Without having studied the subject at all, I believe that little 'duck tail' appendage adds considerable positive buoyancy at the tail end when compared to a 'cruiser stern' ( e,g,. QE2). Would welcome some other thoughts on this.
When a ship accelerates up to speed, it develops a trim by the stern known as 'squat'. When the ship trims by the stern, it's 'angle of attack' changes as the bow lifts, and this adds to the hull drag and fuel consumption (as stated by Magic Pipe), and also the aft draft (the depth of lowest aft part of the hull immersed in water). This 'squat' resulted in the QE2 running aground and incurring considerable damage - I seem to recall being told that she would squat by some 3+ metres when doing 30 knots, and, when this was added to the 'suction' effect experienced in shallow waters, the squat value would be even greater, hence the grounding.
By adding the 'duck tail', when the ship tries to 'squat' at speed, the additional buoyancy provided by the duck tail when it becomes immersed would act to counteract the amount of squat, and thus the aft draft increase is reduced. The effect of the wake pushing against the underside of the ducktail would also add dynamic lift as well.
Between the 'duck tail' at the stern and the bulbous bow at the other end, the two appendages combined would also reduce the ship's pitching moment.
Some ro/ro ferries have 'sponsons' added to their sides to give additional buoyancy - effectively giving the same result as the 'duck tail', but in this case to assist the heeling tanks in trying to reduce the list when rakes of heavy wagons are rolled on and off the ship from the outermost rail lines.
I was born in Christchurch, NZ. My relatives down there tell me they have been doing a lot of 'involuntary' squatting recently.
Cheers
Skilly