The Cumberland Hotel, Marble Arch, London
Monday 7th April 2008
The definitive conference on quantitative trading, models and technology.
9.00am - 9.45am
Registration, Coffee, Exhibition and Networking
9.45am - 9.50am
Opening remarks from chair
9.50am - 10.30am
Keynote Lecture: Commodity Derivatives and the Impact of
Algorithmic Trading on this Asset Class
10.30am - 11.00am
Building an Algorithmic Trading System and the Application
of Maniford Learning in the FX-market
This talk will give a brief overview of the work of introducing machine
learning intelligence in the Nordea e-markets system, to fa- cilitate
auto-hedging, smart price engine algorithms and proprietary automatic
positioning within the foreign exchange market. Since the work began one
year ago, steady progress has been made and the sys- tem is ready to be
employed in a "semi"-live setting this spring. In this talk we will give a
brief overview of the steps taken so far 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 mar- ket. 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 chal- lenging task. Hence, robust methods for detection and
exploitation of such correlations are of high interest for model trading and
quanti-tative strategies. To this end, we study the application of a
recently 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.
Dr Erik Alpkvist, Analyst,
Nordea Markets Division
11.00am - 11.30am
Teas, Exhibition and Networking
11.30am - 12.00pm
Short Horizon Covariance Forecasting
Asset return covariances at intra-day horizons are known to bebiased towards
zero due to market microstructure effects. Thus, traders who simply scale
their daily covariance forecast to match their trading horizon, are likely
to over-estimate the actual experienced asset dependence. In this paper we
discuss some of the key challengesencountered in intra-day covariance
forecasting. Based on extensive empirical analysis, we make specific
recommendations regarding model design and data sampling.
Mr Roel Oomen, Quantitative Analyst,
Deutsche Bank AG
12.00pm - 12.30pm
Adaptive Arrival Price
Electronic trading of equities and other securities makes heavy use of"arrival price" algorithms, that balance the market impact cost of rapid
execution against the volatility risk of slow execution. In the standard
formulation, mean-variance optimal trading strategies are static: they donot
modify the execution speed in response to price motions observed during
trading. We show that with a more realistic formulation of the mean-variance
tradeoff, with no momentum or mean reversion in the price process,
substantial improvements are possible by using dynamic trading strategies.
We develop a technique for computing optimal dynamic strategies to any
desired degree of precision. The asset price process is observed on a
discrete tree with a arbitrary number of levels. We introduce a novel
dynamic programming technique in which the control variables are not only
the shares traded at each time step, but also the maximum expected cost for
the remainder of the program; the value function is the variance ofthe
remaining program. The resulting adaptive strategies
are"aggressive-in-the-money": they accelerate the execution when the price
moves in the trader's favor, spending parts of the trading gains to reduce
risk. The improvement is larger for large initial positions.
Mr Julian Lorenz, PhD Student,
ETH Zurich
Mr Robert Almgren, Head of Quantitative Strategies,
Bank of America Securities
12.30pm - 12.45pm
Open Forum
12.45pm - 2.00pm
Lunch, Exhibition and Networking
2.00pm - 2.30pm
The Next Generation in Algorithmic Trading
Mr Ali Pichvai,
Quod Financial
2.30pm - 3.00pm
Calibrating Market Impact Model with Exponential Decay
The market impact model of Almgren-Thum-Hauptmann-Li (ATHL) has to some
extent become an industry standard in the area of transaction cost analysis.
We improve upon the baseline specification in several aspects. Firstly, by
calibrating the decay speed of the temporary impact (after Obizhaeva-Wang),
we can both ensure finite (as opposed to infinite in case of ATHL) cost of
instantaneous execution and the property that temporary impact depends not
just on the speed of trading but also on the time it takes to trade
(maintaining a 30% trading rate for 1 minute is substantially cheaper than
doing the same for several hours). Secondly, we introduce spreads into the
model. Thirdly, we argue that temporary market impact is a linear function
of a properly modelled expected trading rate. Our model provides a
substantially better fit to the available transactions data and exhibits
consistent behavior if scaled to smaller and/or faster traded orders.
Dr Alexander Gerko, Quantitative Analyst,
Deutsche Bank AG
3.00pm - 3.30pm
Q & A Session
3.30pm - 4.00pm
Teas, Exhibition and Networking
4.00pm - 4.30pm
Algorithmic Trading: Market Impact Models and Trade
Scheduling
Ms Ekaterina Kochieva,
Brunel University
4.30pm - 5.00pm
Market Analytics: A New Front in the Algo Wars
In the early days of electronic trading, electronic execution channels were
made available to save brokers the trouble of dealing with annoying small
orders, so they could focus on their primary task of executing large orders
on the floor.
Now with the proliferation of algorithms, and markets of makers and takers,
the electronic channels are rapidly becoming the primary venue for execution
of ever larger orders. 10,000 shares looks increasingly like 100 shares, 100
times. Algorithm designers are fighting a multi-front war for an edge in
speed, and real-time analysis. The task is further complicated by an ever
growing population of execution venues.
Exchanges are moving aggressively to provide a friendly environment for this
order flow. They build ever faster trading systems, whose performance is
measured in milliseconds, and offer collocated hosting for major
participants, to further shave delays so small that the speed of light
enters into their calculation.
Executions are not the only type of electronic traffic needed to support
leading edge algo trading. Analytics, calculations that convey information
about markets, are of growing importance. All of the major global stock
exchanges have added real time analytics to their data offerings.
These analytics convey information derived from much higher volume data
flows. The raw information used includes all order book events, all order
flows, including cancels and replacements, and partial fills. Using this
data directly is generally not possible or practical. It is not possible,
since markets do not disclose the micro details of their customers' order
working process, and not practical, since there is far too much data.
Order traffic rates go as high as millions per minute, and are still rising.
There is only so much that can be conveyed over even today's broadband
channels, and not every market participant wants to deal with the
computational task of analyzing that torrent in real time.
This creates a need for a new class of high speed analytics, sent via the
same low latency systems developed for rapid order execution. These
analytics seek to convey information needed by algo participants, without
conveying information that those participants want to hide from each other.
They replace the "market color" and "sense of the floor" familiar to traders
of the past.
The customers for these market analytics fall into two groups: human traders
and computers running algos. Humans read, but the computers act, as rapidly
as they can, on the information in these analytics. These forces result in a
new set of low latency computing machinery being deployed in this latest
front in the Algo Wars.
This paper gives a quick overview of the state of real time market analytics
offered today, using examples from US and European Exchanges. An example
showing the application of low latency analytics to order data shows a
significant and potentially exploitable speed advantage.
Mr David Leinweber, President, Leinweber & Co
5.00pm - 5.15pm
Concluding remarks from chair
5.15pm - 5.20pm
Exhibitors Prize Draw
5.20pm - 6.00pm
Coffee, Exhibition and Networking
Please send all enquiries to Seromanie Bernard at
seromanie@chameleonproductions.co.uk (please mention MoneyScience)
or log onto the event website: www.ChameleonProductions.co.uk for further details and a delegate registration form.
THIS EVENT IS LISTED IN THE MONEYSCIENCE FINANCIAL CONFERENCE CALENDAR.
Featured Product
