As per my understanding, you have not modelled the body diode and the MOSFET is still conducting even if you apply positive potential at source and negative potential at drain.
This is the expected behaviour for any "practical" MOSFET.
MOSFET can block the voltage in only one direction and it can conduct current in both the directions, owing to its construction
Let me explain....
You can see that the P-type body is shorted with the source and a P-N junction is formed between Drain and source. Therefore, the "body diode" is inherent, owing to its construction.
Having selected : "Model Body Diode = No", will not remove that body diode. Instead, when you do not model body diode, Simulink will use the default model for it. When you choose "Model Body Diode = yes", simulink will allow you to enter your desired diode characteristics to be modelled.
Practically, you can never find 3-terminal MOSFET without body diode. Therefore, you are witnessing drain currents even with negative Vds
Consider setting the "Number of terminals" to "four.
When you set the "Number of terminals" properties to "four" as shown below,
You can see that there is change in construction. The P-type body is not shorted with the source, instead, a seperate terminal is connected to the P-type body.
The Simulink block for "N-channel MOSFET" will now look like:
Now, if you provide positive potential between "g" terminal and "b" terminal, the channel will be formed, which can allow the conduction in both the directions, from Drain to source and vice-versa. But when
is lower than the threshold voltage, the channel will not be formed and the device will be capable of blocking voltage of either polarity accross drain and source terminals. Therefore, you can see that in this case, the MOSFET can allow flow of current in both the directions and it can block the voltage of either polarity.
In both the cases, MOSFET allows the current to conduct in both the directions
I hope you find this explaination usefull !