1: \begin{abstract}

2:   We prove mean-square convergence of a novel numerical method, the tamed-splitting method, for a generalized Ait-Sahalia interest rate model. The method is based on a Lamperti transform, splitting and applying a tamed numerical method for the nonlinearity.

3: The main difficulty in the analysis is caused by

4: 	the non-globally Lipschitz drift coefficients of the model.

5:         We examine the existence, uniqueness of the solution and boundedness of moments for the transformed SDE. We then prove bounded moments and inverses moments for the numerical approximation.

6:         The tamed-splitting method is a hybrid method in the sense that a backstop method is invoked to prevent solutions from overshooting

7: zero and becoming negative.

8: We successfully recover the mean-square convergence rate of order one for the tamed-splitting method. 

9: In addition we prove that the probability of ever needing the backstop method to prevent a negative value can be made arbitrarily small.

10: In our numerical experiments 

11: %We use Backward Euler method (BEM) as a reference numerical 

12: %scheme to approximate the original equation.

13: we compare to other numerical methods in the literature for realistic parameter values.

14: %Numerical experimental results show that Tamed-Splitting method

15: % performs well.

16: 

17:  \keywords{Ait-Sahalia model,  tamed-splitting method, 

18: 		Lamperti transformation,

19: 		mean-square convergence rate \\

20: 		AMS subject classification: {\rm\small 60H35, 60H15, 65C30.}}

21: 	% \PACS{PACS code1 \and PACS code2 \and more}

22: 	% \subclass{MSC code1 \and MSC code2 \and more}

23: \end{abstract}

24: