First you may use back testing. The last half a year or whatever period you select. This doesn’t take too long because you can rapidly scroll through historical charts searching for the signals that would have led you to make a trade if you had been operating your system live at that time.
Back testing should give you an idea of whether a system has potential. Naturally the market isn’t going to copy in exactly the same way so you should take into account the proven fact that you could have struck lucky or unfortunate and picked a point in time when the system performed abnormally well or badly. For this reason, it is best to back test over the longest possible time and maybe split your tests so that instead of testing, for instance, one entire year when the market might have been especially powerful or feeble, take the 1st quarter of year one, quarter two of year two, etc so that you test one 3-month period from every year of four years. Here you are dealing with the live market but not using real money. This strategy is slower because you have to wait for your signals to come up for real. On the other hand, it simulates real live trading techniques with the possibility of slippage and other factors which aren’t gong to show up in back testing. Remember that you can test many systems at the same time in a demo account, provided you keep separate records of their performance. Or you may use many demo accounts. In this way you’ve a better likelihood of ending up with at least one moneymaking system at the end of your period of testing. This gives you solid real time training to prepare you at present when you go live with real cash. Most currency exchange brokers will supply free demo accounts which you can use to check currency exchange systems.