Copyright © 2011-2014 airxc.com, airxcell.com
Table of Contents
AirXCell a Calculation Software dedicated to Quantitive Research in Finance. It intents to provide the financial engineers and quantitative researchers in financial institutions with a dedicated tool to perform portfolio valuation, analysis, simulation, rebalancing and optimization as well as structured product pricing. AirXCell combines a powerful computation engine with a cutting edge, user friendly, efficient and complete Graphical Interface.
AirXCell is also a software framework for building online (RIA / Web 2.0) Quantitative Research Applications. It is specifically designed to address QR needs in Finance.
AirXCell currently supports a programmable spreadsheet (online clone of excel with an evolved formula language), an R development environment, various financial calculation forms and many other things related to Quantitative Research Tasks in Finance.
AirXCell is using R - The GNU R Project for Statistical Computing as calculation backend. Current version is still a beta version, yet it is fully functional.
AirXCell is a Web application, i.e it runs in a browser, enabling the customer to install it once and for all on a production server while every user accesses simply by opening a web browser and reaching the AirXCell site.
AirXCell supports several core applications within a single package, all applications being tightly coupled and deeply integrated with each other:
-
A spreadsheet application: just as Excel in a browser, only better.
The User Interface emulates Microsoft Excel while the calculator, on the other hand, goes much further. Instead of the poor and unexpressive Excel formula language built of keywords and parenthesis, AirXCell supports R as the formula language. While the syntax is close to Excel, the possibilities are unlimited, thanks to a true functional language and the thousands of available functions.
-
An R Code Editor: unleashes the power of the R calculation platform.
While the Graphical User Interface of AirXCell is sufficient for most of the financial analysis or quantitative research tasks, it may still be convenient sometime to access the core of the calculation engine. The R Code Editor is implemented for this. It enables the user to tweak AirXCell’s calculation, define new functions or rewrite existing ones, etc.
-
A Chart Wizard: A simple yet powerful charting tool.
The wizard is used by the user to chart data displayed on the spreadsheet, a data frame, or any other graphical widget within AirXCell. The user simply needs to select the target data and define the parameters of the chart with the wizard.
-
A Data Frame Editor: data visualization and editing easier than ever. In finance, most
quantitative or financial analysis requires manipulating a lot of data such as, typically, financial timeseries. The Data Frame Editor provides an Excel like User Interface to visualize and edit data frames such as timeseries or any other structured data.
These core applications coupled to the powerful R calculation engine form the buidling blocks on top of which AirXCell’s financial computation applications are built. As of version 1.0, AirXCell supports the following set of financial applications presented to the user as forms within AirXCell:
-
Asset Quote Management: visualize, edit, chart and analyze asset quotes.
A financial analyst uses this form to load a set of asset quotes within his workspace (his session in the AirXCell software). He’s provided with a set of tools to visualize, edit or chart the quotes timeseries and a lot of analysis tools. This form is also used to load the portfolio constituents within the workspace which is a prerequisite to building the portfolio.
-
Forex Quote Management: visualize, edit, chart and analyze Forex quotes.
This form enables financial analysts to load a set of Forex quotes within their workspace. In addition, a set of tools are implemented to visualize, edit or chart the quotes timeseries. One should not that loading a Forex through this form is not a pre-requisites to work with different currencies since Forex Loading is completely automated in AirXcell and performed on the fly when required.
-
Stock volatility Calculation: Compute and chart stock volatility using standard methods.
Stock volatility is a key financial indicator. For instance, the computation of the theoretical price of a financial Option requires its volatility. This form enables users to compute the theoretical volatily of financial instruments.
-
Asset Quotes Analysis: perform statistical analysis on asset or forex quotes.
This form provides financial analysts with various statistical analysis tools on asset or forex quotes such as Standard Linear regression against time, scatter plots, correlation analysis, rolling performances, etc.
-
Option Pricing: Compute theoretical Option price for most common option models.
This form is used by quantitative researchers to compute the theoretical price of financial Options form most option models such as european, american, asian, barrier, lookback, etc. option models.
-
Portfolio Analysis: compute, analyse and chart key indicators on portfolios.
This form makes portoflio analysis easier than it ever was. One can either create a new portfolio, open an existing portfolio or import one from an external portfolio management system for analysis. AirXCell loads the quotes of every instrument in the portfolio and provides the user with means to compute, chart or analyse the following key indicators:
- peformance: what is the value of the portfolio, what is its daily, monthly, yearly, etc. performance (i.e. gain or loss) ?
- exposure: how is the portoflio exposed to a specific aspect such as currency, country. industry sector ?
- risk: what is the risk value of the portfolio ?
Each of this indicator can be computed globally or broken down by one or more variables (such as industry sector, sub-industry, country, etc.).
-
Portfolio Simulation: define a management strategy and backtest a portfolio.
The quantitative researcher can start with an existing portfolio or an empty one, define the constituents of the universe and a portfolio management strategy. Then, he can backtest the simulated portfolio over the past time and see how it behaves. At every moment of the backtesting, he can extract the key financial indicators presented above (Portfolio Analysis Form).
-
Portfolio Optimization: use mathematical optimization methods on the set of a portfolio constituents.
This form enables the researcher to create a new portfolio, open an existing one or import one from an external portfolio management system and performs a mathematical optimization on it. Optimizing a portoflio in a mathematical way means computing the optimized constituent weights in order to minimize the risk, maximize the returns or maximize the returns over risk ratio. Several models of portfolios are supported such as Markovitz’s Mean Variance Portfolios or Mean Conditional Value at Risk Portfolios.
These financial applications are designed from the ground's up with customization and extensibility in mind. Customers can adapt them to their own needs and ways of working with portfolio analysis.
AirXCell provides the low-level mechanisms for easily building RIA Applications coupled to the powerful GNU R calculation engine. It is a tool that overtakes the advantages of the usual products used in Quantitative Research and eliminates their drawbacks (See presentation).
AirXCell is a framework, not an application by itself.
Applications are either modules ...
- Each module is dedicated to a specific task related to Quantitative Research.
- Every module designed from the grounds up with simplicity and ease of use in mind.
- Modules are isolated but can communicate or be built on top of each other.
... or Dynamic Forms
- Dynamic Forms are a possibility offered to users to develop their own apps on top of AirXCell.
- Dynamic Forms are kind of like VBA apps or macros in Excel or UI scripts in Matlab.
AirXCell's current dynamic forms are oriented towards QR in Finance, making it a pretty laboratory for QR involved tasks in Finance.
There are currently only three kinds of software adressing the specific computation needs in quantitative research or financial analysis. .
-
Microsoft Excel
Excel is the major choice of most small financial institutions. Excel is used for financial analysis and even portfolio management. These companies have built their own software on top of Excel using macros and VBA applications (custom application built as extensions to Excel).
Microsoft Excel is great when it comes to manipulate and visualize data. On the other hand, the computation facilities implemented in Excel are rather limited, sometimes even buggy.
Thus, while such applications can achieve the required level of user friendliness, they suffer from the limitation inherent to the Excel calculator wich is weak and very slow. These limitations make it a poor choice to build high-end and evolved computation software.
Yet most of the targeted financial institutions have no real other alternative except Matlab (or other computation software, see below) and are pretty much stuck to Excel.
-
Matlab or other computation software
Some companies are using Matlab (or SaS, Mathematica, etc.) to perform their quantitative analysis tasks or other financial calculations. The problem with Matlab is that it has a steady learning curve, and institutions making that choice are condemned to hire between 2 or more software engineering FTE (Full Time Employee) to support the Matlab codebase and maintain it over time. While this can be an acceptable cost for medium to big companies, it definitely has room for something with a less expensive TCO\footnote{Total Cost of Ownership}.
In adition, custom applications built on top of Matlab suffer from limitations regarding the User Interface inherent to the Matlab platform. Matlab applications mostly suffer from the lack of User Friendliness.
-
Core Banking Software packages
Most banking software packages such as Finova, Avaloq or Temenos support a limited set of features for quantitative research and financial analysis. The problem here comes with the price of such packages. The pricing models of all these software packages is very dependent of the deployed features and the size of the customer. The entry price for these products for a small financial company is unaffordable. Here again there is room for something cheaper.
In adition, the financial computation facilities in these packages are rather limited.
The banking software packages won't be covered further in this document since they really address much bigger companies than our target market, usually companies above a hundred employees. Smaller companies of typically only a dozen of employees have no chance of buying such a package since its way beyond their financial means.
An additional issue with Excel and Matlab comes with the fact that such software run on the end user local computer. Since the user can save his spreadsheets or Matlab scripts on his own computer, there is no centralized management place where the most essential audit and quality checks can be run. Financial Management decisions taken on the ground of Excel spreadsheets or Matlab scripts are not auditable, nor assessable, reproducible, etc.
AirXCell is a pretty unique piece of software in the sense it is dedicated to quantitative and financial analysis. While other software are generaly dedicated to some other task, for instance Excel is a spreadsheet, Matlab a calculation environment, Avaloq a private banking ERP, etc. and yet contains some (rather limited) quantitative analysis features, none of them is dedicated to it as is AirXCell.
In addition, AirXCell is fully customizable and adaptable to fit 100\% the specific needs of a given financial analyst, a quantitative researcher or another. AirXCel makes it easy and fast to perform analysis that the other alternative software make rather complicated, long and difficult, and most of the time only at the cost of integrating several different pieces of software.