Phase plot of a first order process.

Hello, I have a very basic question about phase plot of first order process G(s)=K/(tau*s+1), tau>0.

My understanding is since we have a negative pole here, there show be a lag in the frequency response, i.e. phase angle <0. It should hold no matter what the sign of the gain, K, is. However, when I used bodeplot command to produce phase plot, I got a lead, a positive phase angle when K<0. The following are the code I used:

clear all;
k_dc = 5;
Tc = 10;
s = tf('s');
G = k_dc/(Tc*s+1)

Above produces a phase plot with negative phase angle value, a lag, as expected.

When I but if change k_dc=-5, phase shift is positive.

How, should I interpret this? Should I think the starting point of second phase plot 180 as 0, since the max phase change is 90 deg? Thanks a lot!