John McDonald/Co owner

In 2018, I began day trading with a small account, and the early success was exhilarating. My returns defied expectations, fueling a belief that I had an edge over other traders—or so my ego thought. That confidence led me to a risky gold trade, where I invested my entire account just before a Federal Reserve rate announcement. Holding a long position, I watched my account plummet to under $100, a devastating loss that taught me four critical lessons: I had the potential to trade, but I needed to master risk management, deeply understand every sector I traded, and learn to code to rigorously test every scenario.

​

Determined to grow, I honed my skills in coding custom indicators. While I wasn’t the best, I gained a clear understanding of my goals. Along the way, I collaborated with coding communities and met Laz, who later became my co-owner. Together, we tested relentlessly—forward and back—to build the “perfect algo.” Through this process, I discovered that most indicators on daily timeframes underperformed. Randomizing signal entries revealed that risk management, particularly a specific trailing stop, could achieve a 57% win/loss ratio even with random entries. This breakthrough became the cornerstone of our approach. We’ve since identified indicators that enhance this metric, but only for specific symbols and timeframes.

​

MarketFragments.com is designed to drastically shorten the learning and testing curve for traders—far less than the five years it took me. Whether you’re a beginner or an advanced trader, our tools provide real-time data, testing every possible trading combination. Our current features, which we aim to expand, are built to empower all traders with profitable, data-driven strategies.

​

Lastly, my main goal is to grow through with the community.  Think tinkerer-builder type who genuinely loves the process. I am working on releasing every python code that optimized these indicators. They have strengths and weaknesses, which I am open about what to use and what's in the creation pipeline, however I'm already in the process of making Multi AI Agents that are essentially a meta-optimizer pipeline. Think:

​

• Generate constant ranges

• Build market regime filters

• Combine indicators (1000s of combinations)

• Backtest (30 different combinations of exits with 5 possible at once- think next level dynamic ** see the Doji grid)

• Forward-test

​

Since we are not using the standard indicators that everyone else has, we are generating different signals. Edge on the hedge fund guys sort of thing. You can't build a system on sticks, right ? After that, I will sit at my computer all day making trade plans that are unheard of while algos work. Life goals! 

-

-Single dad building quant tools at night.

​

​