Performance-weighted ensembles of random forests for predicting price impact

For any large player in financial markets, the impact of their trading activity represents a substantial proportion of transaction costs. This paper proposes a novel machine learning algorithm for predicting the price impact of order book events. Specifically, we introduce a prediction system based on ensembles of random forests (RFs). The system is trained and tested on depth-of-book data from the BATS and Chi-X exchanges and performance is benchmarked using ensembles of other popular regression algorithms including: linear regression, neural networks and support vector regression. The results show that recency-weighted ensembles of RFs produce over 15% greater prediction accuracy on out-of-sample data, for 5 out of 6 timeframes studied, compared with all benchmarks. Feature importance ranking is used to explore the significance of various market features on the price impact, finding them to be highly variable through time. Finally, a novel procedure for extracting the directional effects of features is proposed and used to explore the features most dominant in the price formation process.     

Dynamic liquidation under market impact

The optimal liquidation problem with transaction costs, which includes a positive fixed cost, and market impact costs, is studied in this paper as a constrained stochastic optimal control problem. We assume that trading is instantaneous and the dynamics of the stock to be liquidated follows a geometric Brownian motion. The solution to the impulse control problem is computed at each time step by solving a linear partial differential equation and a maximization problem. In contrast to results obtained from the static formulation of Almgren and Chriss [J. Risk, 2000 Almgren, R and Chriss, N. 2000. Optimal execution of portfolio transactions. J. Risk, 3: 5–39. [Crossref] , [Google Scholar], 3, 5–39], when risk is not considered, the optimal liquidation strategy from our stochastic control formulation depends on temporary market impact cost and permanent market impact cost parameters. In addition, our computational results indicate the following properties of the optimal execution strategy from the stochastic control formulation. Due to the existence of a no-transaction region, it may not be optimal for some individuals to sell their assets on some trading dates. As the value of the permanent market impact parameter increases, the expected optimal amount liquidated at the terminal time increases. As the value of the quadratic temporary impact cost parameter increases, the expected optimal amount liquidated at trading times tends to be uniform, and the no-transaction region shrinks. In the presence of quadratic temporary market impact costs, in contrast to optimal strategies that result from fixed and/or proportional transaction costs alone, portfolios in the selling region are neither re-balanced into the no-transaction region nor into the sell and no-transaction interface.     

Round-the-clock market efficiency and home bias: Evidence from the international Japanese government bonds futures markets

This paper examines how information is processed between almost identical international futures markets: London (LIFFE) and Tokyo (TSE) JGB futures. In these markets, variations in open-to-open changes are virtually the same as those of close-to-close changes, suggesting that information is transmitted efficiently across markets with small opening pricing errors. The overall results confirm market efficiency around the clock, yet the intraday U-shaped patterns in volume/volatility of the London JGB futures suggest home bias in international investments, indicating a less global view of trading than expected. Specifically, at the LIFFE open, London investors rush to rebalance portfolios instead of doing so at the TSE close, which is only one hour before the LIFFE opens.     

Determinants of bid and ask quotes and implications for the cost of trading

Financial transaction costs are time varying. This paper proposes a model that relates transaction cost to characteristics of order flow. We obtain qualitatively consistent model results for different stocks and across different time periods. We find that an unusual excess of buyers (sellers) relative to sellers (buyers) tends to increase the ask (bid) price. Hence, the ask and bid components of spread change asymmetrically about the efficient price. For a fixed order imbalance surprise these effects are muted when unanticipated total volume is high. Unexpected high volatility in the transaction price process tends to widen the spread symmetrically about the efficient price. Our findings are consistent with predications from market microstructure theory that the cost of market making should depend on both the risk of trading with better-informed traders and inventory risk. We also find that order flow surprises have a significant impact on the efficient price and can also explain a substantial amount of persistence in the volatility of the efficient price. This dependence does not violate the efficient market hypothesis since the surprises, by definition, are not predictable.     

Optimal high-frequency trading with limit and market orders

We propose a framework for studying optimal market-making policies in a limit order book (LOB). The bid–ask spread of the LOB is modeled by a tick-valued continuous-time Markov chain. We consider a small agent who continuously submits limit buy/sell orders at best bid/ask quotes, and may also set limit orders at best bid (resp. ask) plus (resp. minus) a tick for obtaining execution order priority, which is a crucial issue in high-frequency trading. The agent faces an execution risk since her limit orders are executed only when they meet counterpart market orders. She is also subject to inventory risk due to price volatility when holding the risky asset. The agent can then also choose to trade with market orders, and therefore obtain immediate execution, but at a less favorable price. The objective of the market maker is to maximize her expected utility from revenue over a short-term horizon by a trade-off between limit and market orders, while controlling her inventory position. This is formulated as a mixed regime switching regular/impulse control problem that we characterize in terms of a quasi-variational system by dynamic programming methods. Calibration procedures are derived for estimating the transition matrix and intensity parameters for the spread and for Cox processes modelling the execution of limit orders. We provide an explicit backward splitting scheme for solving the problem and show how it can be reduced to a system of simple equations involving only the inventory and spread variables. Several computational tests are performed both on simulated and real data, and illustrate the impact and profit when considering execution priority in limit orders and market orders.     

Testing market efficiency with the pricing kernel

Market efficiency and the pricing kernel are closely related. A non-monotonic decreasing pricing kernel implies the existence of a trading strategy in contingent claims that stochastically dominates a direct investment in the market. Moreover, a market is assumed to be efficient only if no dominating strategies exist. Empirically, many studies of the pricing kernel find non-monotonicity, apparently ruling out market efficiency. However, these results are often unreliable, because the pricing measures of the pricing kernel are estimated using differing filtration sets. We show this effect both theoretically and empirically, and we discuss recent approaches in the literature for achieving more reliable estimates of the pricing kernel, potentially leading to better tests of market efficiency.     

Mechanics of good trade execution in the framework of linear temporary market impact

We define the concept of good trade execution and we construct explicit adapted good trade execution strategies in the framework of linear temporary market impact. Good trade execution strategies are dynamic, in the sense that they react to the actual realisation of the traded asset price path over the trading period; this is paramount in volatile regimes, where price trajectories can considerably deviate from their expected value. Remarkably, however, the implementation of our strategies does not require the full specification of an SDE evolution for the traded asset price, making them robust across different models. Moreover, rather than minimising the expected trading cost, good trade execution strategies minimise trading costs in a pathwise sense, a point of view not yet considered in the literature. The mathematical apparatus for such a pathwise minimisation hinges on certain random Young differential equations that correspond to the Euler–Lagrange equations of the classical Calculus of Variations. These Young differential equations characterise our good trade execution strategies in terms of an initial value problem that allows for easy implementations.     

On the real estate market efficiency

This reseach reexamines the efficiency hypothesis of the real estate market using monthly data and the vector autoregressive (VAR) modelling technique. The tests focus on the causal linkage between real estate returns and a number of relevant financial and economic variables. An eight-by-eight VAR model is estimated using the FPE and the specific gravity criteria, in conjunction with an extensive series of specification tests. The empirical results distilled from system estimations suggest that the real estate market is efficient with respect to available information on the industrial production, the risk premia, the term structure of interest rates, and the monetary base. Movements in these variables are quickly and fully utilized by market agents, perhaps owing to the intensity with which their relationship with stock returns has been discussed in the literature and the popular media. However, the results also suggest the presence of a significant lagged relationship between real estate returns and fiscal policy moves, even when the paths through other potential determinants of these returns are taken into account. Of course, our finding that the fiscal policy measure is useful in predicting stock returns does not necessarily imply that the real estate market is inefficient. At a minimum, inefficiency is revealed only if a careful analysis of the budgetary process can help design a profitable (exploitable) trading strategy.     

A change in market microstructure: the switch to electronic screen trading on the New Zealand stock exchange

The New Zealand Stock Exchange (NZSE) switched from open outcry trading to an electronic screen trading system on June 24, 1991. The change was made by the members of the exchange to improve the trading system and to reduce costs. This paper investigates empirically whether improvement was achieved through a reduction in transaction costs. The tests and results focus on order-flow migration to the exchange from alternative execution locations and changes in bid-ask spreads. On balance, we conclude that transaction costs have declined.     

Who moves the stock market in an emerging country – Institutional or retail investors?

By analyzing the dynamic behavior of institutional and retail investors in the Indonesia Stock Exchange using their completed transactions (comprising over 250 million observations), this study highlights that their trading strategies and behavior, in which institutions play a more important role than individuals in the market, are indeed different. Specifically, past trading activities by individual (institutional) investors have significantly affected the current trading behaviors and strategies of individual investors (both investor types). Furthermore, retail (institutional) investors are most likely to perform contrarian (momentum) strategies and trade frequently (infrequently) with small (large) amounts of money and short (long) holding periods.     

Very fast money: High-frequency trading on the NASDAQ

This paper provides evidence regarding high-frequency trader (HFT) trading performance, trading costs, and effects on market efficiency using a sample of NASDAQ trades and quotes that directly identifies HFT participation. I find that HFTs engage in successful intra-day market timing, spreads are wider when HFTs provide liquidity and tighter when HFTs take liquidity, and prices incorporate information from order flow and market-wide returns more efficiently on days when HFT participation is high.     

A fully consistent,minimal model for non-linear market impact

We propose a minimal theory of non-linear price impact based on the fact that the (latent) order book is locally linear, as suggested by reaction–diffusion models and general arguments. Our framework allows one to compute the average price trajectory in the presence of a meta-order that consistently generalizes previously proposed propagator models. We account for the universally observed square-root impact law, and predict non-trivial trajectories when trading is interrupted or reversed. We prove that our framework is free of price manipulation and that prices can be made diffusive (albeit with a generic short-term mean-reverting contribution). Our model suggests that prices can be decomposed into a transient ‘mechanical’ impact component and a permanent ‘informational’ component.     

不完全竞争、交易效率和证券风险

本文提出了一种崭新的理论观点———不完全竞争的市场结构是证券风险的一个重要来源,并利用市场数据进行了实证分析。此外,本文还发现不完全竞争的市场结构会通过影响买卖双方的交易策略而降低市场的交易效率。     

An empirical examination of the impact of market microstructure changes on the determinants of option bid-ask spreads

This paper examines the determinants of bid-ask spreads in the Australian Options Market before and after it switched from a quote-driven floor-traded market to an order-driven screen-traded market. This study reports that both put and call option bid-ask spreads are positively related to the option's value, its remaining term-to-maturity, its absolute hedge ratio and the volatility of returns from the underlying asset and negatively related to the level of trading activity in that option series. The study also reports that spreads are generally less when market makers are obliged to maintain continuous quotes in the market. The paper also finds that following the change in trading regime, both call and put option spreads became more sensitive to the absolute value of the option's delta. This finding is consistent with previous theoretical and empirical work from equities markets that has suggested that a switch to an electronic trading regime results in an increase in the adverse selection component of the bid-ask spread. There is also some limited evidence that suggests that the switch to electronic trading resulted in call option spreads being less sensitive to the return volatility of the underlying asset but more sensitive to the option's price.     

A brief history of market efficiency

Every finance professional employs the concept of market efficiency. The theory, evidence and counter-evidence focus on a couple of dozen highly influential articles published during the twentieth century. We summarise the origins of and interlinkages between these contributions to the history of finance.     

Evaluation of the adaptive market hypothesis as an evolutionary perspective on market efficiency: Evidence from the Tehran stock exchange

This study evaluates the existence of the adaptive market hypothesis (AMH) as an evolutionary alternative to the efficient market hypothesis (EMH) by applying daily returns on the TEPIX index in the Tehran stock exchange (TSE) in Iran. The data span of daily returns is from 1999 to 2013. In this paper four different tests in the form of two distinguished classes (linear and nonlinear) have been used to study adaptive behavior of returns. The results that were obtained from linear (automatic variance ratio and automatic portmanteau) and nonlinear (generalized spectral and McLeod–Li) tests represent the oscillatory manner of returns about dependency and independency which corresponds with the adaptive market hypothesis.     

The intra-day performance of market timing strategies and trading systems based on Japanese candlesticks

Abstract

We develop market timing strategies and trading systems to test the intra-day predictive power of Japanese candlesticks at the 5-minute interval on the 30 constituents of the DJIA index. Around a third of the candlestick rules outperform the buy-and-hold strategy at the conservative Bonferroni level. After adjusting for trading costs, however, just a few rules remain profitable. When we correct for data snooping by applying the SSPA test on double-or-out market timing strategies, no single candlestick rule beats the buy-and-hold strategy after transaction costs. We also design fully automated trading systems by combining the best-performing candlestick rules. No evidence of out-performance is found after transaction costs. Although Japanese candlesticks can somewhat predict intra-day returns on large US caps, we show that such predictive power is too limited for active portfolio management to outperform the buy-and-hold strategy when luck, risk, and trading costs are correctly measured.     

(How) has the market become more efficient?

Using a portfolio of Dow Jones Industrial Average index constituents and the index ETF, we document significant intraday deviations from the law of one price. These are especially pronounced at very short time intervals. The extent of deviations is related to volatility, liquidity, and transaction costs of both the index constituents and the ETF. Further, the influence of news arrival, and liquidity (volatility) shocks on the deviations persists for several hours. Finally, we document significant decline (by at least 80%) in the deviations between 1998 and 2010. We find that this decline is largely due to decimalization, the repeal of the uptick rule, and the introduction of automated updating of the NYSE order book. Overall, our findings indicate an increase in operational market efficiency.     

Real-time market microstructure analysis: online transaction cost analysis

Motivated by the practical challenge in monitoring the performance of a large number of algorithmic trading orders, this paper provides a methodology that leads to automatic discovery of causes that lie behind poor trading performance. It also gives theoretical foundations to a generic framework for real-time trading analysis. The common acronym for investigating the causes of bad and good performance of trading is transaction cost analysis Rosenthal [Performance Metrics for Algorithmic Traders, 2009]). Automated algorithms take care of most of the traded flows on electronic markets (more than 70% in the US, 45% in Europe and 35% in Japan in 2012). Academic literature provides different ways to formalize these algorithms and show how optimal they can be from a mean-variance (like in Almgren and Chriss [J. Risk, 2000, 3(2), 5–39]), a stochastic control (e.g. Guéant et al. [Math. Financ. Econ., 2013, 7(4), 477–507]), an impulse control (see Bouchard et al. [SIAM J. Financ. Math., 2011, 2(1), 404–438]) or a statistical learning (as used in Laruelle et al. [Math. Financ. Econ., 2013, 7(3), 359–403]) viewpoint. This paper is agnostic about the way the algorithm has been built and provides a theoretical formalism to identify in real-time the market conditions that influenced its efficiency or inefficiency. For a given set of characteristics describing the market context, selected by a practitioner, we first show how a set of additional derived explanatory factors, called anomaly detectors, can be created for each market order (following for instance Cristianini and Shawe-Taylor [An Introduction to Support Vector Machines and Other Kernel-based Learning Methods, 2000]). We then will present an online methodology to quantify how this extended set of factors, at any given time, predicts (i.e. have influence, in the sense of predictive power or information defined in Basseville and Nikiforov [Detection of Abrupt Changes: Theory and Application, 1993], Shannon [Bell Syst. Tech. J., 1948, 27, 379–423] and Alkoot and Kittler [Pattern Recogn. Lett., 1999, 20(11), 1361–1369]) which of the orders are underperforming while calculating the predictive power of this explanatory factor set. Armed with this information, which we call influence analysis, we intend to empower the order monitoring user to take appropriate action on any affected orders by re-calibrating the trading algorithms working the order through new parameters, pausing their execution or taking over more direct trading control. Also we intend that use of this method can be taken advantage of to automatically adjust their trading action in the post trade analysis of algorithms.     

Quote-based competition and trade execution costs in NYSE-listed stocks

This study examines quotations, order routing, and trade execution costs for seven markets that compete for orders in large-capitalization NYSE-listed stocks. The competitiveness of quote updates from each market varies with measures of the profitability of attracting additional order and with volatility and inventory measures. The probability of a trade executing on each market increases when the market posts competitive quotes. Execution costs for non-NYSE trades when the local market posts competitive (non-competitive) quotes are virtually the same (substantially exceed) costs for matched NYSE trades. Collectively, these results imply a significant degree of quote-based competition for order flow and are consistent with off-NYSE liquidity providers using competitive quotations to signal when they are prepared to give better-than-normal trade executions.