Context : Observations suggest that the major fraction of long duration gamma-ray bursts ( GRBs ) are connected with broad-lines supernovae Ib/c , ( SNe-Ibc ) .
The presence of GRB-SNe is revealed by rebrightenings emerging from the optical GRB afterglow 10 – 15 days , in the rest-frame of the source , after the prompt GRB emission .

Aims : Fermi -GBM has a field of view ( FoV ) which is about 6.5 times larger than the FoV of Swift , therefore we expect that a number of GRB-SN connections have been missed due to lack of optical and X-ray instruments on board of Fermi , which are essential to reveal SNe associated with GRBs .
This fact has motivated our search in the Fermi catalogue for possible GRB-SN events .

Methods : The search for possible GRB-SN associations follows two requirements : ( 1 ) SN should fall inside the Fermi -GBM error box of the considered long GRB , and ( 2 ) this GRB should occur within 20 days before the SN event .

Results : We have found 5 cases , within z < 0.2 fulfilling the above reported requirements .
One of them , GRB 130702A-SN 2013dx , was already known as GRB-SN association .
We have analyzed the remaining 4 cases and we concluded that three of them are , very likely , just random coincidences , due to the Fermi -GBM large error box associated with each GRB detection .
We found one GRB possibly associated with a SN 1998bw-like , GRB 120121B/SN 2012ba .

Conclusions : The very low redshift of GRB 120121B/SN 2012ba ( z = 0.017 ) implies a low isotropic energy of this burst ( E _ { iso } = 1.39 \times 10 ^ { 48 } ) erg .
We then compute the rate of Fermi low-luminosity GRBs connected with SNe to be \rho _ { 0 ,b } \leq 770 Gpc ^ { -3 } yr ^ { -1 } .
We estimate that Fermi -GBM could detect 1 – 4 GRBs-SNe within z \leq 0.2 in the next 4 years .