When the program starts the optimization from the initial strategy, it makes many intermediate optimization steps until it reaches the highest Net Balance. Normally the Net Balance is in direct proportion to the SQN, so it is not surprising that the first optimization reaches a high SQN result.

When you start the optimizer to optimize SQN, it tries to change the parameters in order to reach the highest possible SQN. However, it may miss some step from the first optimization and to not be possible to reach so high SQN. For example it is possible to have an intermediate step where the Net Balance to be higher than the Net Balance from the previous optimization step, but the SQN to be lower than the previous. In such case, the optimizer will accept this step only if it searches for higher Net Balance, but not when searches for higher SQN. The opposite case is also possible.