1: \begin{abstract}

2: We demonstrate that the processes underlying on-line auction

3: price bids and many other longitudinal data can be represented by an

4: empirical first order stochastic ordinary differential equation with

5: time-varying coefficients and a smooth drift process. This equation

6: may be empirically obtained from longitudinal observations for a

7: sample of subjects and does not presuppose specific knowledge of the

8: underlying processes. For the nonparametric estimation of the

9: components of the differential equation, it suffices to have

10: available sparsely observed longitudinal measurements which may be

11: noisy and are generated by underlying smooth random trajectories

12: for each subject or experimental unit in the sample. The drift

13: process that drives the equation determines how closely individual

14: process trajectories follow a deterministic approximation of the

15: differential equation. We provide estimates for trajectories and

16: especially the variance function of the drift process. At each

17: fixed time point, the proposed empirical dynamic model implies a

18: decomposition of the derivative of the process underlying the

19: longitudinal data into a component explained by a linear component

20: determined by a varying coefficient function dynamic equation and an

21: orthogonal complement that corresponds to the drift process. An

22: enhanced perturbation result enables us to obtain improved

23: asymptotic convergence rates for eigenfunction derivative estimation

24: and consistency for the varying coefficient function and the

25: components of the drift process. We illustrate the differential

26: equation with an application to the dynamics of on-line auction

27: data.\looseness=-1

28: \end{abstract}