Actually the fact that wages stopped rising along with the economy as a whole (e.g. less of the generated profits get payed out as wages) seems to play an important role, too.
The retirees represent a cost in absolute numbers (e.g. they are entitled to pension benefits of a certain size). The workforce is collectivly paying out the retirees with a share of their income.
That system works fine even with a declining retiree/workforce ratio (which actually shrinks since at least the 1960s) as long as real wages continue to grow along with GDP and productivity.
Which they don't, and that's a problem.
[0] Figures released by a research institute of the federal government. http://doku.iab.de/aktuell/2015/aktueller_bericht_1514.pdf
I suppose a good outcome would see the unemployment rate of refugees converge against the unemployment rate of the non-migrant population (vaguely: 7.9%[0]), or more specifically the unemployment rate of those parts of the non-migrant population with similar levels of qualification (vaguely: 13.3%). Because as these levels converge, it makes less and less sense to think of it as a migration issue and more of a more general labor market issue.
[0] Numbers strictly to get a feel for the broad range, the first is the overall unemployment rate in 2011, the second the unemployment rate of people with basic education; unemployment numbers are a rats nest in the best of cases. http://www.bpb.de/nachschlagen/zahlen-und-fakten/soziale-sit...
That's only the statutory part though. You are advised to contribute to an occupational and/or private arrangement alongside the statutory scheme. They operate differently.
The underlying problem Germany has is the population dynamics. The population is an ageing one. Lots of people are living longer and not enough children are being born to cover the gap. It might be politically unpopular, but importing refugees and successfully converting them to integrated German tax payers, might well be a long term goal of the state in order to stave off the pensions crisis.