Whatever analysis you use to gain a trading edge and therby determine whether to enter a market or not - it can only tell you what the probability of future price movement is (whether or not that probability plays out or not on that particular occassion) based on it's historical precedent. Therefore depending on the historical strike rate of the analysis that makes up your trading edge (assuming the analysis actually gives you an edge) - some trades as a result of it will fail-ie result in a losing trade, it is - this is the nature of things. This includes any analysis/edge you derive from using tech indicators.
If a trade fails how so/why so ever you entered the market, it is for one of the following reasons.
a. The edge was present but on this occassion it resulted in a losing trade
b. The edge was not present and the trader lost his discipline/patience and entered when it's conditions were not present/all present.
c. The analysis used to gain the edge is flawed and no actual edge is present over a large sample.
G/L