Why 2020 is Different: FX Risk View on 43 Currency Pairs vs USD

In many risk management practice, especially for those who are dealing with market risk in daily basis, year 2008 must be still lingering in their minds. The great recession of 2008 is so great in effect that it has triggered reforms in many regulatory and accounting standards, in which I mentioned one of them in my last post (Basel 4). The year is also used in stress scenarios of many financial market business, strategy, and even regulatory capital calculation. However, how about 2020?

The Great Pandemic of 2020

Year 2020, now dubbed the great pandemic year, is full of many surprises that we have never seen since World War II. As I briefly mentioned in last 2020 GARP Virtual Risk Convention event, when addressing the question of “how latest AI/ML implementation in dealing with Covid-19 shock?”, I mentioned that for AI/ML based strategies, 2020 is full of Black Swans. Though there can be many more, most notable events are the followings:

1. Q1-2020 S&P 500 30% sell-off in 22-days is recorded as fastest ever in history
2. First time negative settlement price of Crude Oil WTI in history
3. Highest US unemployment figure since WW II

We can see how severe the shock was to AI/ML based funds/trading strategies, to the point that big names, like Renaissance Technology and AQR, reported over 20% loss. Long story short: it was horrible.

“So, how does it compare to 2008?”

Year 2008 is also full of shocks, but they are different in many ways. The damages done in 2008 were mostly attributed from flaws within the financial system itself. For example, the rating of securitized subprime mortgage products where at one point was rated with highest rating from credit risk perspective (AAA), without looking further into the inherent risk of such product. With large institutions leverage their trading books greatly through complex derivatives, the systemic effect was so great to the point that one’s default impact (Lehman Brothers) can be felt by people from other side of the world. However, without undermining 2020 pandemic shock, apparently from FX point of view, 2008 shock is still more significant compared to 2020 shock.

Early this week, I started compiling the information for this analysis, with initial purpose to analyze FRTB (Fundamental Review of the Trading Book) for simple FX portfolio on the two different approach: Standardized Approach (SA) and Internal Model Approach (IMA). In this comparison, I basically would like to see how large the regulated FX delta sensitivity shock (60% or 42.55% – depend on the currency pairs) of SA-FRTB vs IMA-FRTB’s 97.5% Expected Shortfall (ES), for all currencies in which their central banks are members of the BIS (Bank of International Settlement). However, since IMA-FRTB emphasized the use of 1-year stressed period data, observed historically since 2007, I noticed there are 2 possible FX scenarios. After collecting the data for 43 currencies against US Dollars, I noticed that there are only 4 currencies, in which their 10-days 97-5% Expected Shortfall (in return presentation), for either the left-tail or right-tail, are higher in 2020 (full-year) vs 2008 (full-year):

1. US Dollar / Chinese Yuan (USD/CNY)
2. US Dollar / Kuwait Dinar (USD/KWD)
3. US Dollar / Norwegian Krone (USD/NOK)
4. US Dollar / Russian Ruble (USD/RUB)

Key Takeaways

With this information, following actions or conclusions management-wise can be derived:

1. For common trading/treasury desk (including non-financial corporates), 2008 stress test scenario is still generally valid for FX intensive portfolio, despite 2020 shocks
2. IMA-FRTB require each risk factor to take the worst 12 months expected shortfall since 2007 in its minimum regulatory capital calculation. With above information, there are currency pairs in which the worst 12 months periods are not using 2008 scenario anymore.
3. For FX risk class, through comparing the 10-days 97.5%-ES figures for the 43 BIS member currencies vs USD, it appears that IMA-FRTB ES-based equivalent risk-weight is significantly lower than SA-FRTB (42.55% or 60% for most emerging market currencies)

Of course, there can be more analysis and studies to be made on 2020 vs 2008 comparison, especially within risk domain. In relation to the new Basel 4 standards, I expect one of them is the business-case of implementing IMA vs SA, to benefit from the regulatory capital discount, with explanation in point 3 as an indication of this benefit.

___________________________________________

Gior Simarmata is a risk/finance professional located in The Netherlands. He has worked as a senior risk advisor with ING Bank N.V. in the Risk Management Department, responsible for the risk reporting and analysis over the XVA portfolio of the bank’s global derivative positions. He started his career in The Netherlands banking sector with Rabobank as Product Controller in the Finance and Risk Center Department, with trading desks reporting and control responsibilities covering Equity Derivatives, Commodity Derivatives, and Group Treasury’s Structured Notes portfolios. Prior to leaving Indonesia for his master study, he worked with Standard Chartered Bank in Jakarta as a management trainee.

You can reach me by email: giorevinus.simarmata {at} finrisk.eu

This article along with my other articles are also available in LinkedIn and Medium