Assuming no irregularities in the performance of the reflow oven, the reason that immediately comes to mind is that the solder paste was not properly dried before being subjected to the elevated temperature of reflow.
There are volatile ingredients in the non metallic portion of the paste - like alcohol -and if it is still "wet" when it reaches a zone at about 100Cit rapidly boils off generating copious amounts of gas -like a mini explosion - which could easily displace a component.
The presence of humidity in the paste or the board would have the same effect. I would expect that if this were the cause the same phenomenon would be noticed on many of the small components.
However, It is difficult to believe that the reflow oven's thermal profile was not set up with a suitable ramp up to cause the solder paste to dry out correctly. It would be worth while checking the reflow oven profile to make sure it is correct. Baking of the boards prior to pasting and assembly should resolve any board humidity problems.
Another possibility, that may be more likely, is that the assembly was allowed to stand at ambient temperature after component placement and before being passed into reflow.
If this is the case, it is not uncommon for the surface of the solder paste to dry out creating a skin which encapsulates the volatile material and prevents it evaporating during the slow ramp up. Here again, when it arrives at a high enough temperature it explodes with the same result!
