Pinned Tweet
Hey guys! I’ve just released a refined, compact edition of one of my newsletter series on the foundations of quantitative trading. It is a theoretical introduction designed for aspiring quants who are beginning to explore the field. The focus is on first principles, method, and k behind quantitative trading.
Link in the comments. I hope you enjoy it.
Link in the comments. I hope you enjoy it.
13
47
519
31,706
The problem is that the additional reward depends on price changes and sentiment, not on how much of each asset the agent actually holds or trades.
The authors acknowledge in Section 3.1 that the dynamic program uses the entire realized order-flow path of the evaluation day, including information that becomes available after each trading decision 😬 in other words, look-ahead bias
2
1
7
499
There’s only one reason I haven’t told X to fuck off. 🤣
Auditor resignations are another data source because it can reflect governance and risks. Indeed, his job can be assessed as any other event. Here, the bearish setup carries high conviction when a resignation intersects with deteriorating liquidity.
1
1
9
741
Auditor resignations are another data source because it can reflect governance and risks. Indeed, his job can be assessed as any other event. Here, the bearish setup carries high conviction when a resignation intersects with deteriorating liquidity.
Resolved filing delays create opportunities when the 10-Q or 10-K report numbers that match expectations. The long thesis strengthens once accounting issues are clear, and the stock begins recovering the NT drop. Here the entry triggers after the filing posts.
2
1
13
1,497
The authors acknowledge in Section 3.1 that the dynamic program uses the entire realized order-flow path of the evaluation day, including information that becomes available after each trading decision 😬 in other words, look-ahead bias
Interesting interpretation for the Sunday morning. I only have to say that learning a representation of market microstructure is not equivalent to learning a predictive, economically useful, or causally accurate model of financial markets.
2
3
24
1,794
Resolved filing delays create opportunities when the 10-Q or 10-K report numbers that match expectations. The long thesis strengthens once accounting issues are clear, and the stock begins recovering the NT drop. Here the entry triggers after the filing posts.
NT 10-K is like a disclosure-risk event. With it you can score the stated reason for delay (may accounting, expected resolution time, or management). Use it as a long-avoidance filter first, then short only when price and liquidity confirm deterioration.
2
3
14
1,850
Interesting interpretation for the Sunday morning. I only have to say that learning a representation of market microstructure is not equivalent to learning a predictive, economically useful, or causally accurate model of financial markets.
The NFE achieves a Sharpe ratio of 1.49 against 1.45 for AFM using the same residual factors. This difference of 0.04 is too small to distinguish from sampling noise.
3
1
24
3,049
NT 10-K is like a disclosure-risk event. With it you can score the stated reason for delay (may accounting, expected resolution time, or management). Use it as a long-avoidance filter first, then short only when price and liquidity confirm deterioration.
What about the 8-K? Use it as an information-quality event where high promotional intensity combined with low financial specificity creates an urgency-risk score. Trade normally only when subsequent price or fundamentals validate the claims.
3
2
7
2,152
