As usual for me, I compute here the Hurst exponent with the lag=1 approximation of
Koutsoyiannis (2003) [https://doi.org/10.1623/hysj.481.3.43481] equation (5)
ACF(j)=H(2H-1)j^(2H-2), j>0
setting j=1 (i.e. lag=1) we derive: ACF(1)=H(2H-1) or
H=(1.+sqrt(1.+8.*acf(1)))/4
This value, at its best, can be considered a lower limit of the real (correct) Hurst exponent.