1: \begin{abstract}

2: The ${\overline{\rm MS}}$-like schemes in QCD have in general the running coupling which

3: contains Landau singularities, i.e., singularities 

4: outside the timelike semi-axis, at low squared momenta. 

5: As a consequence, evaluation of the spacelike quantities, such as

6: current correlators, in terms of (powers of) such a coupling then results in

7: quantities which contradict the basic principles of Quantum Field Theories.

8: On the other hand, in those QCD frameworks where the running coupling remains 

9: finite at low squared momenta (IR freezing), the coupling usually does not 

10: have Landau singularities in the complex plane of the squared momenta. 

11: I argue that in such QCD frameworks the spacelike quantities should not be evaluated 

12: as a power series, but rather as a series in derivatives of the coupling with respect 

13: to the logarithm of the squared momenta. Such series show considerably better 

14: convergence properties. Moreover, Pad\'e-related resummations of such 

15: logarithmic derivative series give convergent series, thus eliminating 

16: the practical problem of series divergence due to renormalons.

17: \keywords{low-energy QCD, IR freezing, logarithmic derivatives, Pad\'e-related resummation}

18: \end{abstract}

19: