It's possible that there is too much paste.  In addition, if you have a lead free component and a tin/lead solder, you might need a slightly higher temperature than a standard tin/lead process to get the solder to flow on the lead free component.  However, it does appear that there is solder mask between the pads AND that the pads appear to be defined by the solder mask.  Typically, when the solder melts, the component will settle down and spread the solder, with the solder flowing under the component with the flux.  Then as the flux burns off the solder coalesces and pulls back to the pads.  This is how solder beads are formed with conventional capacitors.  With mask defined pads, it is more difficult for the solder to coalesce so it cannot pull back to the pads and is left under this big component.  It cannot flow out to the edge to form a bead.

It's also possible that when the component is placed, the pressure is spreading the solder.

I suggest that you lift some components right after component placement to see if pressure is spreading the solder.  You can design the pad with a small solder mask exclusion area along the inside edge of the pad.  Finally, enlarge the pad area along the other three sides without enlarging the solder stencil.  These two actions will give the solder someplace to go if there is too much solder or if it needs someplace to flow temporarily.

I hope that this helps.