The Quants by Scott Patterson
The Quants by Scott Patterson
Throughout the book, Patterson returns to the theme of poker/card games: from Ed Thorp's Beat the Dealer to Aaron Brown's Poker Face of Wall Street to the annual quant poker tourney. Patterson seems to not have made the mental leap that when people talk to him, major errors may accumulate: mis-perception, narrative fallacy, self-deception, deception of others, and so on....
Normally, these errors accumulate in a manageable fashion, i.e. you ask enough people and you can begin to get a balanced view. However, it may be that all the major players (including Patterson) are biased in the same direction, and we end up with a situation not unlike that which set up the crashes of '07 and '08.
(Aaron) Brown crunched the numbers and came up with a key insight: A game of Liar's Poker follows one of two paths...
Knowing how the two paths differed, along with the odds that accompanied each challenge, helped Brown crack the game...
He started circulating his strategy on electronic bulletin boards, and even created a simulator that let the quants practice on their home computers. They focused on speed. Rapid-fire bets would un-nerve the traders...
Their testing complete, the quants finally decided to put their strategy into action on Kidder Peabody's trading floor...
The traders were predictable, playing it safe: four 2s
When the quants' turn came, the bids came in fast and furious. Bid ... bid ... bid. Ten 7s, Twelve 8s. Thirteen 9s. They machine-gunned around the circle back to the top trader, who had started the bidding. Kidder's traders were dumbfounded. The silence lasted a full minute. The quants struggled to keep straight faces. Brown nearly doubled up with laughter.
The head trader finally decided to challenge the last quant. Bad move, there were fifteen 9s in the circle. He lost, but he refused to pay, accusing the quants of cheating. The quants just laughed, high-fiving. Brown had expected this. Traders never admit to losing.
Had Brown actually expected non-payment, he would have engineered a solution, e.g. "How do I know you all have money to play? The market's been wicked recently." "Ok, we put money down in front of us."
Perhaps Brown learned a valuable lesson in counter-party risk that day.
Just like Citadel, Saba was getting mauled...
(Boaz) Weinstein remained outwardly calm, quietly brooding in his office overlooking Wall Street. But the losses were piling up rapidly and soon topped $1 billion. He pleaded with Deutsche's risk managers to let him purchase more swaps so he could better hedge his positions, but the word had come down from on high: buying wasn't allowed, only selling. Perversely, the bank's risk models, such as the notorious VAR used by all Wall Street banks, instructed traders to exit short positions, including credit default swaps...
He explained that the bank's ability to see around the subprime model in 2007 had earned it a fortune. Now the right move was the same -- think outside the quant box.
It didn't work. Risk management was on autopilot.
Perhaps. And perhaps since the market was burning, the time to buy insurance had already passed.
Renaissance has a concept known as the "second forty hours". Employees are each allotted forty hours to work on their assigned duties -- programming, researching markets, building out the computer system. Then, during the second forty hours hours, they're allowed to venture into nearly any area of the fund and experiment. The freedom to do so -- insiders say there are no walled-off segments of the fund to employees -- allows for the chance for breakthroughs that keep Medallion's creative juices flowing.