Solving Old Puzzles with New Tricks: Addressing Endogeneity and Nonlinearity in Time-on-Market Research