1: \begin{abstract}

2: We study different estimators of the radius of convergence of 

3: the Taylor series of the pressure in finite density QCD. 

4: We adopt the approach in which the radius of convergence is estimated

5: first in a finite volume, and the infinite-volume limit is taken later. 

6: This requires an estimator for the radius of convergence that is

7: reliable in a finite volume. Based on general arguments about the

8: analytic structure of the partition function in a finite volume, we

9: demonstrate that the ratio estimator cannot work in this approach, and

10: propose three new 

11: estimators, capable of extracting reliably the radius of convergence,

12: which coincides with the distance from the origin of the closest

13: Lee-Yang zero. We also provide an estimator for the phase of the

14: closest Lee-Yang zero, necessary to assess whether the leading

15: singularity is a true critical point. We demonstrate the usage of

16: these estimators on a toy model, namely 4 flavors of unimproved

17: staggered fermions on a small $6^3 \times 4$ lattice, where both the

18: radius of convergence and the Taylor coefficients to any order can be

19: obtained by a direct determination of the Lee-Yang zeros.

20: Interestingly, while the relative statistical error of the Taylor

21: expansion coefficients steadily grows with order, that of our

22: estimators stabilizes, allowing for an accurate estimate of the radius

23: of convergence. 

24: In particular, we show that despite the more than 100\% error bars on

25: high-order Taylor coefficients, the given ensemble contains enough

26: information about the radius of convergence. 

27: \end{abstract}

28: