Testing the effect of technical analysis on market quality and order book dynamics

ABSTRACT

We find empirical support for the theoretical finding in agent-based models of limit order book markets that the effect of technical trading on market quality is not positive. When signals occur, technical traders lower liquidity as proxied by the relative spread, the effective spread, the realized spread, the dispersion and the slope in the order book. Technical trading is also found to be accompanied by rising volatility. There is overall strong empirical support against the hypothesis that technical trading has no effect on order book dynamics.

KEYWORDS:

• Agent-based model
• liquidity
• technical analysis
• limit order market

JEL CLASSIFICATION:

• G14
• C50

Acknowledgement

We are grateful to NYSE Euronext in Paris for providing the data. We also thank Alexis Cellier, Catherine D’Hondt, Rudy De Winne, and Christophe Majois for their early work on the dataset, without forgetting various seminar participants for their useful suggestions. Any remaining errors are the responsibility of the authors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ In preliminary analyses, we have simulated the process on a set of arbitrary thresholds. The results remain unchanged if we consider 50% or 60%.

² These measures are computed with the sum over each interval.

³ By accumulated, we mean the sum of the quantities outstanding at that limit and the sum of all quantities outstanding at each better quote.

⁴ Hendershott and Moulton (2011) also examined two other price efficiency proxies: the 5/30-min variance ratio and the pricing error. We argue against the use of the first one since the proxy does not bring new elements in the analysis. The latter is based on Hasbrouck (1993) which uses a vector autoregressive representation (VAR) to evaluate market quality and price efficiency. This model grounds on quote revisions after trades and on the informational content of trades. We choose not to use this proxy for several reasons that are also invoked by O’Hara and Ye (2011).

⁵ To save space, we only report the dummies related to the three technical indicators, with separate buy and sell signals.