Denmark uses @RISK for
Road and Infrastructure Projects
Story Covered Widely in the Press
The Centre for Traffic and Transport (CTT), part of the Technical University of Denmark (DTU), is using @RISK to evaluate the feasibility of new roads and transport projects. On behalf of a department of the Danish Ministry of Transport, it also assesses large infrastructure programs such as proposals for bridges.

In particular, the CTT undertakes extensive evaluations with @RISK to produce a cost-benefit ratio that determines whether the benefits of a proposed transport venture justify its economic, social, and environmental cost. Risk assessments include calculating if a new road will shorten journey times, reduce accidents, increase pollution, and more.

The story was picked up by Hoovers, Wachovia, Directions magazine, and nearly a dozen other outlets.

» Read the full story on Hoovers

Join the Party: Palisade User
Conference Miami Almost Here!
@RISK 5.0 Unveiled Oct 25
The brand-new @RISK 5.0 will be presented at the Palisade User Conference in Miami by Palisade President Sam McLafferty, and hands-on training sessions will be offered using the new software. Experience the stunning new interface, powerful analyses, and unprecedented sharing capabilities of this exciting new release. Plus, don’t miss the launch party!

With sessions on topics like utilities regulation, demand forecasting, environmental cleanup, and energy risk assessment from companies like Suncor, BC Hydro, and FutureMetrics LLC, attendance is on track to beat last year. Check out the program and register today.

» View conference schedule
» Register today for only $495 $295

2007 Palisade User Conference
Asia-Pacific Draws Record Crowd
Palisade is pleased to announce the success of its second Asia-Pacific User Conference. Held in Sydney on 13-14 September, 2007, the event attracted nearly double the delegation as the inaugural 2006 conference. Participants included practitioners, academics, and consultants from across the Asia-Pacific region.

“A lot of attendees heard about our conference from colleagues,” noted Palisade Asia-Pacific Managing Director Mark Meurisse. “The word is spreading – the Palisade User Conference is becoming the forum for quantitative risk and decision analysis.”

Customers that took part in Asia-Pacific event included: Sanofi-Aventis, Hatch, BlueScope Steel, Sydney Water, OneSteel, CH3MHILL, Marsh Ltd, Telstra, Shell, Western Australia Department of Agriculture, the University of New South Wales, and many more.

» Download presentations from the Conference
» More about the Conference

Free Live Webcast:
The DecisionTools Suite at Unilever
"Software is Only Part of the Solution..."
Thursday, 8 November 2007 @ 11:00am ET
For the last three years Unilever’s Finance Academy has been rolling out and embedding its Decision Making Under Uncertainty (DMUU) program to its finance managers, project leaders and key decision makers. The key aim of this program is to create a culture where high-quality decisions and probabilistic analysis are the norm. Palisade’s DecisionTools Suite provides the analytical engine which enables the analysts to develop clear insights and gain commitment to action. By building a collaborative relationship with Palisade, Unilever has developed joint courses focused on Unilever-specific requirements and armed over 200 people with the software tools they need to analyze complex decisions. This presentation will share the experiences and learnings Unilever has gained from the past three years.

» Register now for FREE Live Webcast:
"The DecisionTools Suite at Unilever:
Software is Only Part of the Solution..."
Thursday, 8 November 2007, 11:00am - 12:00pm ET / 4:00 - 5:00 pm GMT

@RISK, PrecisionTree
Protecting Endangered Species
Faced with a lack of data, the World Conservation Union’s Conservation Breeding Specialist Group is using @RISK and PrecisionTree to identify and analyze disease in endangered animals. According to Dr. Douglas Armstrong of Omaha’s Henry Doorly Zoo, “It’s an effective approach to minimizing the risks of disease in our efforts to conserve endangered species.”

» Read the full case study

Distribution Fitting
in @RISK
There may be times when you have historical or industry data on a particular input in your model. For example, you may have price data on the material costs of a product you are going to produce. You would like to use this data to represent uncertainty in your model, but how?

@RISK Professional and Industrial include integrated distribution fitting. Just read your data into @RISK and click the Fit button. @RISK will find the distribution which best describes your data. A handy right-click command writes the newfound distribution to your spreadsheet cell. You can also link your fit to the resulting @RISK functions. If the underlying data changes, the fit automatically re-runs and updates the distribution in your model.

@RISK ranks dozens of distribution functions against your data, and provides comprehensive results. Four types of graphs are available to assess accuracy of the fit, and there are complete statistical reports and goodness-of-fit data.

Choose from Comparison, Difference, Probability-Probability, or Quantile-Quantile graphs. All graphs include sliding delimiters and probability bars. Graphs are also easily formatted, and exported to Excel in native Excel format.

@RISK uses three advanced fitting algorithms to optimize its fits – Chi Square, Anderson-Darling, and Komolgorov-Smirnov. You can read in data sets with up to 100,000 points, in sample, density, or cumulative format. You can fit multiple data sets in a single project, and specify which predefined distributions to fit to. BestFit allows full control over Chi-squared calculations, including equal interval binning, equal probability binning, and full custom binning. It will also perform the RMS (root mean square error) test for cumulative and density data.

» More on @RISK

Credit Card Fraud

» Visa Makes Back-to-School Online Shopping Easier
Technique: Neural networks   » see NeuralTools
Source: MarketWire, August 31, 2007

Visa is now adding Verifed by Visa to the multiple layers of protection it offers cardholders through programs such as the three-digit code (CVV2), address authentication and neural networks.

Disease Prevention

» Defra Prepared to Vaccinate - But Not Yet
Technique: Decision Trees   » see PrecisionTree
Source: Farmers Guardian, September 19, 2007

Using decision trees, the UK government contemplates vaccination to combat foot and mouth disease.

Medical Devices

» ISU receives $842,000 from DOD to Develop
'Smart' Prosthetic Hand
Technique: Neural networks, Genetic Algorithms
» see NeuralTools and Evolver
Source: Idaho State University press release,
September 20, 2007

Under a new grant from the Department of Defense, researchers will try to develop an “intelligent” control for prosthetics using a variety of sophisticated computing techniques, including neural networks and genetic algorithms.

Paleontology

» Paleontology and Statistics
Technique: Monte Carlo simulation » see @RISK
Source: Swarthmore College, Daily Gazette,
September 13, 2007

At Swarthmore College, paleontology professor Steve Wang presented research based on Monte Carlo simulation that aimed to uncover the cause of an unexplained extinction event 250 million years ago that brought the annihilation of nearly 95% of Earth’s then-living species.

Retirement Planning

» Retirement Advice: Putting off Retirement
will Improve Income
Technique: Monte Carlo simulation » see @RISK
Source: San Antonio Express, September 2, 2007

A financial columnist relies on widely used Monte Carlo simulation software to help advise a retiree on how to maximize her income over her lifetime.

» How Much Retirement Income to Take
Technique: Monte Carlo simulation » see @RISK
Source: Sympatico MSN Finance, September 15, 2007

A look at how retirement expert William Klinger calculates how changing withdrawal rates affects retirement income. His Monte Carlo simulations make use of 78 years of data on asset and inflation...

» Weathering Much Retirement Income to Take
Technique: Monte Carlo simulation » see @RISK
Source: Barron's, September 17, 2007

One planning approach has improved since the current bear market began. “Monte Carlo simulation programs are now standard for advisers preparing a financial plan. . .”

Sports

» The Odds of Your Team Making the Playoffs Are...
Technique: Monte Carlo simulation  » see @RISK
Source: New Jersey Star-Ledger, August 30, 2007

On his website, Sports Club Stats, Ken Roberts demonstrates how to use Monte Carlo Simulation to estimate the likelihood of any baseball team making the playoffs.

» Playing the Postseason Percentages
Technique: Monte Carlo simulation  » see @RISK
Source: Toronto Globe and Mail, September 1, 2007

Sports writer Jeff Blair also considers the chances of each baseball’s team’s making the playoffs, and he too bases the odds on Monte Carlo simulations.

» How the Mariners' Ship Sank So Fast
Technique: Monte Carlo simulation  » see @RISK
Source: New York Sun, September 12, 2007

In August the Baseball Prospectus' Playoff Odds report used a Monte Carlo simulation based on team stats to predict the Mariners had a 29 percent chance of winning the American League title. Now the think tank has recalculated that model in light of the team’s sudden collapse.

Wealth Management

» Highline Wealth Management Achieves
$500 Million in Assets Under Management
Technique: Monte Carlo simulation  » see @RISK
Source: PR Newswire, September 19, 2007

Highline Wealth Management pursues an alternative investment strategy to reduce the effects of volatility on investments, and leverages a variety of allocation models including Monte Carlo simulation.

Expert Answers to Technical Questions

Dear Amy,

How do I get the RiskLognorm function to output an expected “textbook” mean and standard deviation after a simulated distribution?

—S.B.

Dear S.B.,

@RISK, RISKOptimizer, and the developer kits have two lognormal distributions. RiskLognorm2 is the traditional distribution and will behave in the way described in statistics books. Palisade also offers RiskLognorm, where the mu and sigma you enter are the actual mean and standard deviation of the distribution, subject to the usual sampling fluctuation. The two distributions are the same except for the way you enter parameters.

If you know the actual mean and standard deviation you want in a log-normal distribution, use RiskLognorm. If you want to use parameters that match the lognormal distribution in many textbooks, use RiskLognorm2.

Finally, if you know the desired geometric mean and standard deviation of a log-normal distribution, use RiskLognorm2 but set mu to the natural log of the desired geometric mean, and sigma to the natural log of the desired geometric standard deviation.