In their recent article (J. Econ. Theory 33 (1984), 356-367), J. Z. Drabicki and A. Takayama suggest that the additional flexibility afforded by the introduction of a third asset in Tobin's two assets model leads to the reversal of saddle-point instability into local stability. The present comment questions that suggestion and discusses the seemingly objectionable derivation of this result from Samuelson's correspondence principle. It shows that, when the correspondence principle is correctly applied, the stability result is only ensured by parameter values which can be interpreted as indicating a sufficient sluggishness in stock and price adjustments. Friction, not choice, is the stabilizing factor.