Below is the foreseeable roadmap of further Acadia contributions to ORE in the second half of 2024 and in 2025:

Regulatory Capital

• Market Risk Capital (FRTB-SA)
• Credit Risk Capital (SA-CCR, BA-CVA, SA-CVA)

Performance

• Development of XVA Sensitivity Analysis using AAD continues, which is demonstrated in proof-of-concept stage in ORE example 56.
• The GPU interface implementation will be extended to cover exposure simulations
• A GPU interface implementation in CUDA will follow with benchmarking examples vs the OpenCL implementation that is available in release 12

Pricing & Simulation

• The Heston model will be exposed in ORE for pricing of equity derivatives
• The Multi-factor Cross Asset Model based on n-factor Hull-White will be extended beyond its current coverage of IR, FX, COM asset classes, also adding calibration procedures

Consolidation with QuantLib

To facilitate the use of ORE and QuantLib side by side:

• Merge ORE’s QuantLib modifications into the QuantLib project
• Merge QuantLib extension in ORE’s QuantExt library into the QuantLib project as far as possible

ORE Python

• Extension of the open-source-risk-engine scope to cover more of ORE’s classes and member functions. This will be done tactically as we see concrete demand from clients

See some examples here (https://github.com/OpenSourceRisk/ORE-SWIG/tree/master/OREAnalytics-SWIG/Python/Examples)

ORE Academy

The following 3 topics are currently being produced, and these videos will be released in the coming months. They will all come with an Excel spreadsheet replicating what is happening in our C++ library, while also explaining how to configure the relevant XML files:

• Equity Option Valuation (including explanation on the choice of market data and volatility calibration)
• Sensitivity Calculation
• Parametric VaR

We will also release some tutorials showing how to use our ORE Python library in some Jupiter Notebooks, with similar topics:

• ORE Python – Equity Option Valuation
• ORE Python – OIS Discounting Curve Bootstrapping

See the YouTube channel here (https://www.youtube.com/@oreacademy)

---

---

The rollout of formerly proprietary code (Acadia’s ORE+) continues, but new development has overtaken the latter, notably regarding the extension of instruments, pricing engines, analytics, performance, integration and the ORE Academy.

Instruments & Pricing Engines

• Formula-based leg (see Example 64)
• Callable Swap (see extended Example 5)
• Flexi Swap and Balance Guaranteed Swap (see Examples 65 and 66)
• Outperformance Option, Pairwise Variance Swap
• American Swaption with finite difference pricing in LGM (see extended Example 4)
• Finite difference LGM pricers for Bermudan and European Swaptions
• Scripted Trade pricing using LGM with finite difference and numerical integration
• Improved AMC regression model for Resetting Cross Currency Swaps
• Yield curve building with mixed interpolations (see Example 53)
• SABR pricing for Swaptions and Caps/Floor (see Example 59)
• Analytic pricer for Overnight Index Swaptions
• Numerical integration LGM pricer for European Swaptions
• Burley 2020 scrambled Sobol sequence for faster convergence (see Example 56)
• Separate notional payment lags for Resetting Cross Currency Swaps
• Support for indexed cashflows in American Monte Carlo
• Rules-based Bermudan exercise dates

Analytics

• IM Schedule analytic (see extended Example 44)
• Scenario analytic (see Example 57)
• Bond spread imply
• Support overlapping close-out date grid in exposure/XVA (see Example 60)
• Market risk backtest analytic, calculation of traffic-light bounds
• Historical simulation VaR analytic (see Example 58)
• P&L and P&L Explain analytics (see Example 62)
• Stress test in the par rate domain (see Example 63)
• XVA Stress testing (see Example 67)

SIMM

• SIMM 2.7 Calculation from CRIF Files

Performance

The Scripted Trade framework serves as ORE’s Adjoint Automatic Differentiation (AAD) backbone and interface to external compute devices (GPUs):

• AAD is used case by case, where appropriate, to accelerate sensitivity analysis in Acadia Risk Services
• Near term goal is utilizing GPUs to enhance the performance of backtesting, sensitivity analysis (CRIF generation) and exposure simulation in the Acadia Risk Services

Related new examples:

• XVA Sensitivity using AAD, see Example 56
• Fast Sensitivities using AAD and GPUs, see Example 61

Integration

In the Acadia Risk Services we run ORE wrapped into a Web Service framework, deployed via Docker on multiple nodes for industrial scale (“RESTORE”, still part of Acadia’s proprietary ORE+)

In release 12 we have contributed a similar – proof-of-concept – web service version of ORE based on the ORE Python module and the Flask web framework.

Related new example:

• Examples/API, see user guide section 5.0