Algorithmic trading ... - page 2

New comment
 
I examine the impact of an exogenous trading glitch in a high-frequency market-making firm on institutional trading costs. At the first look, the trading glitch does not appear to affect institutional investors as it leads to dramatic increases in volume without any information content but controlling for various stock- and order-level characteristics, I find that executing a large order on a glitch-affected stock incurs substantially higher costs on the event day. Moreover, the cost increase is persistent up to one week roughly with the same additional cost magnifying the total economic costs. These findings can be interpreted as negative externalities of algorithmic trading which has important policy implications.
 
The concept of Algorithmic Trading emulates via electronic means a brokers core competency of slicing a big order into a multiplicity of smaller orders and of timing these orders to minimize market impact. Based on mathematical models and considering historical and real-time market data, algorithms determine ex ante or continuously the optimum size of the (next) slice and its time of submission to the market. Algorithmic trading models are gaining market share worldwide. As this might impact the order flow on the markets it is self-evident to investigate whether algorithmic trading can be categorized in the traditional way or whether it represents a new category of stylized trader. The paper assesses the upcoming sophisticated trading strategy of algorithmic trading against the background of the traditional categories of stylized traders in the literature, i.e. informed traders, momentum traders and noise traders. As a conclusion, in order to assess the of impact algorithmic trading on financial markets, the set-up of a new simulation model incorporating agents representing the specific properties and the trading behavior of algorithmic trading is proposed.
 
In the past decades, advanced probabilistic methods have had significant impact on the field of finance, both in academia and in the financial industry. Conversely, financial questions have stimulated new research directions in probability. In this survey paper, we review some of these developments and point to some areas that might deserve further investigation. We start by reviewing the basics of arbitrage pricing theory, with special emphasis on incomplete markets and on the different roles played by the 'real-world' probability measure and its equivalent martingale measures. We then focus on the issue of model ambiguity, also called Knightian uncertainty. We present two case studies in which it is possible to deal with Knightian uncertainty in mathematical terms. The first case study concerns the hedging of derivatives, such as variance swaps, in a strictly path wise sense. The second one deals with capital requirements and preferences specified by convex and coherent risk measures. In the final two sections we discuss mathematical issues arising from the dramatic increase of algorithmic trading in modern financial markets.
 
We study the effect of algorithmic trading (AT) intensity on equity market liquidity, short-term volatility, and informational efficiency between 2001 and 2011 in 42 equity markets around the world. On average, AT improves liquidity and informational efficiency but increases volatility. We can attribute the AT-related increase in volatility neither to more “good” volatility that would arise from faster price discovery nor to algorithmic traders’ inclination to enter the market when volatility is high. On the contrary, these volatility-seeking traders are associated with declines in market quality. Our results are surprisingly consistent across markets and thus across a wide range of AT practices. But results vary in the cross-section of stocks. In contrast to the average effect, greater AT intensity reduces liquidity and worsens the volatility increase in the smallest tercile of stocks. Finally, AT becomes less beneficial when market making is difficult.
 
This paper will give a brief overview of the work of introducing machine learning intelligence in the Kineta e-markets system, to facilitate auto-hedging, smart price engine algorithms and proprietary automatic positioning within the foreign exchange market. In this paper we will give a brief overview of the steps taken in the project. A number of quantitative techniques have been implemented in the system and evaluated. As of late we have investigated the use of manifold learning; a class of geometrically motivated nonlinear data mining methods, to predict movements in the foreign exchange market. Financial time series are often correlated over time; and may contain valuable customer specific proprietary information. In principle, such relationships may be exploited for forecasting. However, they may be noisy, nonlinear and changing over time, making this a challenging task. Hence, robust methods for detection and exploitation of such correlations are of high interest for model trading and quantitative strategies. To this end, we study the application of a proposed method for nonlinear regression on manifolds. The approach involves dimensionality reduction through Laplacian Eigenmaps and optimization of cross-covariance operators in the kernel feature space induced by the normalized graph Laplacian.
 
This paper documents a stark periodicity in intraday volume and in the number of trades. We find activity in both variables spikes by about 20% at regular intervals of 5 or 10 minutes throughout the trading day. We argue that this activity is the result of algorithmic trading influenced by human traders/programmers’ behavioral bias to transact on round time marks. An alternative explanation, that algorithms choose to concentrate their trades in time to take advantage of lower costs or to protect themselves from better informed traders, is not supported.
 
Effective prediction of financial asset prices has become a challenge in the present day volatile world. The use of mathematics have become very extensive in the financial world, most of the mathematical models concentrates on the market data rather than the behavior of the market from which the data has been generated. An attempt has been made for the first time to model the prediction of asset prices based on both the market data and the behavior of the market participants. The participants in the financial markets behave differently from each other, these behavioral differences can be attributed to the participants understating or/and his perception about the market. Each investor has his own perception about the market and he feel it is close to reality, but truly speaking it is not so. Each participant has his own impact on the market and the reality is the aggregation of each participant’s perception. The impact of the investor’s behavior has been modeled in the present quantitative behavioral approach by dividing the participants into broad categories based on their trading behavior. To model the participant’s impact first one should predict the proportion of participants in each category. Most of the times, finding the exact number of participants in each category is not easily available from the market data, so an evolutionary based swarm intelligence model has been adopted in the present framework to find the proportion of the participants in each category. Finally the whole methodology has been applied to gold asset class (because gold is an international asset with increasing volatility these days) to validate the present method. The model is tested rigorously using different time varying samples to validate the present methodology; some interesting results have been obtained from the present study. The back testing results prove that the model presented in this paper is very effective in predicting the prices close to reality. The present frame work is very generic and can be applied to any financial asset class to estimate the returns close to reality.
 
We examine the Foreign Exchange (FX) spot price spreads with and without Last Look on the transaction. We assume that brokers are risk-neutral and they quote spreads so that losses to latency arbitrageurs (LAs) are recovered from other traders in the FX market. These losses are reduced if the broker can reject, ex-post, loss-making trades by enforcing the Last Look option which is a feature of some trading venues in FX markets. For a given rejection threshold the risk-neutral broker quotes a spread to the market so that her expected profits are zero. When there is only one venue, we find that the Last Look option reduces quoted spreads. If there are two venues we show that the market reaches an equilibrium where traders have no incentive to migrate. The equilibrium can be reached with both venues coexisting, or with only one venue surviving. Moreover, when one venue enforces Last Look and the other one does not, counterintuitively, it may be the case that the Last Look venue quotes larger spreads.
 
We consider the infinite time-horizon optimal basket portfolio liquidation problem for a von Neumann-Morgenstern investor in a multi-asset extension of the liquidity model of Almgren (2003) with cross-asset impact. Using a stochastic control approach, we establish a "separation theorem": the sequence of portfolios held during an optimal liquidation depends only on the (co-)variance and (cross-asset) market impact of the assets, while the speed with which these portfolios are attained depends only on the utility function of the trader. We derive partial differential equations for both the sequence of attained portfolios and the trading speed.
 
I present evidence that a moving average (MA) trading strategy third order stochastically dominates buying and holding the underlying asset in a mean-variance-skewness sense using monthly returns of value-weighted decile portfolios sorted by market size, book-to-market cash-flow-to-price, earnings-to-price, dividend-price, short-term reversal, medium-term momentum, long-term reversal and industry. The abnormal returns are largely insensitive to the four Carhart (1997) factors and produce economically and statistically significant alphas of between 10% and 15% per year after transaction costs. This performance is robust to different lags of the moving average and in subperiods while investor sentiment, liquidity risks, business cycles, up and down markets, and the default spread cannot fully account for its performance. The MA strategy works just as well with randomly generated returns and bootstrapped returns. I also report evidence regarding the profitability of the MA strategy in seven international stock markets. The performance of the MA strategies also holds for more than 18,000 individual stocks from the CRSP database. The substantial market timing ability of the MA strategy appears to be the main driver of the abnormal returns. The returns to the MA strategy resemble the returns of an imperfect at-the-money protective put strategy relative to the underlying portfolio. Furthermore, combining several MA strategies into a value/equal-weighted portfolio of MA strategies performs even better and represents a unified framework for security selection and market timing.
123456
New comment