What happened Mon 6/21?
Good thing I wasn’t on Auto Trade yet~!
What happened Mon 6/21?
What happened was the “stop” hit. The IB account driving the system (system is TOS) sold at $18.85. Unfortunately that’s the tradeoff of having an intraday “stop”, versus going with the decision made at BOD and holding. A “flash dip” / “fake out” will take you out. But, in the long run having stops beats not having stops.
One note, when the stop hit, after it was sold at $108.85 in my IB account, then autotrader sends same trade to subscribers. All this happens in a second or two, I believe. All those shares of TQQQ gets sold as market orders. This caused subscribers to sell from $108.74 (first ones to sell) and down to $108.53 (last ones to sell). It may have contributed to the length of that red bar pointed to by arrow. So basically a $41million market sell order occurred, which equals about 370k or so shares, which matches the red volume bar below (meaning most of the shares in that red volume bar is this system selling).
Edited/added: BTW we’re trying to figure out a way to better deal with this situation (reduce slippage to get better prices).
fascinating. its one of the weird concerns with popularity and success. everyone wants to make money but not all ETFs can handle large market orders without moving the price. granted 32 cents aint much but considering the daily volume and the price, that’s actually amazing 1set of market sells can move it that much.
Wow! You can really move markets!
The $41million is dollar amount. Calculated to shares it would be about 370,000 shares (a little over 1% of the average daily volume). But yeah no one can corner that market. Also TQQQ is different from a stock, it is backed by entire nasdaq, and synced by AP’s. Though in the short time the $41m order went through, I imagine the AP’s can’t keep up? Does anyone know more about how that works?
OK I see what you mean now, sorry.
But be careful now, traders who can move the price of such a liquid financial market are usually put on the SEC radar…
Nah, just kidding, keep it up.
Haha, I suspect it’s just slightly moving TQQQ the ETF instrument (since it isn’t instantaneously tracking 3xQQQ during the day, and TQQQ dollar volume is small compared to actual nasdaq weighted stock’s average dollar volume), not underlying NASDAQ, no way not even close to doing that!
Added: But yeah for TQQQ itself, the AUM is getting to an amount where when we swap from TQQQ to SQQQ (or vice versa, etc), it’s getting to be an issue.
@TheForexSniper You may be kidding about the SEC, but I was wondering the same thing. What if a trade leader becomes 'too successful? Too much volume in total to stay below the radar? A while ago, C2 itself had an issue with the SEC (or similar organization) and had to pay a fee for ‘moving the markets’ with one of their trade leaders. It’s weird because that’s exactly what the market is about and what bigger players do.
Of course it’s weird, you cannot fine someone because “too many” (??) traders bought or sold a financial instrument at the same time!
I mean what kind of nonsense is that?
Was Soros fined when the broke the Bank of England with his HUGE short GBP/USD position and went home with an extra $1 BILLION in his pocket?
@RaoulSuurmeijer do you know which strategy was that? Were the subscribers of the strategy affected in any tangible way?
@Alvinlee No idea which strategy that was, it wasn’t disclosed as far as I remember. I read it somewhere here on C2 on a sort of news-flash pop-up after login in. That item was written by @MatthewKlein maybe you can still find it in the news archives.
Why don’t you break up the buy and sell orders into 3 or 4 orders with 1-2 minutes between each order?
IB has an accumulation algorithm as an order type on the buy side. The individual algos on C2 would have to mimic this functionality.
“Once you define the baseline for slicing the full order into smaller increments and submitting them at specific time intervals, you can apply randomization to make your large order behave like a series of unrelated small orders. Use the algo checkboxes to randomize increment size, the time period for submitting increments and more.”
Unfortunately for autotrading, it would only send the order as a “market sell” once the last share is sold. Subscribers will only sell once the trade leader has sold their last share in the order. So, the only way is to make fully seperate orders on IB (unless someone else has more info).
suppose one has 100 followers. if one then make 10 seperate orders, all the followers also make 10 seperate orders (low amounts high fees).
it would be better for thin markets if each follower bought market at a different time during the day. or at least so that the trade leader could mark an order as such.
this would mean followers would get different fills and different performance, but on average they would do a lot better as all market buys did not hit the market at the same time.
something like “buy market at random time inside the next xyz hours of open trading”
That would probably make a stock based strategy able to run 10 times if not 100 times as much money, without incurring crazy fees bc everyone of the 100 copiers, buys 100 times during the day.
the tracking, however, would be much more off as the first and the last buyer probably will get a different fill than the trade leader - simply bc so much time have passed.
so… optional setting on an order - followers buy at random time from now and N hours forward (for instance).
as a side effect, the size of the purchase should probably be a fraction of NAV instead of a number of shares. Actual number of shares to buy could be determined when the market order is created by the collective2 system
Have you considered using the same AI in the futures market with NQ or micro NQ?