@.rmxk0: #التفاعل#

ويلز|🇾🇪⚽️
ويلز|🇾🇪⚽️
Open In TikTok:
Region: EG
Monday 05 October 2026 17:29:13 GMT
522
26
3
5

Music

Download

Comments

reto123543
🎀𝒓𝒆𝒕𝒐 𝒓𝒐𝒃🎀 :
هو انا الوحيده اللي بابايه مبيحبش الشاي وادايما بيشرب قهوه 😂
2026-10-05 19:41:01
1
user8130275614356
Lina Louie :
لا ده جدو
2026-10-05 23:26:22
0
talia.mohammed811
🕯 :
😂😂😂
2026-10-05 19:01:17
1
To see more videos from user @.rmxk0, please go to the Tikwm homepage.

Other Videos

Time series momentum replication on 16 years of futures data the original paper never saw. Moskowitz, Ooi and Pedersen published Time Series Momentum in 2012 with a sample ending in December 2009. QuantPad's continuous futures history starts in mid 2010, so we ran their exact rule on the period after the paper was written and nothing in it was fitted. The rule: once a month, if a market is above where it was 12 months ago go long, if below go short, scale every position to a constant volatility target, equal weight across 29 futures markets from the E-mini S&P to crude to soybeans. 5,261 instrument-months. Net of costs the Sharpe was 0.381, against the 1.2 to 1.4 the paper reported. A third of the published number, and with a t statistic of 1.495 it is not statistically different from zero. It did still beat equal-weighted buy and hold, which returned a Sharpe of 0.269, and it did so with a 31% maximum drawdown against 50%. Costs were not the problem: gross was 0.417 and a 2x slippage stress only took it to 0.352. Two results worth more than the headline. We swept every lookback from 1 to 24 months against holding periods of 1 to 3 months, all 72 combinations, and the best cell in the entire grid was 0.588. There is no parameter setting that recovers the published figure. And the paper's central claim was that it worked on almost all 58 of its instruments; out of sample only 17 of our 29 were positive. By year, 9 of 16 made money, carried by 2013, 2014, 2015, 2020 and 2022, while 2023, 2024 and 2025 all lost. The 16 years of daily futures history across all 29 markets is included with QuantPad, no vendor account and no data licence. #QuantPad #quant #backtesting #futurestrading #systematictrading
Time series momentum replication on 16 years of futures data the original paper never saw. Moskowitz, Ooi and Pedersen published Time Series Momentum in 2012 with a sample ending in December 2009. QuantPad's continuous futures history starts in mid 2010, so we ran their exact rule on the period after the paper was written and nothing in it was fitted. The rule: once a month, if a market is above where it was 12 months ago go long, if below go short, scale every position to a constant volatility target, equal weight across 29 futures markets from the E-mini S&P to crude to soybeans. 5,261 instrument-months. Net of costs the Sharpe was 0.381, against the 1.2 to 1.4 the paper reported. A third of the published number, and with a t statistic of 1.495 it is not statistically different from zero. It did still beat equal-weighted buy and hold, which returned a Sharpe of 0.269, and it did so with a 31% maximum drawdown against 50%. Costs were not the problem: gross was 0.417 and a 2x slippage stress only took it to 0.352. Two results worth more than the headline. We swept every lookback from 1 to 24 months against holding periods of 1 to 3 months, all 72 combinations, and the best cell in the entire grid was 0.588. There is no parameter setting that recovers the published figure. And the paper's central claim was that it worked on almost all 58 of its instruments; out of sample only 17 of our 29 were positive. By year, 9 of 16 made money, carried by 2013, 2014, 2015, 2020 and 2022, while 2023, 2024 and 2025 all lost. The 16 years of daily futures history across all 29 markets is included with QuantPad, no vendor account and no data licence. #QuantPad #quant #backtesting #futurestrading #systematictrading

About