my main goal for 2026 is to continue to branch out into new edges, and question my accepted ideology that I can only compete in the risk premia arena and spend months on end on drawdowns.

that might be true, but I gotta see it for myself.

I also think I've always framed the question in the wrong way and its more of a spectrum rather than a binary answer.

how much more sharpe can I generate, on top of what I currently already do?

I've seem a glimpse of it this year already and I am spending all of my time designing my entire infrastructure to make it feasible.

even if I dont make it work, I now have a more robust infra, unrecognizable to what it was before.

I've ensured that my execution engine can support this and scale by:

- taking any type of model
- standard data sources + alternative data
- cross exchange execution
- scalable record keeping
- risk alerts

now I've been spending a lot of time making sure that my "alpha" generation engine is super efficient and I can take an idea from ideation to simulation (if applicable) into prod with a click of a button.

also making the effect modeling again done with a click of a button and dont have to be restructuring each time I want to test a new idea (@systematicls @quant_arb ).

making alpha generation almost factory like so I can keep myself competing without exhausting myself in adjacent development tasks that are non-pnl generating by nature.

a lot to get done and excited for this new chapter.