Two-loop electroweak top corrections: are they under control?

Abstract

The assumption that two-loop top corrections are well approximated by the $O(G_mu^2 mt^4)$ contribution is investigated. It is shown that in the case of the ratio neutral-to-charged current amplitudes at zero momentum transfer the $O(G_mu^2 mt^2 M_Z^2)$ terms are numerically comparable to the $m_t^4$ contribution for realistic values of the top mass. An estimate of the theoretical error due to unknown two-loop top effect is presented for a few observables of LEP interest.