Home » Market Structure » Low Latency Trading Survey (introduction)

Low Latency Trading Survey (introduction)

Introduction

Is there a profitable instantaneous arbitrage strategy employing a low latency fiber between two major trading centers? In 2001, Lo and MacKinlay popularized the notion that US Equity Prices do not follow random walks in their book, A Non-Random Walk Down Wall Street. Several successful proprietary algorithmic trading desks have established track records of stable profitability based on their knowledge of predictable stock price patterns and market microstructure. Given this commercial success, are these quantitative insights and technologies applicable to other markets and other trading strategies? The use of Barksdale’s recently installed straight run of dark fiber from Carteret, NJ to Chicago South Loop in Low Latency proprietary trading is contemplated here.

Conclusions

Low Latency proprietary trading between NJ and Chicago is plausibly profitable, perhaps even without Barksdale’s fiber. Given the current Dodd-Frank Act regulatory reform implementation, the near-term market making revenue opportunities are numerous. The returns distribution may be skewed more towards Fixed Income than Equities, however, given the relative market sizes and the focus of the US regulatory changes. The OTC Swaps market shifts and displacements expected with the implementation of Dodd-Frank Title VII will be massive and unprecedented. For simplicity and applicability to existing platforms, we have focused this analysis on a Low Latency equity basis market making/arbitrage business. Extension to Fixed Income market making and market microstructure appears plausible as well.

Given the steady profitability of High Frequency trading, the trade will become crowded, presumably sooner rather than later. Diversifying regional and market exposure by investing in Low Latency trading makes sense, but nevertheless does not by itself impose an insurmountable barrier to competition. In a market microstructure trading strategy commonly described as an “arms race,” the incremental infrastructure dollar spent now, principally secures time to spend more dollars on infrastructure later. Expertise in pre-trade algorithms, code optimization, and high performance computational hardware are crucial for competing and thriving in the “arms race.” Diversifying market making and arbitrage trading strategies across Fixed Income and Equities are crucial goals for finessing the “arms race” and establishing a long term measurable risk reward balance.

References

Ars Technica 2008 Nehalem Review

AT&T US Network Latency

Avellaneda & Lee, Statistical Arbitrage in the U.S. Equities Market

Bookstaber, Rich, The Arms Race in High Frequency Trading

Burghardt et.al., The Treasury Bond Basis

Carney, Will Dodd-Frank Help High Frequency Traders Crash the Bond Market Too?

Carney, Yes, High Frequency Traders Will Take Over Swap Trading

Choudhry, The Credit Default Swap Basis

Hasbrouck and Saar, Low Latency Trading

IBM MASS Library

Intel Vector Math Library performance