@ardiekap:

ardiekap
ardiekap
Open In TikTok:
Region: ID
Monday 02 November 2026 20:29:50 GMT
173
1
0
1

Music

Download

Comments

There are no more comments for this video.
To see more videos from user @ardiekap, please go to the Tikwm homepage.

Other Videos

Alpha decay backtest: five published S&P trading edges tested on 15 years of E-mini S&P and Nasdaq futures, and a rule for when to stop trading one A viewer asked what alpha decay looks like on real data. McLean and Pontiff (Journal of Finance, 2016) found that published stock predictors earn 58% less after publication. I tested five famous S&P edges on one-minute ES and NQ data from 2011 to 2025, one contract, $16.50 a round turn: the pre-FOMC drift, the turn of the month, the overnight drift, intraday momentum and the Monday effect. Two died. Shorting Mondays made +$12 a trade in 2011-2015 and -$335 a trade in 2021-2025, -$88,800 on ES since 2011 (on NQ it went -$30, -$224, then -$567 a trade). Intraday momentum got worse every five years. The pre-FOMC drift did the opposite of decay: it lost before 2015 and made almost all its money after. Then a stop rule: bench any edge whose net over the last 1, 2, 3 or 5 years is negative. Walked forward (expanding by year from 2015, rule picked on past years only), it made $161,616 on ES and NQ against $26,217 for trading every edge. But it also benched the overnight drift in 2024, its best S&P year (+$46,095), and 884 of 1,000 random schedules that benched the same number of years did better. It found the dead edges; its timing was worse than random. Takeaway: kill an edge that keeps losing for years. Don't bench one for a bad year. Not tested: other markets, other edges, monthly or shorter checks. Every check setting beat trading everything on both markets; at $33 a trade the overnight drift still made $53,474. I built this in QuantPad. It has every trade on the CME since 2010. Link in bio. #QuantPad #backtesting #quanttrading #futurestrading #algotrading
Alpha decay backtest: five published S&P trading edges tested on 15 years of E-mini S&P and Nasdaq futures, and a rule for when to stop trading one A viewer asked what alpha decay looks like on real data. McLean and Pontiff (Journal of Finance, 2016) found that published stock predictors earn 58% less after publication. I tested five famous S&P edges on one-minute ES and NQ data from 2011 to 2025, one contract, $16.50 a round turn: the pre-FOMC drift, the turn of the month, the overnight drift, intraday momentum and the Monday effect. Two died. Shorting Mondays made +$12 a trade in 2011-2015 and -$335 a trade in 2021-2025, -$88,800 on ES since 2011 (on NQ it went -$30, -$224, then -$567 a trade). Intraday momentum got worse every five years. The pre-FOMC drift did the opposite of decay: it lost before 2015 and made almost all its money after. Then a stop rule: bench any edge whose net over the last 1, 2, 3 or 5 years is negative. Walked forward (expanding by year from 2015, rule picked on past years only), it made $161,616 on ES and NQ against $26,217 for trading every edge. But it also benched the overnight drift in 2024, its best S&P year (+$46,095), and 884 of 1,000 random schedules that benched the same number of years did better. It found the dead edges; its timing was worse than random. Takeaway: kill an edge that keeps losing for years. Don't bench one for a bad year. Not tested: other markets, other edges, monthly or shorter checks. Every check setting beat trading everything on both markets; at $33 a trade the overnight drift still made $53,474. I built this in QuantPad. It has every trade on the CME since 2010. Link in bio. #QuantPad #backtesting #quanttrading #futurestrading #algotrading

About