Thread regarding Goldman Sachs layoffs

Regulators need to put more control and governance in sell side algo trading desk, in particular centralized risk desk. “Central risk desk” are often being used by sell side circumventing from Volcker’s restriction on prop trading. Many sell side central risk desks are their former prop trading desk. For example, BAML central risk book was its prop trading desk QSA. Many people in this desk were prop trading professionals.These are people who know all about regulation and its circumvention.

There shouldn’t exist shady principal alpha trading strategies creating from client data. Central risk book desk need to explain why they are holding overnight position.

Central risk book trading models need to be robust and throughly validated. Traders shall not be allowed tweak parameters. If necessary, trader or other human being shall only have minimal intervention in algo.

Central risk book desk need to document detailed steps and assumption of their models to make sure it doesn’t use client data for principal alpha trading.

Inappropriate use of client data can cause criminal prosecutions. Many actions of central risk book is in grey area now. The so called “facilitation” and “hedge” in central risk book are very similar to prop trading. Regulators need to restrict such actions. “Trading” of central risk book often associate with illicit market activity and abuse of non public client flow information. Central risk book shall only focus risk internalization rather than becoming a platform for alpha principal trading using sensitive and confidential client information. It is not a place for prop trading. Many principal trading in sell side central risk book are causing hidden underestimate risk of market stability. In summary, regulators need to carefully inspect new desks of systematic important banks. In particular, regulators shall make sure central risk book is not used for irresponsible risk taking.

Did you read this Bloomberg article?

Deutsche Bank Said to Lose Money on Risk-Management Trade

U.S. equity central risk book said hurt by algo performance

Book size once topped $2 billion; trader moved to other role

But DB’s not the worst. The bigger issue is central risk book,given its prop trading legacy,may hurt benefit of clients. Many central risk book were converted from its prop trading desk. For example, BAML’s CRB desk was its prop trading desk QSA.This team is poorly managed. Most of bank’s central risk book lost more than 20% YTD. The benefit of CRB is marginal. Yet the reputation damage it created is significant. Central risk book team is lazy. No new models or major model update since 2015. Current model come with many flaws,including some severe issues like principal orders took place while customers gave open orders in the same name (potentially front running) .It also misrepresenting natural liquidity when sending IOIs. Issues in CRB’s existing model caused significant info leakage for clients. Not only trading and quant strategies aren’t profitable,it is not entirely legal.Central risk book’s many activities are in grey area. CRB’s trading system isn’t robust neither. Client order info is widely visible in CRB including some data they suppose to secure safely.

Buy side and regulators need to work together fight sell side misconduct in central risk book. Many of sell side central risk books are secretly doing prop trading. For example, BAML’s central risk book was its prop trading desk QSA. Many people in this desk were prop trading professionals.These are people who know all about regulation and its circumvention. Even worse, client flow data is not safe in central risk book. In some sell side firm,central risk book has real-time visibility of client order details. Trade execution data stored in relation to IOIs may also be used for finding alpha to trade against client. As the nexus of client activity, the desks have a high level of visibility across the market. Some firm’s central risk book team is larger than any of its competitors, which is suspicious. This team talks about alpha trading everyday. They argue that they need the data for purely quant analytical purposes (maybe in a "strategy group"). Desk head instructed team to use client data to backtest and develop alpha trading strategies. Some of central risk book’s action is fraudulent. For example,sell side central risk book misrepresent “natural” liquidity when sending out IOI and it increased info leakage for customer. In low touch child order facilitation, some firm’s central risk book used sensitive and confidential client order fill level percentage done information as alpha signal without any disclosure to client. Clients have no idea that this is going on. With client order data, CRB team found several trading signals to cherry picking client orders in low touch facilitation. CRB’s hedging activities caused CRB frequently trade ahead of customers.The so called “hedge” often cause principal orders took place while customers gave open orders in the same name. Because central risk desks are automated trading teams, it's possible for them to analyze the client data they have access to and to incorporate the signals they find there into their trading models. Trader and the large “quant trading strategists” team found a new way to trading using client data by tweak alpha parameters they found from client data. CRB in sell side often advertised client can access premium liquidity like retail flow or internal derivatives hedging flow. However, clients used to have access through firm’s ATS with better price. Sell side central risk book shall not deprive customers. It's time both buy side and regulator wake up to the dangers. For the sake of stability and transparency in financial markets, central risk books need to be very heavily regulated. Central risk desks are supposed to be about risk internalization and monitoring and as such they don't need so many traders and trading "strategists". The more traders and "strategists" they employ, the greater the likelihood that these desks will simply become a foil for prop trading and the inappropriate use of client data.

central risk book try phishing through IOI

1. Many banks like to fake “Natural IOI” to mislead clients. You need to be careful when bank said it is “natural”.

2. House natural IOI, where bank use its own money to trade with you, can cause data leak and unexpected market impact.

3. Be aware with bank’s IOI “fishing”. They send IOI but not necessarily commit to it. Or only commit a small size of IOI. The goal is to detect your inventory so they can front running.

4. Some brokers just send spam IOIs to confuse you.

Deutsche Bank Said to Lose Money on Risk-Management Trades

-U.S. equity central risk book said hurt by algo performance

-Book size once topped $2 billion; trader moved to other role

https://www.bloomberg.com/amp/news/articles/2018-11-21/deutsche-bank-said-to-lose-money-on-trades-meant-to-improve-risk