What chappy and Sekiyo said stood out for me. If i wasn't happy with what i'm discovering, i'd check out trendlines in a high enough time frame, as in swing trading, that allows you to overcome the spreads and fees that day traders face here and there. That way, you could possibly accept a method that yields only a 1.5 profit factor (risk/reward ratio). With that, you would need a way to control drawdowns. To do that you would trade multiple markets, or slightly different time frames that would also produce around 1.5 profit factors, that can produce profit curves that somewhat hedge each other, such when one is in a drawdown, another might be going up. The idea is to control drawdowns by averaging them out in somewhat uncorrelated profit curves (assuming they are all averaging around 1.5 profit factor). Only when overall account drawdown is smoothed out to something fairly predictable, can leverage be considered. I would call what chappy is doing trend following on price action, a good thing. Its not what i propose, but it's the same general idea, so it should work. The more systematic (rules based) you can make it the better. You really have to collect plenty of data to suggest whatever you are trying, is going to work, assuming you can find an honest broker. You would keep collecting that data to compare with the results you get at any particular broker. Great if you can find a bot to implement some method, that would yield you enough data to suggest it might work. Short of that, you really must find a way to collect data manually, and then to evaluate it. I would challenge you to find, develop, or otherwise brainstorm a method that is always in the market, whatever market, either long or short, for however many hours the market is open. That limits you to some kind of trend following system, which is a good thing. I would try stay away from moving averages and try some kind of price action decision that tells you when to cease and desist going one direction, and turn around and go the other. In other words, your stop losses are actually reversals. There would be no "targets", but you would end up with a win% that is lower than 50%, but an average gain per trade that is much higher than the average loss. Just use the charts for one time frame. Achieving even a 1.3 profit factor is a big deal. You can build on that. For example, you could examine what the profit factor would be after you waited for two losses, or three losses. This does not work in a random walk situation like a casino. That is, if you expected profit factor was 1, you cannot expect to elevate that by waiting for losses. But with positive expectancy to begin with, especially 1.5 PF, it works differently, and you can expect higher PF from those trades. You can separate your data stream into long and short streams, and evaluate how to handle each of those with their own statistics. If you are sitting out a lot of trades, then you would want to be trading enough other markets, or time frames, that most of your capital is productively employed, not just sitting around. Once you have positive expectation, then you can use a higher time frame to guide/filter a lower time frame, and boost your profit factor, whenever you trade in harmony with the upper.
So yah, a lot of work. One way to make room for all the work needed is to limit any other kind of work you do to part time, like 25-30 hours a week. You won't get rich, and might only tread water, but it is not how much money you make, it's what you know you can do with it, once your research establishes a clear path forward. I think there is a clear path and you have to find it, and prove it with data. Believe only data. If you've disposed of data from before, this could be the reason for the struggling so far. Your data should tell you about drawdowns, and the average profit per trade to expect. You would then give some of that up on fees and spreads and that would tell you what time frame you would need to practice on.