The study is based on empirical syntheses in understanding the sagacity of investors, stock returns, and stock market efficiency behavior in the theoretical frame of behavioral finance patterns. The analyses employed a long extent of Pakistan stock market Returns data from June 1994 to November 2019 along the two segments of economic are named as Military phase (1999-2008) and Democratic phase (1994-1998) (2009-2018) to explore the weak form of efficiency of the Pakistan stock market. In this research Autocorrelation, variance ratio tests are conducted on the returns (weekly based) KSE 100 index adaptive pattern of market efficiency across the overall period as well as for both the Military Phase and the Democratic phase. The efficiency tests show trends of a stock behavior, and accordingly forming bounded-adaptive market efficiency. These tests confirmed the presence of asymmetric dynamic behavior of returns obviousness in addition to risk and return relationships across two political states. The overall and democracy results are insignificant that indicates that there is no autocorrelation between a series of returns and its lagged value for the null hypothesis AR=0. Besides this, in the military phase, the results are significant which specify that there is autocorrelation between the series of returns and its lagged value for the null hypothesis AR=1.So the null hypothesis is rejected for the military phase. The overall results of the VR test show that unbiased variance estimation (Homo) is significant at 1% for the null hypothesis VR=1. But heteroskedastic robust standard error estimation (Hetero) is insignificant for the null hypothesis VR=1. In the military phase, the results show that unbiased variance estimation (Homo) is significant at 1% for the null hypothesis VR=1. Heteroskedastic robust standard error estimation (Hetero) is also significant for the null hypothesis VR=1. In democracy phase results show that unbiased variance estimation (Homo) is significant at 1% for the null hypothesis VR=1. But heteroskedastic robust standard error estimation (Hetero) is insignificant for the null hypothesis VR=1. These shreds of evidence provide support to bounded adaptive rationale of investors' behavior, dynamic stock return behavior and accordingly forming bounded adaptive market efficiency.