It's time to short the market.
- Thread starter Junkou
- Start date
Those interested in this AI system might want to consider the impact of "gap openings".
The system generates a signal on the close with the expectation it will be executed on the next opening.
In the case of the latest 2 TQQQ signals... "10/11, Buy @ 53.59", the next opening was actually ~58. And the next sell signal, "10/22, Sell @ ~58, but the next opening was ~54. The system will record this round-trip as "~+10%", but in reality... assuming one executed on the next opening as instructed.... the net result is ~ -8%. That's an 18% differential.
OTOH... on the same "10/22 Sell on Gold Miners"... the gap opening was 3% higher and in your favor.
So... Will the favorable/unfavorable gaps balance each other out? As part of the signal is "change in momentum", I'm thinking they will probably not balance out. Followers will likely be on the disadvantaged end of the majority of the gaps and for larger slippage. This does not negate they system, but one would want to look into the potential negative impact of gaps in its performance. (It would be an easy thing for them to include the "gap opening" in their actual results, so one has to wonder why they're not. To be fair, one should check out all of the signals so far to get an idea of the "gap impact".)
I also noticed... for the NDX, their "indication so far" on the 10/22 Sell signal....
The system has 7141.67 as signal price and 6785.95 as last price. They're showing that as a +9.52%. That calculation is wrong... but maybe just a mistake they haven't caught?? The result as stated would be ~+5.2%. If executed on the next opening, the gap fill was at ~7000... and with last price 6785.95, the actual return so far is ~+3.1%, not +9.52%.
Draw your own conclusions.
Last edited:
Well done Scat
You are absolutely correct. However you forget two things:
1. Indices are not tradable. You have to trade index futures, which are still open after 4pm. If you execute your orders around 4:30pm, your chances of getting big gap is slim.
2. Most ETFs are still trading in after hours with sufficient liquidity to fill your orders. If you want trade the most profitable YINN, you can use FTSE China A futures, which is open at night as well.
Okay, I'll try again.
A commercial product? Whose? Cost?
Homebrew? What broker connection, if any? API? What language/version?
If you wrote it yourself, how long did it take you? Are you selling it? Terms?
Whether commercial or homebrew, if you were to recreate the system right now, what would be your choices amongst brokers, platforms, APIs, languages, etc?
What you say is partially true... but your "performance record" doesn't reflect that, does it?
I can see the latest results of shorting NDX is 5% on the system, not 9.52% you mentioned.
I loaded up minute data of NQ. It was about 7155 between 4:30pm and 7pm on Oct 22. During the same time frame yesterday, NQ was about 6850. A 4.3% return. While it is still different from using closing prices, over the long term, they should balance out. Even trading TQQQ between 4:30pm and 6pm, you would still gain about 13%.
Ideally, if signals can be sent out around 3:50pm will be perfect.
I understand that most traders here are active day traders. They don't and won't believe that trading on daily chart can achieve much better results. The prime principle of success is to control your positions or leverage. If you don't use leverage and trade on NQ, even a 5% draw down is not a nightmare. The MAGIC shows results of NQ would be over 80% without using leverage.
Anyway, the MAGIC score is positive yesterday. With today's sharp sell-off, MAGIC will be likely to tell us buy at the end of the day. Let's see.
Anyway, the MAGIC score is positive yesterday. With today's sharp sell-off, MAGIC will be likely to tell us buy at the end of the day. Let's see.
Apologies for the naive question, but how is this done without getting eaten alive by brokerage fees?