I was able to do it consistently for over 3 years. I had a pretty successful strategy and could execute on that strategy using consumer grade trading APIs from

I was able to do it consistently for over 3 years. I had a pretty successful strategy and could execute on that strategy using consumer grade trading APIs from Interactive Brokers. I had even extricated luck from most of the equation: ~9:1 profit/loss. It was an arbitrage strategy, so the luck variable became how often I had opportunities arise. I quit not because I couldn't make it work, but because I couldn't handle working alone for such a long time from a mental health perspective. Here are my caveats: 1) low capacity strategies are the only ones that have a high enough profit/loss ratios but haven't been completely consumed by the hedge funds. 2) Because they are low capacity strategies, you can forget about compounding. Be happy with consistent cash flow. 3) Because they are low capacity, you can only feasibly scale across more products, not with bigger bets. This means that at some point, your connection bandwidth will become a bottleneck, even after you have graduated to professional data brokers. That then translates to a parallelization engineering problem. For me, it never made sense to deal with that complexity. It was better for me to have suboptimal earnings than to try to horizontally scale heterogeneously-active trading systems across servers in a data center. Once you take into account the caveats, you can make a tech-industry level salary on your own with a lot of work. For some, the freedom may be worth it. For me it was, up until it wasn't.