Quantifying the tension between the Higgs mass and (g-2)_mu in the CMSSM

Supersymmetry has been often invoqued as the new physics that might reconcile the experimental muon magnetic anomaly, a_mu, with the theoretical prediction (basing the computation of the hadronic contribution on e^+ e^- data). However, in the context of the CMSSM, the required supersymmetric contributions (which grow with decreasing supersymmetric masses) are in potential tension with a possibly large Higgs mass (which requires large stop masses). In the limit of very large m_h supersymmetry gets decoupled, and the CMSSM must show the same discrepancy as the SM with a_mu . But it is much less clear for which size of m_h does the tension start to be unbearable. In this paper, we quantify this tension with the help of Bayesian techniques. We find that for m_h > 125 GeV the maximum level of discrepancy given current data (~ 3.3 sigma) is already achieved. Requiring less than 3 sigma discrepancy, implies m_h < 120 GeV. For a larger Higgs mass we should give up either the CMSSM model or the computation of a_mu based on e^+ e^-; or accept living with such inconsistency.