-magnitude. There are algorithms that do work with sign-magnitude, and some early computers were built to use them. (And other computers were built using other not-ready-for-prime-time algorithms). It took several years of experience for computer science to decide that a better way had to be found.

There is a better way. Recall a question and its answer from several pages ago: 

  11111 111
   0000 0001  =   110
   1111 1111  =   ?
   ---------      -----
   0000 0000  =   010

QUESTION 14:

A number added to one results in a zero. What could that number be?