In addition to cam-band location, the bungee has to be the right length. If too loose, it'll always allow the valve to drop - creating a pivot action where the tail rises. The cam-band location should be sufficiently high on the cylinder to create some lateral tension on the tank - pulling the valve down below the armpit. At the same time, it has to be low enough on the cylinder that the bolt-snap is not equal to, or higher than, the fulcrum of the cylinder. If it is, there's no way to control tail rise. The exact location is always person-specific as it varies with torso height/body proportion.
Also look at your cam-band bolt-snaps and the length of line which attach them to the cylinder. Some companies sell huge snaps (i.e. Hollis), which may be necessary for cold-water/thick gloves, but are too long to allow realistic cylinder trimming. If used as meant, in cold water, it's not an issue - as you'd probably be in negative steel tanks. Either get smaller (medium) bolt-snaps and/or shorten the attaching line. The total length of snap and line used there represents the distance variation that your cylinder will transition through when moving from negative to positive. The longer, the more difficulty getting trim sorted. You do need a bit of length though - especially if used for overhead environments/confined areas. Big bolt-snaps are the typical obstruction - and these are often sold as default if included with rigs.
Beyond that, it's just a case of properly positioning D-rings on the harness so that you can move the bolt-snap lower to counter-act progressively positive cylinder buoyancy as the dive progresses. Low-profile (tropical) D-rings also help... as they reduce the distance the cylinder can travel away from the diver positive/negative.
Some rigs, such as the SMS50 make that very hard... because the 'rails' sit lower on the torso than the waist belt. When you move from the rails to the waist D-ring, then you're actually reducing the spacing. That screws up all your adjustment on the cam-band. Ideally, you'd have to re-position the cam-band at the same time to resolve the variance. Obviously, that's a bit impractical... You suffer for the bad design... it's a rig based upon cold-water diving factors and cylinders, but primarily used in tropical locations (buoyancy insufficient for many cold-water divers) where ali tanks are the norm. Hollis should address that failure..