Kaldorian assumptions and endogenous fluctuations: a note on Schinasi’s IS–LM model