It does not reduce drag, but creates more of it.
This concept is called a "box wing" and is one variety of closed wings. Another is called a "joined wing" and was popularized by Julian Wolkovitch. However, only people who don't understand how a wing produces lift can be enticed by this concept - it has too many drawbacks for the modest reduction in induced drag it offers. The general idea is to sweep the horizontal surface of a T-tail forward and stretch it out until it joins the backward-swept wing and relieves it in bending, similar to a strut.
Similar to winglets, the vertical extension of the wing structure will involve more air into the production of lift, which reduces induced drag, but this extended structure must be bought with a hefty mass increase which will reverse the gain in induced drag, because much more lift needs to be produced in order to lift the additional structure.
Note especially that the rear wing (which is supposed to act as a tail surface) will be loaded in compression in order to support the wing. Making this rear wing narrow will make it susceptible to buckling, which can only be avoided by a stiff and heavy structure, and will increase its friction drag due to its much lower Reynolds number. When a winglet-like element is added between wing and horizontal tail, this winglet is now loaded in bending and will also need to be made much stronger and heavier than a conventional winglet.
Calling this a "Prandtl plane" is to completely misunderstand the 1924 publication of Ludwig Prandtl, who very well understood how wings create lift, about the benefit of box wings.
