Home Backtesting and trading The Pitfalls and Power of Automated Trading in 2024

The Pitfalls and Power of Automated Trading in 2024

In one of my current strategies, I was filled en masse today. Usually, that strategy gives me on average 4 fills a day, but today I got 59 fills. Quite unusual! So I stopped the program because I suspected something was wrong. Having continued the program I theoretically would have got over 100 fills/stocks. When doing automation I always see the black swan in the back of my head so I stopped it. Better safe than sorry, I thought. A quick look at the P/L showed indeed something could be wrong: I was down about 6 000 USD in seconds and I was fearing this was just the tip of the iceberg. After having a quick look at some of my fills I discovered an insane volume, and this was in stocks with no news. Looking at those incredible moves I suspected some automated trading program must have gone completely wrong and my strategy actually will turn very profitable when this selling stops. So I sent the program again to buy some more. And yes, after some 30 minutes everything cooled down and the P/L just ticked upwards securing a very profitable day. This was one of the rare moments in a trading year when the opportunity to make really good money suddenly came out of nowhere. And yet again I sit here kicking myself for not trading bigger size. Good opportunities for us who provide liquidity, but you have to go for the jugular!

Currently, I haven’t seen anything on the news except for some rumors that an algo order that should have been executed over 5 days got executed in 5 minutes. I guess someone will be fired….. Automated trading gives power, but can turn almost lethal when executed in the wrong way. And don’t even think about if the quotes feeding the program is wrong.

Look at BRK/B, PL and DDD today just to mention a few. DDD went from 38 to 34 USD in just some minutes. PL had a range of about 27 to 25.5 once every minute.  PL went about 5% after selling abated, and DDD rose about 9% from the bottom!


– Why might traders decide to halt their trading program during unusual market activity?

Traders sometimes stop their trading programs when they suspect something unusual is happening in the market that could result in significant losses.

– How can traders effectively manage the challenges associated with automated trading systems?

Traders need to be prepared for unexpected market events, continuously monitor their automated systems, and implement risk management strategies to safeguard their investments.

– How can traders effectively navigate the challenges of automated trading?

Traders can navigate the challenges of automated trading by being prepared for unexpected events, maintaining continuous oversight of their automated systems, and implementing risk management strategies to mitigate potential losses.

  • Oddmund Grotte says:

    Still holding some of them as I expect many of them to be busted. We’ll see later today.

  • Oddmund Grotte says:

    I’ve been waiting all day for the ruling from the NYSE. So far just 6 tickers have been busted, I have none of them. About 140 tickers are under review by the exchange, I traded in about 1/2 of them.

    Apparently Knight, a market maker, had some problems with their algos this morning. They have issued a press release about this.

    Right now it might seem that there will be no other busts. Knight (ticker KCG) got absolutely crushed towards the close, and I expect there are others player out there that know a lot more than me and sold/shorted the stock. The stock getting hammered is a good sign. Obviously KCG will suffer a big loss if there are no further busts.

  • Oddmund Grotte says:

    I had a look at different discussion forums etc. this morning. It appears several independent traders (traders like you and me having home office) have made many hundred thousand dollars just because of this error at KCG. Some have shown their blotter to back it up. Wow, a lot to prey on…

    Still, I’m surprised that this amok program at KCG was allowed to run for 29 minutes before someone shut it down or found out what was happening.

    Still good to know that the small guy can prey on the “big guys”!

  • Espen says:

    And this must be an pretty good example on the risk when a computer is running the show. If the instruction (codes) are wrong or just contain an error, it can make a big wave in the market.

    And, could this be an example where a stoploss-function would have “smashed” your stake quite a bit?

    You mention that you’ve read about the guys making big profits on this, what kind of stocktalk forum would you recommand to hang out on to keep up with other traders? There are so many to choose from.

    • Oddmund Grotte says:

      Yes, if I had a stop in the market yesterday it would have been a big loss. This was a perfect example of how a stop-loss ruins a potential good day. No way I found the bottom/top but had to go through a nasty loss for about 20 mins before the program stopped. And the slippage to get out in that frenzy would have been enormous.

      A good forum is http://www.elitetrader.com

  • Oddmund Grotte says:

    The day after this algo spectacle, KCG sends out the following press release:

    “Knight Capital Group Inc. said electronic-trading glitches in its system that caused price swings in dozens of stocks this week are likely to cost the brokerage firm $440 million.

    Knight said it is pursuing alternatives, such as financing or strategic business moves, to “strengthen its capital base”.

  • >