We analyze results from the first eighteen months of monthly sub-mm monitoring of eight star-forming regions in the JCMT Transient Survey. In our search for stochastic variability in 1643 bright peaks, only the previously identified source, EC53, shows behavior well above the expected measurement uncertainty. Another four sources, two disks and two protostars, show moderately-enhanced standard deviations in brightness, as expected for stochastic variables. For the two protostars, this apparent variability is the result of single epochs that are much brighter than the mean. In our search for secular brightness variations that are linear in time, we measure the fractional brightness change per year for 150 bright peaks, fifty of which are protostellar. The ensemble distribution of slopes is well fit by a normal distribution with sigma ~ 0.023. Most sources are not rapidly brightening or fading in the sub-mm. Comparison against time-randomized realizations shows that the width of the distribution is dominated by the uncertainty in the individual brightness measurements of the sources. A toy model for secular variability reveals that an underlying Gaussian distribution of linear fractional brightness change sigma = 0.005 would be unobservable in the present sample, whereas an underlying distribution with sigma = 0.02 is ruled out. Five protostellar sources, 10% of the protostellar sample, are found to have robust secular measures deviating from a constant flux. The sensitivity to secular brightness variations will improve significantly with a larger time sample, with a factor of two improvement expected by the conclusion of our 36-month survey.