What's new in Fairmat 1.7.0 Revision 796
Jun 18, 2015
- New:
- Added the possibility of setting a custom date in date transformation / advance. This helps in generating sequences of payments with non-standard schedules
- Sensitivity and Impact analysis on arrays: with the new Fairmat version is possible to assess the impact of changes when elements are defined as vectors
- New Functions are available: ALast (calculates the last index of an array satisfying a given condition), CumSum (calculates the cumulated sum of elements belonging to an array) and IssuerRecovery (calculates the recovery rate from a CDS ticker)
- Projects using Monte Carlo Simulation and Black can be mixed in the same project
- The average operator now uses Monte Carlo simulation for averaging nodes instead of averaging expected values: this allows to preserve the information about the risk of subsequent nodes
New in Fairmat 1.6.2 Revision 775 (Jun 12, 2014)
- New:
- New Calendar Dates processing EndOfMonth function
- Rate function can be used also for calculating zero coupon prices on zero rate functions (i.e the following expression is now handled rate(0;10;@zr) )
- Custom Valuation Attributes: The output of any block can be reported in the valuation stack (see gallery). This feature can be used to automatize unbundling or to customize output generation.
- Automatic Global Correlation: The cross correlations between all risk factors is calculated automatically (If ESG plug-in is loaded). In earlier Fairmat versions correlation was block-diagonal, and a global correlation must be entered manually.
- Improvements:
- Improvement to multivariate time-to-default modelling: first-to-Default models can be calibrated directly by specifying a list CDS tickers (previously it was necessary to enter a list of CDS and ZR curves and a correlation matrix. This procedure Fairmat does everything automatically. Furthermore, it possible to apply indexing to multi-variate random variables: writing U1a+ U1b, it is equivalent to writing U1[1] +U1[2]. The array style syntax implies that array functions (asum, amin, amax, aprod, …) can be applied to random variable allowing compacting of multivariate credit events: For example, the following expressions for the payoff of a Zero-Coupon-Basket are equivalent.
- Goal Seeker: Now it possible to set custom target values (not only zero) and seeker preferences are remembered over user sections.
New in Fairmat 1.6.1 Revision 766 (Mar 17, 2014)
- New product templates: Government bonds, government bonds options and autocalls
- Equity tickers handling: it is now possible to associate tickers to specific stock attributes. For example, by writing stockname#dy it will be possible to refer to the historical series of dividend yields and hence calibrate a models representing stochastic dividend yields.
- Static Constraints: Static constrains can be used to create custom validation procedures which may take into account relationships between several Fairmat objects and hence provide users specific error messages if such relationships are not satisfied. Constraints are expressed as Fairmat expressions. Example: ensuring that every element belonging to vector A precedes every element belonging to vector B: amax(@A) < amin(@B).
- New calendar date function: DaysInMonth returns the number of days for a given month
- The Valuation Date is now reported on the Valuations Stack making easy to compare several valuation
- Add new constant ValuationDate (which is always equal to 0 when evaluated), but it may be referred by dates adjustments functions
- Now it possible to also adjust EffectiveDate passing them as argument to dates adjustments transformations.
- New Plugins
- Fairmat Economic Scenario Generator (ESG), a new solution for generating real-world and risk-neutral market consistent economic scenarios for the following asset classes: Zero Coupon bond Prices, Defaultable bond prices (rating and CDS based), Inflation Indices, Equity (i.e. Euro Stoxx 50, FTSE 100…) prices and dividends. For more information read the presentation and the product's documentation.
- Jarrow-Lando-Turnbull model which may be used to price bonds with default risk, i.e., bonds for which there is the possibility that the issuer does not honour the schedule of payments. From risky zero coupon bond prices, one can calculate credit spreads. JLT is a reduced-form model, i.e., it models the credit migration process (which is the driver of the credit risk) through a Markov Chain with the default state being absorbing.
- Ornstein-Uhlenbeck model open source plug-in for calibrating mean reverting processes to historical series
- Improvements
- Modeling GUI is now far more responsive.
- Error messages improvements: previous Fairmat versions will show you messages like this one: ‘Argument is out of range. Parameter name: index...’ In Fairmat 1.6.1 the same message is shown as follows: ‘Function amin has wrong inputs: starting index is 75 is not coherent with vector rda length which is 72’
- Calibration Proxies: risk neutral calibration may calibrate volatilities over historical series when volatility surfaces are not available and the retrieval mode is set to CarryLastAvailableData.
New in Fairmat 1.6.0 Revision 743 (Sep 17, 2013)
- Equity and FX contracts Valuation, with the new Data-Link. Data-Link 2.0 data providing service has been improved in order to allow you perform end-of-day pricing on new asset classes.
- Data-Link will aggregate public available data and will merge with our provided data (i.e. interest rate volatility surfaces and discounting curves) in order to allow you to price any contract (Interest Rate Linked / Equity Linked / FX-Linked derivatives) on several markets.
- Tickers autocomplete: When pressing CTRL+SPACE in the data-source tab tickers textbox, you will receive autocomplete information about the available equity and currency tickers.
- Automatic calibration of Geometric Brownian Motion models.
- Integration with Yahoo! Finance.
- Access to the European Central Bank echange rates reference data.
- Access to the MEFF market historical prices and option chains.
New in Fairmat 1.5.0 (Sep 17, 2013)
- What’s new:
- [Professional] More than 30 templates on equity linked products are available as ready to use templates which may be used in conjunction of Bloomberg Professional.
- [Academic/Professional] Variance gamma model plug-in allows to calibrate and simulate the Variance Gamma model for simulating stock prices. The model is both simple and robust and it is a very good alternation to the existing models like Heston stochastic volatility model and Black-Scholes model which may be biased when applied to real market circumstances. The calibration procedure is based on closed form calculation of European options.
- [Professional] Geometric Brownian Models can be calibrated against exchange rates, simplifying the valuation of FX derivatives.
- [Professional] Fairmat Server integration plug-in allows Fairmat Professional to access market data stored on a Fairmat Server Instance, making Fairmat Server to act as a Market Data Provider.
- [Professional] When a new templates is created the valuation and the effective date is set to "Today date".
- [Professional] Common label system: the creation of templates for the analysis mode have been simplified
- [Academic/Professional] Data request tab: every valuation is associated to the details of all the historical requests needed for the valuation (i.e price quotes appearing in payoffs).
- [Academic] Provided packages for ARCH Linux on our download area.
- Improvements:
- [Academic/Professional] Smart vector indexing support on matrix rows: the feature allows to write expressions like A[ AnyIndex, c in 1,…,C] if A is a 1xC matrix.
- [Academic/Professional] Calendar Dates Transformations works also on scalars (the result will be a vector with one element)
- [Professional] Several GUI Improvement, such as the new home screen, the simplification of vectors and matrices in Analysis mode.
- [Academic/Professional] Hull-White model and Pelsser allows historical/real-world calibration
- Fixes:
- More than 50 bugs have been fixed.
- If the template is implemented with close form model (Black or Black&Scholes) Numerical Settings tab doesn't appear.
- [Academic] Fixed Activation Under Ubuntu/Unity
New in Fairmat 1.4.1 Revision 677 (Dec 22, 2012)
- IAS 39 Hedge Accounting: The IAS 39 plug-in allows users to implement the accounting of financial products following the IAS 39 principles. Using this plug-in it will be straight forward to perform the prospective and retrospective tests on any contract modeled with Fairmat.
- Commodity Linked contracts pricing: The Convenience Yield calculator plug-in implements the capability of calculating instantaneous forward price for commodities like Natural Gas, Fuel Oil and Gasoil. The deterministic forward price model is calibrated starting from an input matrix of futures prices and other market data and specifications.
- New Special and distribution related functions:
- Special functions: error function (erf), beta function (beta), incomplete beta function (betainc), regularized incomplete beta function (betaincr), gamma function (gamma), incomplete gamma function (gammainc).
- Densities distributions and inverses of the following distributions: normal (normpdf, normcdf, norminv) chi square (chi2pdf, chi2cdf, chi2inv), non central chi square (ncx2cdf, ncx2inv), exponential (exppdf, expcdf, expinv), gamma (gampdf, gamcdf, gaminv), student's t (tpdf, tcdf, tinv) and f distribution (fpdf, fcdf, finv).
New in Fairmat 1.4.0 Revision 675 (Dec 3, 2012)
- Data-link: Data-link is an on-demand pricing service for Fairmat Academic. Data-link enables the pricing of interest rate linked products using either the Monte Carlo simulation or analytic formulations (provided by the PlainVanilla plug-in). The service allows the calibration of interest rate models (for example the Hull and White model, two factors and the Pelsser Squared Gaussian model) against market data while performing valuations on the users laptop thus preserving privacy of information. The Data-Link subscription periods are designed for letting you access market data and the remote valuation services only when you need them, making Data-Link an affordable SaaS solution for derivative pricing.
- Smart vector indexing: indexing on vectors containing one element is relaxed, thus simplifying the building of bullet and amortizing notional or other date dependent parameters.
- Improvements:
- Plain Vanilla Plug-in 1.1: Several improvements have been added: Full support for quanto adjustment, payment dates filtering and handles the retrieval of different discounting and forwarding curves.
- It is now possible to calculate either the clean price or the dirty price of a contract.
- There is improved stochastic processes categorization (processes are categorized into Equity, Inflation, Interest rates, etc)
- You’ll find improved plug-ins management: now online plug-ins come with search and filtering.
- Several usability improvements have been made to the scalar and function editing forms, as well as the Fairmat and the project Preferences windows and the outputs windows which now present outputs in a handy grid format.
- Changes:
- The Fairmat Academic user license is changed. Now it is possible to use Fairmat Academic for commercial purposes when in conjunction with the Data-Link service.
New in Fairmat 1.3.3 (Dec 3, 2012)
- Improvements:
- Improved stochastic processes categorization
- Improved error visualization in the XML Calibration form
- Several 2D Function visualization improvements
- Improved plug-ins management: now online plug-ins come with search functionalities.
- Several usability improvements to the scalar and function editing forms, the Fairmat and the project Preferences windows.
- Enforced Fairmat language syntax checks in recurrence functions
- Fixes:
- Added error handling when opening files using drag and drop
- Fixed wrong display of descriptive statistics of Monte Carlo simulation valuations
New in Fairmat 1.3.2 (Dec 3, 2012)
- Improvements and new features:
- Sensitivity and impact analyses can be applied also on functions: in this case the functions will be translated by constant amounts. The feature is very handy when it is necessary to perform stress tests on discounting curves or other non scalar objects.
- Implementation of first and second order partial derivatives operators. Now it is possible to calculates partial derivatives of user defined functions: for example in order to calculate the fist derivative of the function f in the point .5 it is possible to write partial(@f ; .5).
- Usability improvements: simplification of constants and functions editing forms, categorization of stochastic processes.
- Plug-in robustness: improved behavior of the system in the presence of outdated or plug-ins.
- Fixes:
- Fixed a regression (introduced in version 1.3.1) which prevented editing of constants on Ubuntu.
- Fixed a regression (introduced in version 1.3.1) which prevented the capability of executing valuations.
New in Fairmat 1.3.1 (Dec 3, 2012)
- Improvements:
- Parameters definition intervals can be expressed in absolute terms, instead of depending on the base value B. In previous versions, ranges can be specified by defining a range value R and definition intervals (used in different analyses) were computed as [B-R, B+R]. The actual release allows to define the interval [A, B] by writing the two extremes separated by a semicolon (i.e. A ; B ).
- While editing parameters tab keys can be used to move between editing fields.
- Vector, Matrix and control points function editors have been improved and it is now possible to add/remove columns easily.
- It possible to control either programmatically or through the user interface the metrics that will be calculate on the risk profile.
- Enforced correctness checks on strip options.
- Fixes:
- In some cases, the cursor was positioned in the wrong place after auto-completing an expression.
- Fixed templates directory search.
- Horizon, xi and EffectiveDate were not present in the list of reserved symbols.
New in Fairmat 1.3.0 (Dec 3, 2012)
- Changes:
- .NET 4.0: Fairmat Academic 1.3 runs under the .NET 4.0 profile. Custom plug-ins compiled for the version 1.2 must be recompiled in order to work on the new system. The adoption of the .NET 4.0 run-time will increase Fairmat compatibility with other system and third parties libraries.
- The user settings system has been improved and new user settings has been added, for example the choice of the random number generator, and editor and modeling options.
- The activation process has been postponed, and now Fairmat Academic allows short test sessions.
- Modeling environment improvements:
- A new transformation called Zero Rate generator is now available. With the Zero Rate generator you can control how a zero rate curve is generated starting from the Cash, Fra and Swap rated. The Zr generator front end can be found in the parameters & symbols under the transformation category.
- A new Fairmat Keyword “Horizon” has been defined
- In functional operators, the “Value” function can be also called with the implicit reference arguments. For example, if x1,...,xn are the variables modeling subsequent legs, the following expressions can be written “iif(value(@x1;U1)>0”.
- Improved copy-and-paste of data from MS Excel to custom functions and vectors.
- New open source plug-ins
- With the Fairmat 1.3 the following new plug-ins are available:
- The Random Sourcers Support plug-in improves support for external random sources and generators and allows users to use as source of randomness sequences stored in the file system.
- Berlin University Quantistic Random Number Generator webservice plug-in. This plug-in receive random sequences from a server hosted by department of phisic of the Berlin University.
- Dates Generator: With this plug-in is possible to generate dates sequences by specifying start date, end date and frequency.
- Cox-Ingersoll-Rubinstein (CIR) general equilibrium model is now available.
- Fixes:
- Fairmat 1.3 contains hundreds of fixes over the last 1.2.5 public version. Many issues on the modeler has been fixed, for what concerns the valuation side the following problems has been solved:
- The “EffectiveDate” variable is now initialized correctly
- fixed bug related to evaluation of disabled options and scenarios
- Two race conditions in multi-thread valuations have been fixed.
New in Fairmat 1.1 (Jul 8, 2011)
- Modeling capability enhancements:
- Recurrence functions can be used into stochastic discounting expressions.
- Improved discretization of maturities and payoffs with monte-carlo simulation.
- New functions: afirst, accrual, date functions (day, month, dayofweek), onevery, invnorm (see the online manual for more details).
- Random variables: Fairmat allows to define a certain number of random variables which realizations are consistent with the stochastic process dynamic. Random variables can be used in any payoffs like stochastic processes, but differently from stochastic processes, random variables realizes in a specific date: this allows to define random variables the distribution of which depends on the realization of another random variable.
- Modeling environment enhancements:
- Editing: Import from clipboard handles thousands separators
- Editing: Copy Cut & Paste operations are implemented for every Fairmat object such as symbols, stochastic processes and options map building blocks
- Functions' signature viewer: In addition to auto-complete capabilities, when editing complex payoffs, Fairmat will show you automatically the details about the functions you are typing
- Analysis enhancements:
- Dynamic and volatility analysis (Beta): The dynamic and volatility a analysis is similar to risk analysis, but differently from risk analysis which calculates the conditional distributions of future payments and values, the Dynamic and Volatility analysis calculates the volatility of the evolution of the contract value over time, treating the contract as it would be a stock or fund.
- New plug-ins:
- Hull and White two factors model:
- The Hull and white two factors model is now implemented as a Fairmat plug-in.
- Plain Vanilla:
- With Fairmat 1.1 is available a new plug-in called Plain Vanilla which allows you to model and price structured products starting from a palette of plain vanilla option blocks like Cap/Floor o Digital options (cash or nothing), using the Black model. The Plain Vanilla Plug-in enables:
- A new numerical method: “Analytic Solutions” is available
- A new Toolbar containing a CAP Floor, Cash or Nothing, Floating Rate Leg, Swaption and other blocks is available
- When selecting the Analytic solutions numerical method, it is possible to evaluate projects defined by the composition of the blocks contained in the plain-vanilla toolbar plus the sum/average and the functional operator.
- Analytical Greeks derivatives are automatically calculated in the valuation process.
- Plain-vanilla option blocks are thought to be used easily both in Fairmat Academic and in the commercial versions of Fairmat: contract specific parameters are separated from market data allowing more flexibility in the specification of market data by allowing:
- Manual specification. For example a cap block must contain explicit references to zero rate curves, spot or flat volatility curves and strikes matrices.
- Specification through data provider: Just select the market and Fairmat will do the rest (A commercial version of Fairmat is required).
New in Fairmat 1.0 RC 3 (Oct 12, 2010)
- Changes:
- Plug-ins are now user user dependent: every user can install plug-ins without the need of privileges escalation.
- Fixed bugs:
- Fairmat failed to validate stochastic/option dependent discounting expressions.
- Latent components of stochastic process were not simulated correctly in some cases.
- In certain circumstances Fairmat failed to recognize wrong expressions.