Gold is a highly electronegative element and its compounds are covalent, so talking about electrons being lost to make ions doesn't really fit. Gold's electronegativity is connected to its outer s electron being tightly bound, rather than loosely as in Cs, partly because it is underlain by d orbitals which are poor shielders of nuclear charge. In fact one of gold's few ionic compounds is CsAu which contains Au-! Its electron affinity is close to that of iodine, and it has much in common with that element.
Mercury which follows gold also holds its outer s electrons tightly and has some noble-gas-like properties such as low atomisation energy and low reactivity. Mercury can be called a 'pseudo-noble gas' and as with the true noble gases ions with the same configuration are stable e.g. Au-, Tl+, Pb2+. It has even been predicted that mercury should have a volatile square-planar tetrafluoride fluoride HgF4 - like XeF4.
Gold resembles iodine in easily forming linear [AuCl2]-, square planar [AuCl4]- and also [AuF6]- ions, all of which are covalently bound. Chemically it is hardly metallic at all. I suppose AuF might be ionic but it decomposes at once to free Au and AuF3, a covalent chain polymer in which Au is surrounded by a square of F atoms.
So the oxidation states (not ionic charges except in Au-) of Au are -1 (CsAu etc), +1, linear covalent; +3 (square-planar covalent) and +5 (octahedral covalent). You may note that Au (III) is isoelectronic with Pt(II) which also forms square-planar complexes, and Au(V) is isoelectronic with Pt(IV) which forms countless octahedral complexes.