Not sure what high voltage was supposed to do with it... magnetic fields correspond to currents, besides, spanners have low resistance, how would you expect to drop a high voltage across it other than by arcing to it? And we can find the results of that in everyday life: normal metals aren't substantially magnetized due to welding.
Useful magnets are formulated specifically for retaining strong fields (high coercivity and remenance). Most steels are poor on both of these. (Physically) hard steels tend to be also magnetically hard, so you will find nuisance magnetization on screwdrivers and pliers and tweezers and such (and yes, this even includes stainless steels -- the hard ones tend to be magnetic, and tend to retain a charge). But not really anything useful.
Ye Olde classic iron horseshoe magnets need to be handled carefully, as merely increasing the magnetic path length may be sufficient to cause loss of magnetization -- that is, they're normally stored with a "keeper" plate across the face, which keeps the internal flux density high. Removing it, increases loop reluctance, reducing flux density, apparently enough to cross the opposite coercivity limit and lose magnetization. (Plotting B vs. H, a hysteresis loop is seen, H (horizontal axis) being applied magnetization (amps/meter), with the loop width being coercivity, and B being flux density (tesla, V.s/m^2), the loop height being remenance. Usually one just sees plots of full +/- scale magnetization, but hysteresis applies for smaller loops as well -- it's not that failing to use the keeper causes complete demagnetization I think, it's just that it gets walked into a smaller corner of the loop.)
Alnico magnets had this issue as well ("barrel" style speaker magnets, etc.). Alnico has quite high remenance (Bsat ~ 1.2T like most steels, or is it more like 1.5? and remenance not much below there), but not great coercivity. So it works great when placed in a fixture, but not so great when removed, all by itself.
NdFeB (and to a lesser extent, SmCo) has even higher Brem, as well as all the coercivity -- it's a very permanent magnet, requiring very high currents to charge (100s kA/m). Accordingly, very little of it need be used -- speakers using it can be much smaller. Motors can even hide thin strips of it within rotors, making some really interesting (highly efficient) synchronous machines.
Tim