Recent studies confirm the economical feasibility of the application of superconductors in AC power apparatus. The low-loss conductor which has been proposed still only exists in theory. Simple calculations show that the economic criterion described by Ogasawara and Hlásnik is inaccurate because of an oversimplified hysteresis loss formula. A necessary modification shows that stricter demands will have to be made. As production of ultrafine filaments is a big technological problem, final diameter reduction by rolling instead of drawing is proposed. The expected better performance of a rolled composite has been verified experimentally in test coils as well as in transformer geometry. Experimental set-up and results are presented.