One of the more effective methods for exiting a trade is to employ a Risk Reward Ratio on the trade. By doing so, there is no question of when to exit...the exit occurs when the limit is triggered and you are profitably out of the trade. It is very straight forward and it eliminates any emotion since the stop and limit were put in place prior to the trade ever being entered.
For example, if a trader were to employ a 1:2 Risk Reward Ratio on a trade (the minimum that we would recommend) and a 100 pip stop were set, then a 200 pip limit would be set. With a 50 pip stop a trader would employ a 100 pip limit and so forth.
Another method that could be employed would involve the trader trading multiple lots. Using this scenario, a trader would open, let's say, three positions on a trade. At a predetermined level of profitability, perhaps +75 pips, one of the lots would be closed thereby locking in that amount of profit. On the remaining lots, the stop would be moved to breakeven...the point at which the trade was entered. Then, should the remaining lots move in the favor of the trader, the stop could be advanced periodically to continue to lock in profit. The worst that would happen on the remaining lots would be that the pair would do an about face and the trader would be stopped out at breakeven...no gain on the remainder of the trade but no loss either. Should the trade continue in the direction of profitability, the second lot could be exited at perhaps +125 pips, thereby adding to the profit, and the stop could be advanced above breakeven on the third and last lot.
A trader could also simply observe levels of support and resistance. As their trade is approaching levels of significant support or resistance, they could exit all or a portion of their position at that point, or perhaps simply tighten their stops. By so doing, should the pair hit the support/resistance level and retrace, a major portion of the profit would be protected and should the pair breakthough and continue on, the trader would still be in the trade.