Alt codes and unicode characters are only vaguely related, and (for the most part) it's up to each application to determine how to deal with them, so it can be different between apps, versions of apps and such.
Essentially it's about character encoding and Code Pages, rather than fonts or user input.
Alt codes for ASCII are only 0-255 (8-bit code page). Many programs (and OS's) deal with larger numbers inputted by the user by wrapping at 256.
256 = 0, 257 = 1, etc.
So entering ALT+8230, is actually equivalent to typing ALT+38 (8230 / 256 = 32.1484375. 256 * 32 = 8192. 8230 - 8192 = 38), which is the ASCII character "&" -- unless the program intercepts it and replaces it with 'proper' Unicode codes, or by using a different code page (ie: Unicode), as MS Word, and newer WordPad versions do, for example.
Windows uses an extended version of the MSDOS 8-bit Codepage (850) called the "Windows-1252" code page (for English languages anyway). But it only uses this for legacy parts of the OS that don't fully support the Unicode code pages.
These are the reasons why you can easily see different results in different OS's and applications, as well as within different parts of Windows itself.
It's a confusing issue to say the least. :)
I'm pretty sure Alt-0133 is the "official" Windows extended ALT code for the ellipsis (in the Windows code page), so you'll probably have more consistent luck using it.
Here's some more related info.