I know this has been asked as a general question before and the consensus here is that each item that adds a resist magic percentage is multiplicative but UESP reckons it is additive with an 85% hard cap. With 1.3 out now have we any more confirmation which is correct?
Resist magic sources are additive, but are multiplicative with element resists. I have copied text below I wrote in another post:
Magick damage received calculation
Magic resist does reduce elemental damage, but separately from the resists.
( Damage | Absorbed ) = 0 * damage ( Damage | ¬Absorbed ) = (( 1 - magic resist ) * ( 1 - elemental resist )) * damage
Elemental Resists and Magic Resists cap at 85%. Magick absorption caps at 80% (without the breton daily power). Probability( Absorbed ) is just your magic absorption. Each effect is additive with itself up to the cap, then multiplicative with the other effects.
Expected damage = ( 1 - magic absorption ) * ( 1 - magic resist ) * ( 1 - elemental resist ) * damage
Min expected damage is therefore 0.2*0.15*0.15*damage = 0.45% of original
N.b. we can see, therefore, that it is better to stack a single source of resistance up till the cap than spread it evenly with the possible exception of absorption which will put you in the random hand of fate: Assuming equal chance of element types. However, this is not the case as fire and frost are much more common than shock. So it may be better to have single element fire and frost resistance, than a lower magnitude resist all.
- The UESP
EDIT: apologies for formatting. I am new to stack exchange and haven't worked out a clean post style yet.