Quants, robo-advisors, HFT, and algorithmic based trading desks dominate Wall Street now. It’s has gotten so egregious, I literally base my investment decisions on how a quant would respond to sell off, the criteria it’d be interested in, and how negligent the programmers of said quant might be. Gone are the days when retail meant something to Wall Street. Trades are now done, based off AI, machine learning, and quantitative analysis. Certain triggered get hit, and a cascade of orders follow.
Goldman’s DJ Sol is out with a ‘duh’ moment, suggesting robots are to blame for the recent slide.
Source: CNBC
“There’s no question when you look at last week, some of the selling is the result of programmatic selling because as volatility goes up, some of these algorithms force people to sell,” Solomon told CNBC’s Wilfred Frost. “Market structure can, at times, contribute to volatility and one of the things that we’re spending a bunch of time thinking about at the firm is how changes in market structure over the course of the last 10 years will affect market activity.”
He continued.
“All those things are untested over any duration of time with severe stress,” Solomon said. “Now, when we see a little bit of stress, you can see reactions that might lead you to believe that there’s a risk that with more significant stress that could play a bigger role. I wouldn’t predict that, but it’s certainly something we watch.”
According to JP Morgan, as of last year, discretionary traders accounted for just 10% of trades.
Looking even deeper into quant trading, the high frequency varietal accounts for ~52% of all trades. That is literally the dumbest form of trader — the algorithm that rips trades to and fro in fractions of a second — with the goal of milking each trade for a small profit. Reminiscent of the scheme purported in the cult classic flick Office Space, those pennies quickly add up, and produce billions in profits to Wall Street’s top trading desk.
How profitable are these strategies?
Back during the hey day of HFT, 4 investment banks posted a record 61 straight days without a loss.
Even more perverse, and just to show you how the deck is stacked against the plebeian retail investor, JP Morgan lost money two days over 4 years.
If you enjoy the content at iBankCoin, please follow us on Twitter
Dumb one thousand times a second appears brilliant.
Digital can go through the most rudimentary yes-no-and-or sequence, but because it does it a million times a second, the brain dead nature of the process is hidden. In the audio area, elegant analog simplicity of a resistor and capacitor is replaced with dozens of lines of code in digital. And it took nearly twenty years for the digital implementation to sound better.
As you pinted out, program trading is not some brilliant AI-brain making split second decisions – it is just a ordinary computer program folliwing the instructions of its programmers. So blaiming “program trading” on market movements is like crediting fire hoses for putting out a fire or blaming kerosene for causing one.
HFTs provide 0 value to the market, yet Democrats and Republicnas let them continue. The calim is that they provide liquidity, but at EVREY instance where liquidity is needed (ie, crashes), they disappear.
In 2012: “Goldman Sachs had 236 profitable days of trading last year (and just 15 days of losses)” https://qz.com/58535/goldman-sachs-had-236-profitable-days-of-trading-last-year-and-just-15-days-of-losses/