One of the best posts on trading that I've encountered in years is Brandt's analysis of how randomness affects trading profitability. The same exact trading methodology can yield winning or losing outcomes over a limited time horizon simply due to the random strings of winning and losing trades. This is demonstrated extremely well by Brandt's post, where he assumes a given win rate and ratio of average winning size to losing size and then shows how those metrics can lead to profitability or unprofitability over a run of 100 trades.
There are two important implications of Brandt's analysis:
1) If you ramp up your trading size, the increased risk over a random series of losing trades can devastate your account. Your trading size should not be a function of some high target return that you hope to make, but rather a function of the random strings of losers that you can survive.
2) Just because you're going through a losing period doesn't mean you're trading a losing methodology. Changing sound methods in the middle of a random string of losing trades would be like a batter changing his swing after striking out a few times. That is what helps turn normal setbacks into prolonged mental and performance slumps.
The bottom line is that we can read far too much into short-term trading outcomes. Losing trades don't necessarily mean we're trading poorly and winning trades don't necessarily suggest that we have a hot hand. We can gain knowledge from analytical mistakes and creative insights, but it's also important to retain the wisdom that sometimes we will trade well and lose and other times we can trade poorly and win.
Not every move--of markets, or of profit/loss--is meaningful.
Further Reading: Lessons From a Successful Trader
.