Trading activity and price impact in parallel markets: SETS vs. off-book market at the London Stock Exchange

Abstract

We study empirically the trading activity in the electronic on-book segment and in the dealership off-book segment of the London Stock Exchange, investigating separately the trading of active market members and of other market participants who are non-members. We find that (i) the volume distribution of off-book transactions has a significantly fatter tail than that of on-book transactions, (ii) groups of members and non-members can be classified in categories according to their trading profile, (iii) there is a strong anticorrelation between the daily inventory variation of a market member due to on-book market transactions and an inventory variation due to off-book market transactions with non-members, and (iv) the autocorrelation of the sign of the orders of non-members in the off-book market is slowly decaying. We also analyse the on-book price impact function over time, both for positive and negative lags, of the electronic trades and of the off-book trades. The unconditional impact curves are very different for the electronic trades and the off-book trades. Moreover, there is a small dependence of the impact on the volume for the on-book electronic trades, while the shape and magnitude of the impact function of off-book transactions strongly depend on volume.

Keywords:

• Market microstructure
• Market segmentation
• Order flow
• Price impact
• Brokerage data

Acknowledgements

The authors acknowledge financial support from the MIUR PRIN project 2007TKLTSR ‘Indagine di fatti stilizzati e delle strategie risultanti di agenti e istituzioni osservate in mercati finanziari reali ed artificiali’.

Notes

¶http://www.sec.gov/rules/concept/2010/34-61358.pdf

⊥http://www.cesr-eu.org/index.php?page=consultation_details&id=158

†In this last case the member trades a security on his own behalf and enters into a CDF with the clients, meaning that no shares are exchanged between themselves, but the client assumes the risk associated with handling the shares.

†It is worth noting that the inventory variation time series in the on-book and off-book market segments are computed in a slightly different way. The daily on-book inventory variation is obtained by considering transactions that occur during market hours, whereas the daily off-book inventory variation include transactions that might occur after market hours, although of course before the next opening of the market.

†When we sort according to the cross-correlation values between the price return and the off-book inventory variation we obtain a similar structure.

†Note that while in the on-book market one order can generate more than one trade, in the off-book market we observe trades and we infer the order that generated it.