How will the new @RISK be different?

It’s been a very long time coming, four years in fact, but Palisade Corporation has announced that it will be showcasing a new version of @RISK and the DecisionTools Suite.

What can we expect?

Price

There will almost certainly be a price increase. @RISK is already the most expensive risk analysis add-in on the market, but the venture capital company that now owns Palisade will be wanting to get a venture-capital style return on their investment - since Palisade already dominates the market, that means cutting staff and raising prices.

@RISK users are incredibly loyal too. There are many alternatives available - some more sophisticated (ModelRisk Complete), some much cheaper or free (RiskAmp or ModelRisk Basic) and some that are better and cheaper (ModelRisk Complete) but @RISK users tend to stick to what they have. Palisade's new owners will surely be hoping to take advantage of that.

New features

History tends to repeat itself. Looking back over past updates of @RISK we saw:

Version 6 (2012)

This introduced a number of features, unashamedly 'inspired' by tools already available in ModelRisk, namely:

• Time series – a selection from the time series functions featured in ModelRisk, but with more complexity
• Distribution fitting – the same live fitting as ModelRisk and picking up ModelRisk’s use of information criteria as the fitting statistic, and a version of ModelRisk’s bootstrap feature for estimating parameter uncertainty, but missing its purpose a bit by not simulating the uncertainty in the spreadsheet
• New distributions – a selection of distributions from ModelRisk like Bernoulli, Double Triangle, Extreme Value Min, F, Laplace, and Levy
• Adding the OptQuest optimizer – ModelRisk had done this but later dropped it
• Spider plot and double-sided tornado plot for sensitivity – an exact replica of the ModelRisk features
• Connecting to MS Project – a feature they thought of themselves, revisiting an old idea (@RISK for Project) but not overcoming the simulation performance issues inherent in that idea

Version 7 (2015)

It took another three years to be inspired by ModelRisk again, but finally Palisade introduced:

• Copulas – the exact same set as ModelRisk
• A data viewer – the name inspired by ModelRisk, but without the features
• Efficient frontier – similar to the ModelRisk tool by the same name
• Some more distributions from the ModelRisk set – Burr, Cauchy, Dagum, Fatigue Life, Frechet, Hyperbolic Secant, Kumaraswamy and Reciprocal
• A few new statistical functions – including one that they thought of without ModelRisk’s help

And Version 8 (2019)?

The most reliable approach to guessing what @RISK Version 8 might include would be to do a gap analysis – i.e. see what ModelRisk features are still missing from @RISK. There are many, of course, but here are a list of the most important things Palisade could copy:

• Switch to time-based licenses. Not so much a feature, but a way to generate more revenue
• A separate Results Viewer application to make it easy to share results electronically and print reports to PDF, PowerPoint or Word
• Ditch the #VALUE! response when an @RISK function generates an error because of incorrect input parameters, and return sensible messages instead like all ModelRisk functions do
• Introduce ‘objects’. This makes it possible to do a lot more manipulation with distributions. For example, when Palisade tried to replicate the VoseAggregateMC function to make its RiskCompound function, it didn’t use objects and had to force the function to disobey Excel’s add-in rules which causes all sorts of model behaviour unpredictability
• Fix the correlation functions so they are visible not placed in hidden sheets
• Allow fitting to data residing in databases. ModelRisk’s DataObject function helps keep models clean and automatically updated with the latest values but it is difficult to develop
• Introduce extreme value calculations. ModelRisk does not describe how these remarkable tools work, so that may be a challenge
• Have a View Function button. A very helpful feature in ModelRisk is that one click on this button will open up an interface illustrating the function(s) in the cell
• Adding an Input function. Currently, the equivalent of the VoseInput function in ModelRisk is a hefty combination of several @RISK functions (RiskMakeInput, RiskName, RiskUnits, etc)
• Making the data viewer a dynamic data exploration tool like ModelRisk, though this may cannibalise some of the other tools it sells
• Allowing a section of the spreadsheet to appear in the abovementioned Results Viewer, very useful for capturing tables of cells with simulation statistical functions
• Some of the 50+ other functions and tools unique to ModelRisk like StopSum, insurance tools, ODE and numerical integration, etc

What will be the reaction from ModelRisk?

To many people, ModelRisk and @RISK are about the same since both products have features enough for most users. ModelRisk is half the price though, and we have a version of ModelRisk that is pretty powerful and completely free – it attracts 100+ new users a week. It also has an @RISK model converter.

Our primary use of ModelRisk is as a support tool for our Enterprise Risk Management system, Pelican. It allows models to be run, shared, connected to data in Pelican, and it allows you to store and compare simulation results - all within your web browser.

If you like the cool things that @RISK can do, you will probably really appreciate the other risk analysis software products from the people who first introduced all these cool features. Visit www.vosesoftware.com and check them out.