This is easily the most significant factor. At privatisation it was realised that separate ownership of trains with TOCs leasing them was the most effective way of allowing franchises to change hands while keeping trains in place to allow operations to continue more or less unchanged. The model was seen to be that which exists in aviation but failed to recognise that part of that model includes lessors having a variable number of spares not being used by anyone, sometimes for lengthy periods. With only three ROSCOs to begin with there was never any real competition in the train leasing market leading to inflated rates. In fact the government willingly created the situation because it wasn't sure that anyone would be interested otherwise!
The fact the new Greater Anglia franchise will lead to a complete fleet renewal and subsequent glut of trains with no homes may well be a warning shot to the ROSCOs to move on from the relatively easy profit they have enjoyed up to now.
Against that, there are more leasing agents these days, and several ways of funding new trains.
Manufacturers are much more involved in funding their own production and maintenance.
Also the (public procurement of) IEP, Crossrail and Scottish 385 deals, all by-passed the main ROSCOs.
So there is more competition in the leasing market.
Despite that, the big recent (private) orders have gone to the principal ROSCOs (eg the CAF and Stadler deals, or the First AT300s).
So the ROSCOs can't be all that uncompetitive.
The DfT business model was until recently to authorise the minimum new stock and "work it harder", eg 2 new vehicles for 3 old ones.
Now we have a growth model which for the first time says "fill more new trains and pay us the increased premiums".
