Thoughts on Central Risk Books & Market Access Rules

Central Risk Books (CRBs) Are Coming of Age

An article in Financial News goes into some depth examining CRBs, which is a favorite topic of mine since I built the first version of a CRB at Salomon Smith Barney over 15 years ago. This article, which delves deeply into the trader view of CRBs, makes a few important points:

  • CRBs are becoming a necessity for trading firms to compete
  • CRBs are viewed by traders with suspicion & gaming is a serious concern
  • CRBs can potentially obfuscate proprietary trading by providing a justification for hedging for trades.

On the first point, all I can say is:its about time! When we built the 1.0 version of the CRB at Salomon Smith Barney, we were limited by the internal politics between my quantitative team and the cash traders, to handling the positions that resulted from retail orders only. As this article reports, however, the concept has gained traction over the past 15 years, which is welcome news for the industry, as liquidity is at a premium. While it is not a trivial undertaking, a CRB that handles the full portfolio of risk taken by all the individual trading desks is the optimal method for minimizing trading costs, managing risk, and, therefore, providing liquidity.

The second point, suspicion and gaming, has always been the most difficult to overcome. Proper architecture of a CRB includes a couple of important design points. First, traders must have well-defined rules for keeping positions and such activities must include proper attribution and analysis. A CRB which did not allow traders to express opinions on future price direction would sacrifice a major source of profits. Such decisions, however, must be attributed to the decision-maker and analyzed to ensure that those traders add value over time. In addition, the method of transferring risk to the CRB must be clearly defined and not subject to manipulation. Systems that transfer positions without regard to estimated transaction costs, for example, are problematic as they are open to being gamed. To explain this, consider a trader that facilitates a client and buys a large position. Without a transfer mechanism that is sensitive to the size of positions, however, the trader would be incentivized to expand the position in the market before the transfer, as that will push the price higher. The result would be unrealized profit for the trader, while the CRB would inherit a more expensive position to liquidate.

The third point, that CRBs can obfuscate proprietary trading is contentious. I will admit that it is a possibility that some firms will allow the risk managers in charge of the CRB to have significant discretion over the hedge instruments to use, which could encourage bets on specific instruments. That said, a well-designed CRB, would control most macro factor risks that traders would be tempted to use for establishing a proprietary position. (This means that such positions would need to be explicit and segregated to avoid the CRB from effectively cancelling out the bet by hedging.) It is important to note, however, that if Volker was loosened, then the banks with the most quantitative and technologically adept CRBs will be in pole position to deploy their capital most effectively,

Market Access Fines; a Reminder of Why the Rule Was Created

Traders Magazine reported over $4.5 million in fines to four brokersfor market access violations. This rule was implemented in order to ensure that all firms who trade in the US equity market obey the key rules of the market, even if they are not regulated brokers. It also is meant to ensure that clients of agency brokers dont trade too much to risk the creditworthiness of the settlement system in the case their trades go bad. The rule was necessary because of the evolution of Direct Market Access (DMA) platforms that enabled both hedge funds and proprietary trading firms to trade without the intervention of a broker dealer. I point this out, not to rub salt into the wounds of the identified firms, but rather to explain that much of the activity that gets labelled predatory trading is done via these channels. While proprietary trading has been done since the late 1800s, the technology available to any trader who wants to build a strategy has advanced significantly. In previous eras, proprietary trading firms needed to set up a brokerage account at member firms and trade through a human intermediary, or the system of the broker itself, which both had significant compliance controls and procedures. Today, however, with the hyper-competitive market for providing DMA platforms, both the cost of trading and the level of scrutiny, pre-trade, have been lowered in order to deliver latencies of a couple of microseconds or lower.

In addition to creating surveillance issues for FINRA, particularly when trading firms use multiple broker DMA platforms, it also explains why information leakage is such a serious consideration for institutions. With more trading firms able to operate at lower costs, including the ability to build models that scrutinize real-time market data, institutional trades that provide clues about their existence are more likely to discovered than ever before. This is why HFT, which includes both market-makers and the aggregation of all these proprietary trading firms, is blamed by institutions for increasing their trading costs. It is not that proprietary trading is a new thing, but rather that the technology advancement has enabled proprietary traders to exploit sloppy trading more than ever.