Quick note before I dive in: this is a personal account, not a recommendation. I traded a small account by professional standards, and a lot of what I picked up over the year was specifically about how to operate when every basis point matters. The motto I kept coming back to was the trader who learns to manage cents will eventually learn to manage millions.
I’m writing this down because going through the year on paper is the only way I get any real honesty out of it. Staring at the P&L curve never works. So here it is, blunt and short.
How the year started
I started with a stock-picking setup that looks pretty naïve to me now. Rank a universe of equities with a quantitative model, hold a small handful of names for a few weeks, rotate. Simple on paper. The execution wasn’t. My first models were over-engineered and tuned way too tightly to history. I’d talked myself into believing that complexity equals edge, and that good in-sample fit means good prediction. Neither is true.
The first months didn’t really teach me anything. The tape was kind and the numbers were fine, but I couldn’t tell you why. That’s a horrible spot to be in as a trader: results without a clean explanation behind them.
Adding options to the mix
Around the middle of the year I started using options. Two basic structures: OTM cash-secured puts on stocks I’d actually be okay owning at the strike, and covered calls on positions I already held. Two reasons for it. One, options give you a much finer instrument to express a view. Two, premium is a source of return that doesn’t depend solely on picking the right stock.
A chunk of what came next was just the tax of learning a new instrument with real money on the table. There’s a real difference between reading the textbook and watching the position breathe at you in real time.
The four mistakes that stand out
Gamma on short options got me a few times. Textbook delta is one thing. Watching your exposure shape-shift on a Friday afternoon close to expiry is something else. I was late to adjust more than once. A short option looks linear until it doesn’t.
I sized in notional when I should have sized in vol. Sometimes I was right on direction but still had way too much exposure for how much the underlying actually moves day to day. You want to size to the worst plausible move, not the one you’re expecting. On a small account the urge to size up is constant. Resisting it is most of the job.
My early ranking models were too clever. Too many parameters, fitted way too closely to history, and not enough of a real economic story behind them. I ended up replacing them with simpler systems built on fewer factors, with cleaner out-of-sample testing and an actual reason for each piece. Robust over optimal, every single time.
My risk framework assumed nicer markets than the ones that actually show up. The early version baked in correlations and vol that were closer to “calm Tuesday” than to an average day. You plan for ordinary stress, not for sunny weather.
None of these blew anything up on their own. Together though, they ate time, attention, and capital I needed elsewhere. The decent news is that every one of them is fixable, and the fixes are what made the second half of the year a lot calmer.
What I changed
Honestly, the back half of the year was less about new ideas and more about cleaning house. Fewer moving parts. Smaller average size on options. Tighter rules for adjusting and rolling. A genuinely separate out-of-sample test instead of the half-honest version I used to run. And, embarrassingly basic, an actual pre-trade checklist written down somewhere I’d see it.
The lesson behind it all
The whole year, looking back, was a long exercise in scaling down before scaling up. Learning to size. Learning to adjust. Learning to call a mistake early instead of nursing it. Learning to say no to complexity I didn’t need. The small account part forced me to take every basis point seriously, and that discipline is the thing I’m carrying into next year, more than any specific trade idea.
The trader who learns to manage cents will eventually learn to manage millions. I believe that more now than I did twelve months ago.