Erratum to: Eur. Phys. J. C (2012) 72:2138 DOI 10.1140/epjc/s10052-012-2138-3

In the original version of the paper, there was an ambiguity between the value of μ before and after the shift due to the Giudice–Masiero (GM) term. Here, we will clarify the equations which were affected. We define μ 0 as the μ-term in the superpotential defined at the input universality scale M in. μ(M in) will refer to the μ-term after the shift induced by the GM contribution to the Kähler potential also defined at the input scale. Then Eq. (12) becomes

$$\mu(M_{\mathrm{in}}) = \mu_0 + c_H m_0 . $$
(12)

Similarly, μB(M in) is defined as

$$\mu B (M_{\mathrm{in}}) = \mu_0 B_0 + 2 c_H m_0^2 , $$
(13)

which replaces Eq. (13). As a consequence, we would find

$$B(M_{\mathrm{in}}) = (A_0 - m_0) \mu_0/\mu(M_{\mathrm{in}}) + 2 c_H m_0^2/\mu(M_{\mathrm{in}}) . $$
(14)

This clarification affects the result only in Sect. 2 of the paper. For M in=M GUT, and when the Giudice-Masiero term (11) is included [15], one can deduce the (GUT) boundary conditions for μ and B:

(16)
(17)

This allows us to solve for c H where we obtain an equation similar to Eq. (30):

$$\begin{aligned}[b] c_H &= (B(M_{\mathrm{GUT}}) - A_0 + m_0)\mu(M_{\mathrm{GUT}})/(3m_0^2 - A_0 m_0) . \end{aligned} $$
(18)

These changes affect the contours in Figs. 2–4. In Fig. 2, with A 0=0, all contour labels should be multiplied by 2/3. In Fig. 3, with A 0=2.5m 0, all contours should be multiplied by 4.0. In Fig. 4a, with A 0=0, all contour labels should be multiplied by 2/3. Finally, in Fig. 4b, with A 0=2.0m 0, all contour labels should be multiplied by 2.0.

All results and figures in Sects. 3 and 4 remain unaffected.