Monte Carlo Value-at-Risk (VaR)

In today’s highly uncertain commodity price environment, portfolio managers need an efficient and accurate way to monitor the value and risk within their portfolios. Global pandemics, catastrophic supply chain disruption, geo-political conflict, and extreme weather events are just some of the drivers of rapid fluctuations in market prices that can present an existential threat to a portfolio of contracts and/or assets. Having clear visibility into changes in portfolio value and risk as these types of events unfold can be the difference between taking decisive action to cover exposed positions and being forced to permanently close shop.

While a classical parametric VaR approach can be appropriate for portfolios heavy in linear instruments, it falls significantly short in cases where the portfolio contains options, complex structured transactions (e.g., renewable PPAs, tolling agreements, full requirements load contracts, etc.) or physical assets. In these cases, the closed-form approximations of value required by a parametric VaR model can mis-state risk and leave the portfolio exposed.

A simulation-based approach that takes into account the important nuances of complex contracts and physical assets is required for a more accurate picture of VaR. Such an approach allows more accurate modeling of prices and is not bound by a priori assumptions about the nature of the distribution of price returns. Conversely, a parametric VaR model typically assumes price returns are normally distributed; an assumption that is typically not the case in energy and commodities prices. Additionally, a simulation-based approach allows each portfolio element to be re-valued for each simulation path, yielding a distribution of future value outcomes that truly reflects optionality, hourly shaping and volatility (e.g., in the case of renewable assets or contracts), contractual nonlinearities, and even physical asset operational characteristics (e.g., for batteries or thermal generators).