Forum on News Analytics  applied to Trading, Fund Management  and Risk Control
GAUTAM MITRA <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
<ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul><ul><li>Background and ove...
DAN DI BARTOLOMEO <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
Incorporation of Quantified News into Portfolio Risk Assessment Dan diBartolomeo Northfield and Brunel University London, ...
Motivation for the Short Term  Risk Forecasts <ul><li>Risk models for asset management (as distinct from trading operation...
A Short Chronology <ul><li>A call from Blair Hull in 1996 </li></ul><ul><li>diBartolomeo and Warrick (2005) in Linear Fact...
Simple Approach to Short Term Modeling <ul><li>The usual answer: </li></ul><ul><ul><li>Increase the frequency of observati...
What’s the Problem with High Frequency Data? <ul><li>Financial markets are driven by the arrival of information in the for...
Some Surprising Things Appear Anticipated <ul><li>Lets look at a precipitous decline in the implied volatility of options ...
Investor Response  to Information Flow  <ul><li>Several papers have examined the relative market response to “news” and “a...
Our Approach is Different <ul><li>Continue to use the existing risk models that are estimated from low frequency return ob...
One Form of Working with “External Information” <ul><li>Risk estimates in our short term model of US equities have been co...
“ Variety” as External Information <ul><li>Solnik and Roulet (2000) examine the dispersion of country returns as a way of ...
Other Conditioning Information <ul><li>Estimates of volatility based on high/low/open/close information instead of the dis...
What Makes People Buy or Sell a Particular Stock? <ul><li>They WANT to trade the stock </li></ul><ul><ul><li>They believe ...
The Potential for “Have To’s” <ul><li>We can fundamentally evaluate the potential for “have to” trades </li></ul><ul><ul><...
The Potential for “Want To” Trades <ul><li>Investors are responding to information, so just measure variations in the volu...
Incorporating News Flows into Risk Assessments <ul><li>diBartolomeo, Mitra and Mitra (2009) forthcoming in  Quantitative F...
Incorporating Investor Attention <ul><li>Our next step will be to directly measure the degree of investor attention to a s...
Crucial Refinement <ul><li>diBartolomeo and Warrick (2005),  and diBartolomeo, Mitra and Mitra (2009) both assume that the...
Other Differences Between Long and Short Horizon Risk <ul><li>Negative serial correlation </li></ul><ul><ul><li>Daily over...
Conclusions <ul><li>The key to good short term risk assessment is understanding how conditions now are different than they...
ARMANDO GONZALEZ <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
ARUN SONI <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
JAMES CHENERY <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
INCOPRORATING NEWS ANALYSIS INTO TRADING AND INVESTMENT PROCESSES FORUM ON NEWS ANALYTICS November 9, 2009 James Chenery B...
EXPLOITING NEWS CONTENT <ul><li>News is emerging as differentiated, value generating content set </li></ul><ul><ul><li>Qua...
EXPLOITING NEWS CONTENT <ul><li>News flow is a good indicator of volume and volatility </li></ul><ul><li>Pricing movements...
MACHINE READABLE NEWS USE CASES <ul><li>Wolf detection / circuit breaker </li></ul><ul><li>News flow algorithms  </li></ul...
NEWSSCOPE PRODUCT PORTFOLIO <ul><li>NewsScope Archive   </li></ul><ul><ul><li>Historical database of Reuters and select th...
HOW TO MAKE MONEY WITH THE NEWSSCOPE ANALYTICS Buy on good news Sell on bad news Outperform S&P500 by 5000 basis points ov...
Questions?
MARK VREIJLING <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
GANGADHAR DARBHA <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
GURVINDER BRAR <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control  </li></ul>
Exploiting news-flow signals Macquarie Quantitative Research Gurvinder Brar, Christian Davies, Adam Strudwick, Andy Moniz ...
The global presence of Macquarie Quant <ul><li>17 professional staff across the globe </li></ul>Page  Australia (8)  Resea...
Recent publications Page  <ul><li>Monthly ‘ Quantamentals’  report </li></ul><ul><ul><li>Unwrapping value, Oct 08 </li></u...
Page  News-flow research in vogue Source: Macquarie Research, November 2009 <ul><li>Academic literature </li></ul><ul><ul>...
Page  Data vendors QUANTAMENTALS: HAVE I GOT NEWS FOR YOU? MAY ‘09 Source: Macquarie Research, November 2009 <ul><li>Bloom...
Page  Exploiting news flow strategies QUANTAMENTALS: HAVE I GOT NEWS FOR YOU? MAY ‘09 Strategy Performances Information ra...
Page  Important disclosures: Recommendation definitions –  For quarter ending 30 September 2009 Recommendation definitions...
Page  Analyst Certification:   The views expressed in this research accurately reflect the personal views of the analyst(s...
Upcoming SlideShare
Loading in …5
×

Forum On News Analytics

1,808 views

Published on

Forum on News Analytics applied to Trading and Risk Control, 9 November 2009, London

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
1,808
On SlideShare
0
From Embeds
0
Number of Embeds
7
Actions
Shares
0
Downloads
29
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • News is emerging as an exploitable content set to drive alpha and manage event risk. We’ve been in market for about 2 years and while most of the focus is more on black-box trading, regardless of trading frequency, we see many ways to utilize it for advanced human decision support especially with our news analysis solutions. Customers are utilizing machine readable news in several key areas from the obvious speed advantage to managing the scale and scope of their portfolio to creative risk management and loss avoidance strategies. NewsScope represents a robust set of capabilities and data for customers to utilize – from historical data for deep research, algo building and back testing to real-time ultra-low latency feeds for optimal execution, to news analytics where we convert qualitative text into quantitative values making it easier for machines and humans-alike to analyze and comprehend.
  • A number of academic and practitioner research has focused on news and the subsequent market impact. Some of the main conclusions fall into the following areas: News flow is a good indicator of volume and volatility. The more news on a company, the higher the volume traded &amp; the more volatile it is (more about abnormal volume spikes – see Event Indices) Enhanced by relevance, sentiment, and novelty indicators Pricing movements accompanied by news tend to be momentum in nature; those with a lack of news tend to reverse to average trends. Applications for statistical arbitrage, momentum trading, technical analysis, risk, human decision support The market tends to overreact when there is a lot of news on something and under-react when there is a small quantity of news. Reversals as news is “repeated” and syndicated. Longer lead times / position building for more novel stories. For direction and magnitude, find cause:effect relationships Key words – upgrade/downgrade, raises/lowers guidance, misses/exceeds Market sentiment – results vs consensus Author sentiment – tone of news articles
  • Wolf detection / circuit breaker – automatically stop the algorithm when news is published – tends to be more defensive in nature News flow algorithms – because news flow can predict potential volume or volatility spikes, one might switch from VWAP to News VWAP, a more participatory algorithm, which accelerates an order to get a better price in the market. Alpha generating signal – being able to predict returns is key use to many Post Trade Analysis – assist in proving best execution and trader performance Stock screening tool – give me all of the good news and bad news stocks Risk Management – help understand event risk and what kinds of news can impact your portfolio risk profile Monitor for Compliance and market abuse – track insider trading Fundamental research – assist in calculating stock, sector, and market forecasts Trader decision support for confirming or contrarian trading signals
  • Note hosted and deployed solutions, fault tolerant, fully resilient for the best mix of flexibility and performance (as a differentiator vs Ravenpack, of course) and we also provide the market data with common symbology and timestamp making it easier to test and then deploy into production.
  • Note that given the limited amount of time, you can share some more detail and examples after the session, but I like to call this slide: “How to make money with the NewsScope Analytics” Buy on the good news (green); sell on the bad news (red) and outperform the S&amp;P 500 by 5000 basis points in 60 days! (if only it were that easy)
  • Forum On News Analytics

    1. 1. Forum on News Analytics applied to Trading, Fund Management and Risk Control
    2. 2. GAUTAM MITRA <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    3. 3. <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul><ul><li>Background and overview of news analytics in finance </li></ul><ul><li>Sentiment classification </li></ul><ul><li>News and abnormal returns </li></ul><ul><li>News and volatility </li></ul><ul><li>Industry insights, technology, products and services </li></ul><ul><li>Directory of news analytics service providers </li></ul><ul><li>Bibliography </li></ul>
    4. 4. DAN DI BARTOLOMEO <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    5. 5. Incorporation of Quantified News into Portfolio Risk Assessment Dan diBartolomeo Northfield and Brunel University London, 2009
    6. 6. Motivation for the Short Term Risk Forecasts <ul><li>Risk models for asset management (as distinct from trading operations) have traditionally focused on estimating portfolio risk from security covariance over time horizons of a year or more </li></ul><ul><ul><li>Suitable for long term investors such as pension funds </li></ul></ul><ul><li>Investment performance of asset managers is often evaluated over shorter horizons so they are interested in shorter term risk assessment. Hedge funds and other portfolios with high portfolio turnover are even stronger in this preference </li></ul><ul><li>The proliferation of high frequency trading and algorithmic execution methods have created demand for very short horizon risk assessment </li></ul>
    7. 7. A Short Chronology <ul><li>A call from Blair Hull in 1996 </li></ul><ul><li>diBartolomeo and Warrick (2005) in Linear Factor Models in Finance, edited by Satchell and Knight </li></ul><ul><li>“ Short Term Risk from Long Term Models”, Northfield research series, Anish Shah, 2007-2009 </li></ul><ul><li>“ Equity Portfolio Risk Using Market Information and Sentiment” by diBartolomeo, Mitra and Mitra, 2009 </li></ul>
    8. 8. Simple Approach to Short Term Modeling <ul><li>The usual answer: </li></ul><ul><ul><li>Increase the frequency of observations (daily or shorter) </li></ul></ul><ul><ul><li>Use a shorter sample period </li></ul></ul><ul><ul><li>Generally need different factors </li></ul></ul><ul><li>There are serious problems with this approach at the individual security level </li></ul><ul><ul><li>High degree of kurtosis in return distributions (well maybe?) </li></ul></ul><ul><ul><li>Negative serial correlation due to short term reversal effects </li></ul></ul><ul><ul><li>Positive serial correlation on illiquid instruments </li></ul></ul><ul><ul><li>Asynchronous trading across time zones makes correlation estimation very difficult </li></ul></ul><ul><li>Address “shocks” through a GARCH process </li></ul>
    9. 9. What’s the Problem with High Frequency Data? <ul><li>Financial markets are driven by the arrival of information in the form of “news” (truly unanticipated) and the form of “announcements” that are anticipated with respect to time but not with respect to content. </li></ul><ul><li>The time intervals it takes markets to absorb and adjust to new information ranges from minutes to days. Generally much smaller than a month, but up to and often larger than a day. That’s why US markets were closed for a week at September 11 th . </li></ul><ul><li>GARCH models don’t work well on announcements </li></ul><ul><ul><li>Market participants anticipate announcements </li></ul></ul><ul><ul><li>Volume and volatility dry up as investors wait for outcomes </li></ul></ul><ul><ul><li>Reduce volatility into the announcement and boost it after the announcement, so they are wrong twice </li></ul></ul>
    10. 10. Some Surprising Things Appear Anticipated <ul><li>Lets look at a precipitous decline in the implied volatility of options on LUV </li></ul><ul><ul><li>All days in 2001 prior to September 7, average of .45 with a s.d. of .13 </li></ul></ul><ul><ul><li>September 7, LUV implied = .22 </li></ul></ul><ul><ul><li>September 10, LUV implied = .15 </li></ul></ul><ul><ul><li>All days subsequent to September 17, average of .54 with a s.d. of .18 </li></ul></ul><ul><ul><li>September 10 is in bottom 1% of the universe in implied volatility, September 17 is 91 st percentile </li></ul></ul><ul><li>Could this be driven by fundamentals? </li></ul>
    11. 11. Investor Response to Information Flow <ul><li>Several papers have examined the relative market response to “news” and “announcements” </li></ul><ul><ul><li>Ederington and Lee (1996) </li></ul></ul><ul><ul><li>Kwag Shrieves and Wansley(2000) </li></ul></ul><ul><ul><li>Abraham and Taylor (1993) </li></ul></ul><ul><li>Jones, Lamont and Lumsdaine (1998) show a remarkable result for the US bond market </li></ul><ul><ul><li>Total returns for long bonds and Treasury bills are not different if announcement days are removed from the data set </li></ul></ul><ul><li>Brown, Harlow and Tinic (1988) provide a framework for asymmetrical response to “good” and “bad” news </li></ul><ul><ul><li>Good news increases projected cash flows, bad news decreases </li></ul></ul><ul><ul><li>All new information is a “surprise”, decreasing investor confidence and increasing discount rates </li></ul></ul><ul><ul><li>Upward price movements are muted, while downward movements are accentuated </li></ul></ul>
    12. 12. Our Approach is Different <ul><li>Continue to use the existing risk models that are estimated from low frequency return observations </li></ul><ul><li>Use new information that is not part of the risk model to adjust various components of the risk forecast to short-term conditions </li></ul><ul><ul><li>Just ask yourself “How are conditions different now than they were on average during the sample period used for estimation?” </li></ul></ul><ul><li>This approach has multiple benefits </li></ul><ul><ul><li>We sidestep almost all of the statistical complexities that arise with use of high frequency data </li></ul></ul><ul><ul><li>We get to keep the existing factor structure of the model so risk reporting remains familiar and intuitive </li></ul></ul><ul><ul><li>Since our long term and short term forecasts are based on the same factor structure, we can quickly estimate new forecasts for any length time horizon that falls between the two horizons </li></ul></ul><ul><ul><li>Can be applied to any of our existing models </li></ul></ul>
    13. 13. One Form of Working with “External Information” <ul><li>Risk estimates in our short term model of US equities have been conditioned for many years based on analysis of stock option implied volatility </li></ul><ul><ul><li>Every day we look at the implied volatility of options on all US stocks. We keep a 30 day moving average of the ratio of implied volatility to historic volatility </li></ul></ul><ul><ul><li>If the implied volatility/historic ratio jumps because of an information flow to the market (e.g. Bill Gates gets run over by a bus), the specific risk of that stock is adjusted </li></ul></ul><ul><ul><li>If implied volatility ratio of many related stocks changes, the implied changes in factor variance are also made. Risk forecasts change even for stocks on which no options trade </li></ul></ul><ul><ul><li>Requires non-linear optimization process for adjustments </li></ul></ul><ul><ul><li>See Chapter 12, by diBartolomeo and Warrick Linear Factor Models in Finance , Satchell and Knight, editors (2005) </li></ul></ul>
    14. 14. “ Variety” as External Information <ul><li>Solnik and Roulet (2000) examine the dispersion of country returns as a way of estimating correlations between markets </li></ul><ul><li>Lilo, Mantegna, Bouchard and Potters use the term Variety to describe cross-sectional dispersion of stock returns </li></ul><ul><ul><li>They also define the cross-sectional dispersion of CAPM alpha as idiosyncratic variety (noted as v(t)) </li></ul></ul><ul><ul><li>They find that the average correlation between stocks is approximately: </li></ul></ul><ul><li>C(t) = 1 / [1 + (v 2 (t)/r m 2 (t) ] </li></ul><ul><li>diBartolomeo (2000) relates periods of high cross-sectional dispersion to positive serial correlation in stock returns (i.e. momentum strategies working) </li></ul>
    15. 15. Other Conditioning Information <ul><li>Estimates of volatility based on high/low/open/close information instead of the dispersion of returns </li></ul><ul><ul><li>Parkinson, Garman-Klass, Satchell-Wang, etc. </li></ul></ul><ul><li>Yield spreads for different classes of fixed income securities provide an implied default rate and the potential for large negative skew in stock returns </li></ul><ul><li>Implied distribution of asset returns given the implied vols of options on market indices across strike prices </li></ul><ul><li>Direct measures of information flow to investors, and investor attention that can create imbalances between supply and demand for a given stock </li></ul>
    16. 16. What Makes People Buy or Sell a Particular Stock? <ul><li>They WANT to trade the stock </li></ul><ul><ul><li>They believe the information that supports a valid forecast of abnormal future return </li></ul></ul><ul><li>They HAVE to trade the stock </li></ul><ul><ul><li>They are trading to implement a change in asset allocation </li></ul></ul><ul><ul><li>They are trading to implement a cash versus futures arbitrage trade on a stock index </li></ul></ul><ul><ul><li>They are a mutual fund or ETF sponsor responding to investor cash flows in or out of the portfolio </li></ul></ul><ul><ul><li>They are hedge fund that is forced to transact because of a margin call </li></ul></ul><ul><ul><li>They are forced to cover a short position by having the stock called </li></ul></ul>
    17. 17. The Potential for “Have To’s” <ul><li>We can fundamentally evaluate the potential for “have to” trades </li></ul><ul><ul><li>Index arbitrage trades only occur with index constituents and we know the open interest in futures </li></ul></ul><ul><ul><li>Short interest information is published </li></ul></ul><ul><ul><li>We know what big hedge and mutual funds have big positions in particular stocks </li></ul></ul><ul><ul><li>We have somewhat out of date information on full mutual fund holdings and cash flow statistics </li></ul></ul><ul><ul><li>We have fairly up to date information on ETF flows </li></ul></ul>
    18. 18. The Potential for “Want To” Trades <ul><li>Investors are responding to information, so just measure variations in the volume of information about a particular stock over time </li></ul><ul><li>Judge the magnitude of information flow of news text coming over services such as Dow-Jones, Reuters and Bloomberg </li></ul><ul><ul><li>Ravenpack and Thomson Reuters offer real time statistical summaries of the amount and content of text news distributed </li></ul></ul><ul><ul><li>Lexicons of over 2000 popular phrases are used to score the content as “good news” or “bad news” </li></ul></ul><ul><ul><li>Judge investor attention directly by measuring the number of Google and Yahoo searches on ticker symbols </li></ul></ul>
    19. 19. Incorporating News Flows into Risk Assessments <ul><li>diBartolomeo, Mitra and Mitra (2009) forthcoming in Quantitative Finance </li></ul><ul><ul><li>Follow the diBartolomeo and Warrick mathematical framework </li></ul></ul><ul><ul><li>Allow the conditioning information set to include both option implied volatility and variations in text news flows from Ravenpack (derived from Dow-Jones text feeds) </li></ul></ul><ul><ul><li>Empirical tests on Euro Stoxx 50 during January 17-23, 2008 and Dow Jones 30 stocks September 18 to 24, 2008 </li></ul></ul><ul><ul><li>Evaluate both individual stocks, full index and financial/non-financial subset portfolios </li></ul></ul><ul><li>In all cases, inclusion of quantified news flows improved the rate of adjustment of risk estimates to time variation in volatility faster than implied volatility alone </li></ul>
    20. 20. Incorporating Investor Attention <ul><li>Our next step will be to directly measure the degree of investor attention to a stock </li></ul><ul><ul><li>Judge investor interest directly by measuring the number of Google and Yahoo searches on trading symbols </li></ul></ul><ul><ul><li>Avoid company names to eliminate product or service related searches </li></ul></ul><ul><ul><li>Try it yourself with Google Trends </li></ul></ul><ul><li>Da, Engleberg and Gao (2009) have already documented a strong relationship between abnormal search frequency and price momentum </li></ul><ul><li>Investor attention is not always a good thing </li></ul><ul><ul><li>Bolster and Trahan (2009) document predictable price behavior in stocks mentioned on the Jim Cramer television show </li></ul></ul><ul><ul><li>Clear strategy: wait two days, then short every stock mentioned positively or negatively </li></ul></ul>
    21. 21. Crucial Refinement <ul><li>diBartolomeo and Warrick (2005), and diBartolomeo, Mitra and Mitra (2009) both assume that the full impact of the conditioning information should applied to ex-ante risk estimates </li></ul><ul><li>Shah (2008) introduces formal Bayesian framework for incorporating conditioning information into models </li></ul><ul><ul><li>Requirement for orthogonal factors is removed </li></ul></ul><ul><ul><li>Non-linear optimization to “fit” the adjustments to correlated factors is even more complex </li></ul></ul><ul><ul><li>Introduced into Northfield “near-horizon” models in May 2009 </li></ul></ul><ul><ul><li>Reduces noise and allows for fitting to different time horizons </li></ul></ul>
    22. 22. Other Differences Between Long and Short Horizon Risk <ul><li>Negative serial correlation </li></ul><ul><ul><li>Daily overreactions & reversals, which cancel out over time, become significant e.g. under leverage </li></ul></ul><ul><li>Contagion / panic </li></ul><ul><ul><li>Liquidity demands can drive up short-term correlations </li></ul></ul><ul><li>Transient behavior </li></ul><ul><ul><li>A long term model intentionally integrates new phenomena slowly: Is the future like the past or are we in and concerned about a present shift? </li></ul></ul><ul><li>Lots more extreme events in the short-term </li></ul><ul><ul><li>3 std deviations contains less probability mass. 99% VaR is farther away from the mean </li></ul></ul>
    23. 23. Conclusions <ul><li>The key to good short term risk assessment is understanding how conditions now are different than they usually are </li></ul><ul><li>A broad set of information other than stock characteristics and past returns are clearly useful in improving risk estimates </li></ul><ul><li>Among the most useful sets of conditioning information appears to be summaries of textual news flows to investors </li></ul><ul><li>A rigorous Bayesian framework should be employed to intelligently combine long term and short term information sets </li></ul>
    24. 24. ARMANDO GONZALEZ <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    25. 25. ARUN SONI <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    26. 26. JAMES CHENERY <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    27. 27. INCOPRORATING NEWS ANALYSIS INTO TRADING AND INVESTMENT PROCESSES FORUM ON NEWS ANALYTICS November 9, 2009 James Chenery Business Development Manager
    28. 28. EXPLOITING NEWS CONTENT <ul><li>News is emerging as differentiated, value generating content set </li></ul><ul><ul><li>Quant strategies – all trading frequencies </li></ul></ul><ul><ul><li>Human decision support – especially with analytic enhancements </li></ul></ul><ul><li>Key uses </li></ul><ul><ul><li>Speed – beat the humans, beat the machines </li></ul></ul><ul><ul><li>Manage scale and scope of events affecting portfolio </li></ul></ul><ul><ul><li>Risk management and loss avoidance </li></ul></ul><ul><li>NewsScope product line - Robust set of capabilities </li></ul><ul><ul><li>Historical data to back test and build algorithms </li></ul></ul><ul><ul><li>Real-time feeds for deployment, including ultra-low latency feed </li></ul></ul><ul><ul><li>News Analytics which convert qualitative text into quantitative data </li></ul></ul>
    29. 29. EXPLOITING NEWS CONTENT <ul><li>News flow is a good indicator of volume and volatility </li></ul><ul><li>Pricing movements accompanied by news tend to be momentum in nature; those with a lack of news tend to reverse to average trends </li></ul><ul><li>The market tends to overreact when there is a lot of news on something and under-react when there is a small quantity of news </li></ul><ul><li>For other direction and magnitude signals, find cause:effect relationships </li></ul>
    30. 30. MACHINE READABLE NEWS USE CASES <ul><li>Wolf detection / circuit breaker </li></ul><ul><li>News flow algorithms </li></ul><ul><li>Alpha generating signal </li></ul><ul><li>Post trade analysis </li></ul><ul><li>Stock screening tool </li></ul><ul><li>Risk Management </li></ul><ul><li>Compliance / Market abuse </li></ul><ul><li>Fundamental research </li></ul><ul><li>Trader decision support </li></ul>
    31. 31. NEWSSCOPE PRODUCT PORTFOLIO <ul><li>NewsScope Archive </li></ul><ul><ul><li>Historical database of Reuters and select third-party market moving sources </li></ul></ul><ul><li>NewsScope Direct </li></ul><ul><ul><li>Ultra-low latency feed of highly structured news and economic data </li></ul></ul><ul><li>NewsScope Analytics (aka NewsScope Sentiment Engine) </li></ul><ul><ul><li>Automated news analysis solution measuring sentiment, relevance, and novelty of text along with a host of other valuable metadata </li></ul></ul><ul><li>NewsScope Event Indices </li></ul><ul><ul><li>Automated news analysis solution indicating when abnormal amounts of news occur across various categories </li></ul></ul>
    32. 32. HOW TO MAKE MONEY WITH THE NEWSSCOPE ANALYTICS Buy on good news Sell on bad news Outperform S&P500 by 5000 basis points over a 60 day period! S&P1500 stocks in 2008; Daily items >50; Pos vs Neg >50%
    33. 33. Questions?
    34. 34. MARK VREIJLING <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    35. 35. GANGADHAR DARBHA <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    36. 36. GURVINDER BRAR <ul><li>Forum on News Analytics applied to Trading, Fund Management and Risk Control </li></ul>
    37. 37. Exploiting news-flow signals Macquarie Quantitative Research Gurvinder Brar, Christian Davies, Adam Strudwick, Andy Moniz [email_address] Macquarie Capital (Europe) Ltd Level 2, Moor House, 120 London Wall , London EC2Y 5ET November 2009
    38. 38. The global presence of Macquarie Quant <ul><li>17 professional staff across the globe </li></ul>Page Australia (8) Research George Platt John Conomos Portfolio Products Scott Hamilton Burke Lau Quant Applications Connah Cutbush Simon Rigney George Ferizis Charles Lowe Europe (4) Gurvinder Brar Christian Davies Andy Moniz Adam Strudwick South Africa (1) Hannes Uys Asia (2) Martin Emery Viking Kwok Japan (2) Custom Products Patrick Hansen Ayumu Kuroda
    39. 39. Recent publications Page <ul><li>Monthly ‘ Quantamentals’ report </li></ul><ul><ul><li>Unwrapping value, Oct 08 </li></ul></ul><ul><ul><li>Quality Control, Nov 08 </li></ul></ul><ul><ul><li>When the tide turns, Dec 08 </li></ul></ul><ul><ul><li>Style Outlook 2009, Jan 09 </li></ul></ul><ul><ul><li>Exploiting dividend uncertainty, Feb 09 </li></ul></ul><ul><ul><li>Do Technicals Add Value, Mar 09 </li></ul></ul><ul><ul><li>Asymmetric Style, Apr 09 </li></ul></ul><ul><ul><li>Have I got News for You? May 09 </li></ul></ul><ul><ul><li>Positioning for Recovery, June 09 </li></ul></ul><ul><ul><li>Preparing for regime changes, July 09 </li></ul></ul><ul><ul><li>Spotting Growth, Sept 09 </li></ul></ul><ul><ul><li>Arming Models with industry-specific data, Oct 09 </li></ul></ul><ul><li>‘ Global Dynamics’ report </li></ul><ul><ul><li>Focusing on earnings revisions, Oct 08 </li></ul></ul><ul><ul><li>Beyond Minimum Variance, Jan 09 </li></ul></ul><ul><ul><li>Asset Allocation: Spoilt for choice?, Apr 09 </li></ul></ul><ul><li>‘ Risky Business’ report </li></ul><ul><ul><li>Reality bites: Report on risk and implementation, Nov 08 </li></ul></ul><ul><ul><li>Stopping losses, taking profits, Feb 09 </li></ul></ul><ul><ul><li>Portfolio Turnover: Friend or Foe, June 09 </li></ul></ul>
    40. 40. Page News-flow research in vogue Source: Macquarie Research, November 2009 <ul><li>Academic literature </li></ul><ul><ul><li>The impact of public information on the stock market, 1994 </li></ul></ul><ul><ul><li>The market impact of corporate news stories, 2004 </li></ul></ul><ul><ul><li>More than words: Quantifying language to measure firms’ fundamentals, 2007 </li></ul></ul><ul><ul><li>Equity portfolio risk (volatility) estimation using market information and sentiment, 2008 </li></ul></ul><ul><ul><li>Investor inattention and Friday earnings announcements, 2008 </li></ul></ul><ul><ul><li>In search of attention, 2009 </li></ul></ul><ul><ul><li>Impact of news sentiment on abnormal stock returns, 2009 </li></ul></ul><ul><li>Data challenges </li></ul><ul><ul><li>Timeliness of news – Key newswires, stock exchange statements, press releases from company websites, national newspapers </li></ul></ul><ul><ul><li>Relevance of news – Company names mentioned in headlines/1st paragraph of text </li></ul></ul><ul><ul><li>Classification of news – Accounting-related versus strategic news </li></ul></ul><ul><ul><li>Independence of news – Mixed versus standalone events </li></ul></ul><ul><ul><li>Informational content of news – Identifying good versus bad news (computational linguistics/market based) </li></ul></ul>QUANTAMENTALS: HAVE I GOT NEWS FOR YOU? MAY ‘09
    41. 41. Page Data vendors QUANTAMENTALS: HAVE I GOT NEWS FOR YOU? MAY ‘09 Source: Macquarie Research, November 2009 <ul><li>Bloomberg - “ Black Box Newsfeed ” and “ Black Box ECO Stats ” . Low-latency delivery with 10,000+ daily corporate and economic headlines and text, 1500 category codes, 18month history </li></ul><ul><li>CapitalIQ – Website based search engine with categorized headlines, though without full text of article. </li></ul><ul><li>Dow Jones Elementized News feed   - Low latency, tagged data feed (from Jan 2004). News categorized by corporate event </li></ul><ul><li>Dow Jones News and Archives -  Text feed, 20+ years archive with identifiers, headlines & full stories. News items are not tagged into categories </li></ul><ul><li>Ravenpack - Sentiment scoring for traditional news wires (DJ News), internet sources (CNN Money) and blogs </li></ul><ul><li>Factiva.com – Website based search engine with categorized headlines and text </li></ul>
    42. 42. Page Exploiting news flow strategies QUANTAMENTALS: HAVE I GOT NEWS FOR YOU? MAY ‘09 Strategy Performances Information ratios (2001-2009) Source: Macquarie Research, November 2009 <ul><li>How to define the event? Companies may announce several news items over a month, should we react to all? </li></ul><ul><li>Informational content? Once we see an event, how do we systematically decide on its significance? </li></ul><ul><li>Holding period? Dealing with conflicting signals, excessive turnover and breadth of strategy </li></ul>
    43. 43. Page Important disclosures: Recommendation definitions – For quarter ending 30 September 2009 Recommendation definitions Macquarie - Australia/New Zealand Outperform – return > 5% in excess of benchmark return Neutral – return within 5% of benchmark return Underperform – return > 5% below benchmark return Macquarie – Asia/Europe Outperform – expected return >+10% Neutral – expected return from -10% to +10% Underperform – expected <-10% Macquarie First South - South Africa Outperform – return > 10% in excess of benchmark return Neutral – return within 10% of benchmark return Underperform – return > 10% below benchmark return Macquarie - Canada Outperform – return > 5% in excess of benchmark return Neutral – return within 5% of benchmark return Underperform – return > 5% below benchmark return Macquarie - USA Outperform – return > 5% in excess of benchmark return Neutral – return within 5% of benchmark return Underperform – return > 5% below benchmark return Recommendation – 12 months Note: Quant recommendations may differ from Fundamental Analyst recommendations Volatility index definition* This is calculated from the volatility of historic price movements.   Very high–highest risk – Stock should be expected to move up or down 60-100% in a year – investors should be aware this stock is highly speculative. High – stock should be expected to move up or down at least 40-60% in a year – investors should be aware this stock could be speculative. Medium – stock should be expected to move up or down at least 30-40% in a year. Low–medium – stock should be expected to move up or down at least 25-30% in a year. Low – stock should be expected to move up or down at least 15-25% in a year.   * Applicable to Australian/NZ stocks only Financial definitions All &quot;Adjusted&quot; data items have had the following adjustments made: Added back : goodwill amortisation, provision for catastrophe reserves, IFRS derivatives & hedging, IFRS impairments & IFRS interest expense Excluded: non recurring items, asset revals, property revals, appraisal value uplift, preference dividends & minority interests EPS = adjusted net profit /efpowa* ROA = adjusted ebit / average total assets ROA Banks/Insurance = adjusted net profit /average total assets ROE = adjusted net profit / average shareholders funds Gross cashflow = adjusted net profit + depreciation *equivalent fully paid ordinary weighted average number of shares All Reported numbers for Australian/NZ listed stocks are modelled under IFRS (International Financial Reporting Standards). AU/NZ Asia RSA USA CA EUR Outperform 45.08% 54.02% 40.00% 42.31% 62.86% 43.61% Neutral 39.77% 19.10% 45.00% 43.36% 31.90% 39.85% Underperform 15.15% 26.88% 15.00% 14.34% 5.24% 16.54%
    44. 44. Page Analyst Certification: The views expressed in this research accurately reflect the personal views of the analyst(s) about the subject securities or issuers and no part of the compensation of the analyst(s) was, is, or will be directly or indirectly related to the inclusion of specific recommendations or views in this research. The analyst principally responsible for the preparation of this research receives compensation based on overall revenues of Macquarie Group Ltd ABN 94 122 169 279 (AFSL No. 318062 )(MGL) and its related entities (the Macquarie Group) and has taken reasonable care to achieve and maintain independence and objectivity in making any recommendations. General Disclaimers: Macquarie Securities (Australia) Ltd; Macquarie Capital (Europe) Ltd; Macquarie Capital Markets Canada Ltd; Macquarie Capital Markets North America Ltd; Macquarie Capital (USA) Inc; Macquarie Capital Securities Ltd; Macquarie Capital Securities (Singapore) Pte Ltd; Macquarie Securities (NZ) Ltd; and Macquarie First South Securities (Pty) Limited are not authorized deposit-taking institutions for the purposes of the Banking Act 1959 (Commonwealth of Australia), and their obligations do not represent deposits or other liabilities of Macquarie Bank Limited ABN 46 008 583 542 (MBL) or MGL. MBL does not guarantee or otherwise provide assurance in respect of the obligations of any of the above mentioned entities. MGL provides a guarantee to the Monetary Authority of Singapore in respect of the obligations and liabilities of Macquarie Capital Securities (Singapore) Pte Ltd for up to SGD 35 million. This research has been prepared for the general use of the wholesale clients of the Macquarie Group and must not be copied, either in whole or in part, or distributed to any other person. If you are not the intended recipient you must not use or disclose the information in this research in any way. If you received it in error, please tell us immediately by return e-mail and delete the document. We do not guarantee the integrity of any e-mails or attached files and are not responsible for any changes made to them by any other person. MGL has established and implemented a conflicts policy at group level (which may be revised and updated from time to time) (the &quot;Conflicts Policy&quot;) pursuant to regulatory requirements (including the FSA Rules) which sets out how we must seek to identify and manage all material conflicts of interest. Nothing in this research shall be construed as a solicitation to buy or sell any security or product, or to engage in or refrain from engaging in any transaction. In preparing this research, we did not take into account your investment objectives, financial situation or particular needs. Before making an investment decision on the basis of this research, you need to consider, with or without the assistance of an adviser, whether the advice is appropriate in light of your particular investment needs, objectives and financial circumstances. There are risks involved in securities trading. The price of securities can and does fluctuate, and an individual security may even become valueless. International investors are reminded of the additional risks inherent in international investments, such as currency fluctuations and international stock market or economic conditions, which may adversely affect the value of the investment. This research is based on information obtained from sources believed to be reliable but we do not make any representation or warranty that it is accurate, complete or up to date. We accept no obligation to correct or update the information or opinions in it. Opinions expressed are subject to change without notice. No member of the Macquarie Group accepts any liability whatsoever for any direct, indirect, consequential or other loss arising from any use of this research and/or further communication in relation to this research. Clients should contact analysts at, and execute transactions through, a Macquarie Group entity in their home jurisdiction unless governing law permits otherwise. Country-Specific Disclaimers: Canada: In Canada, securities research is prepared, approved and distributed by Macquarie Capital Markets Canada Ltd, a participating organisation of the Toronto Stock Exchange, TSX Venture Exch ange & Montréal Exchange. Macquarie Capital Markets North America Ltd., which is a registered broker-dealer and member of FINRA, accepts responsibility for the contents of reports issued by Macquarie Capital Markets Canada Ltd in the United States and to US persons and any person wishing to effect t ransactions in the securities described in the reports issued by Macquarie Capital Markets Canada Ltd should do so with Macquarie Capital Markets North America Ltd. Securities research is issued and distributed by Macquarie Securities (Australia) Ltd (AFSL No. 238947) in Australia, a participating organisation of the Australian Securities Exchange; Macquarie Securities (NZ) Ltd in New Zealand, a licensed sharebroker and New Zealand Exchange Firm. The Research Distribution Policy of Macquarie Capital Markets Canada Ltd is to allow all clients that are entitled to have equal access to our research. United Kingdom : In the United Kingdom, securities research is issued and distributed by Macquarie Capital (Europe) Ltd, which is authorised and regulated by the Financial Services Authority (No. 193905). Hong Kong: In Hong Kong, securities research is issued and distributed by Macquarie Capital Securities Ltd in Hong Kong, which is licensed and regulated by the Securities and Futures Commission. Japan: In Japan, securities research is issued and distributed by Macquarie Capital Securities (Japan) Limited in Japan, a member of the Tokyo Stock Exchange, Inc., Osaka Securities Exchange Co. Ltd, and Jasdaq Securities Exchange, Inc. (Financial Instruments Firm, Kanto Financial Bureau (kin-sho) No. 231, a member of Japan securities Dealers Association and Financial Futures Association of Japan). South Africa: In South Africa, securities research is issued and distributed by Macquarie First South Securities (Pty) Limited in South Africa, a member of the JSE Limited. Singapore: In Singapore, securities research is issued and distributed by Macquarie Capital Securities (Singapore) Pte Ltd (Company Registration Number: 198702912C), a Capital Markets Services license holder under the Securities and Futures Act to deal in securities and provide custodial services in Singapore. Pursuant to the Financial Advisers (Amendment) Regulations 2005, Macquarie Capital Securities (Singapore) Pte Ltd is exempt from complying with sections 25, 27 and 36 of the Financial Advisers Act. United States: In the United States, securities research is issued and distributed by Macquarie Capital (USA) Inc., which is a registered broker-dealer and member of FINRA. Macquarie Capital (USA) Inc, accepts responsibility for the content of each research report prepared by one of its non-US affiliates when the research report is distributed in the United States by Macquarie Capital (USA) Inc. Macquarie Capital (USA) Inc.’s affiliate’s analysts are not registered as research analysts with FINRA, may not be associated persons of Macquarie Capital (USA) Inc., and therefore may not be subject to FINRA rule restrictions on communications with a subject company, public appearances, and trading securities held by a research analyst account. Any persons receiving this report directly from Macquarie Capital (USA) Inc. and wishing to effect a transaction in any security described herein should do so with Macquarie Capital (USA) Inc. Important disclosure information regarding the subject companies covered in this report is available at www.macquarie.com/research/disclosures, or contact your registered representative at 1-888-MAC-STOCK, or write to the Supervisory Analysts, Research Department, Macquarie Securities, 125 W.55th Street, New York, NY 10019. © Macquarie Group

    ×