# SACCR vs CEM for FX Products

Blame our active community of readers if you must. But we’ve had the most requests for a comparison of FX products under SACCR and CEM in response to my original blog.

Therefore, here we go…

## SACCR is Coming

If you need a refresher on SACCR and CEM, then please check out our comprehensive coverage below:

Suffice to say that we expect SACCR to be live by January 2022 at the very latest across all jurisdictions and even earlier for leverage ratio calculations:

## SACCR for FX Products

When we delved into the SACCR calculations, we found that it is necessary to start from the bottom-up when calculating Risk Weighted Assets under SACCR. However, I realise that not all of our readers want the absolute detail. Fortunately we provide a simple Excel add-in that calculates Exposure at Default under SACCR.

SACCR for Excel has free trials available here.

## SACCR for Excel

I will use our SACCR for Excel tool to demonstrate some simple portfolios. Alongside the calculations I will do some manual comparisons to what the equivalent CEM number would have been.

## SACCR for FX

It is useful to use some simple examples in FX to illustrate the main parts of SACCR. Crucially, FX in SACCR has only one maturity bucket (very different to CEM). This increases the netting opportunities under SACCR.

Let’s start with 3 NDFs in USDBRL, all uncleared and with 3 different counterparties. They operate under a daily CSA with a zero threshold:

Showing;

• For each NDF, the SACCR “Exposure at Default” is shown. For CEM, the “Potential Future Exposure” is shown. These are the two key measures that dictate the Risk Weighted Assets (RWAs) under each methodology.
• You can see that the CEM calculations are very simple. It is a given percentage of notional, which varies by asset class and maturity. For FX, these vary from 1% (sub 1 year) to 7.5% (5y+).
• I’ve shown NDFs, but these could be any product under which the main risk factor is FX – such as cross currency swaps.
• For FX options, the calculations are the same under CEM. Under SACCR there is an adjusted notional for option products, which our SACCR for Excel product calculates for you.
• For SACCR, there are also additional variables to consider. Check out Section 52.51 of BCBS CRE52 here, which I summarise below:
2. If not, what are the details of the bilateral CSA?
3. Does it have daily, weekly, monthly margining?
4. Are there any disputes across the CSA?
5. Are there any “difficult to replace” derivatives in the CSA?

The answer to those questions dictate the variable in SACCR known as the “Margin Period of Risk”. There is not an equivalent to compare to under the Current Exposure Methodology.

## Margin Period of Risk

What we can do for SACCR is try to demystify what the Margin Period of Risk really does. If you think of a SACCR FX exposure as being approximately calculated as:

$$\tag {1} EAD = 1.4 \times NetNotional \times 0.04 \times \dfrac{3}{2}. \sqrt{ \dfrac{M POR}{250}}$$

where;

1.4 is the regulatory defined value of “alpha” (α). This multiplier was completely absent from CEM.
0.04 is the Supervisory Factor (“risk weighting”) for FX of 4%. If we were looking at Rates, this would be 0.005 (0.5%).
$$\dfrac{3}{2}. \sqrt {\dfrac{M POR}{250}}$$ is the “Maturity Factor” which can change depending on how the trade is margined/cleared.

To understand the implication, it is therefore useful to keep a handy reference of the Maturity Factors that can apply to your trades depending on the CSA in place. The values range from:

Optimising the terms of CSAs will therefore have a significant impact on regulatory capital. Disputes and “hard to replace” derivatives are bad, daily margining and clearing are good!

## Clearing

Away from the intricacies of the margin period of risk, the biggest single difference for dealer banks will be the use of net notional when calculating SACCR exposures. Previously, CEM was always based on gross notional and only allowed for a limited amount of netting – up to 40% based on the mark to market of the derivative, not the direction (bizarre!).

The impact of netting is going to be most evident for cleared trades. If we take the 3 NDFs above and move them to LCH ForexClear (other CCPs are available!) we see the following SACCR and CEM exposures:

Showing;

• For a portfolio where I have sold USDBRL in 4 years and 10 years and bought it back in 6 months the amount of netting is significant.
• The SACCR exposure reduces from $672k down to zero (!) for this stylised example. • Note that the CEM number doesn’t change. • This is significant for leverage ratio. Under SACCR, clearing will be far more capital efficient than it currently is. This is mainly thanks to a clearing house being recognised as a single “netting set” under SACCR. ## Directional Portfolios What I find interesting under SACCR is the impact to directional portfolios. SACCR is intended to be a more risk-sensitive measure than CEM. For our simple portfolio, let us assume that I am a seller of USDBRL across all maturities, and that we move them to clearing. We see the following exposures under SACCR and CEM: Showing; • The SACCR exposure does not change. It is the same at$672k, whether these trades are cleared or bilateral (assuming an MPOR of 10 days under both scenarios).
• SACCR is significantly lower than CEM – as it has shown to be for all of these comparisons in FX.

If you have a a lot of FX in your portfolio, it is therefore likely that moving from CEM to SACCR will decrease your exposures and hence result in a decrease in regulatory capital requirements.

## Optimisation in FX

I am going to leave it there for now so that we do not get lost in too many intricacies. However, we can think of more ways to optimise exposures in FX:

• An FX “risk factor” could be neutralised across a whole netting set.
• STM (Settle to Market) can be used under SACCR to reduce exposures. From the latest ForexClear rulebook in April 2019 it looks like STM is now live for FX contracts at LCH.
• I assume that there is optimisation work to be done by taking a net exposure in FX versus one counterparty and converting it into a G10 NDF at a CCP. That should increase netting in dealer portfolios and reduce the MPOR thanks to STM.

Let us know if you want to see a run-down of STM under SACCR across both Rates and FX.

## In Summary

Finally, some SACCR FX soundbites:

• SACCR will reduce exposure measures for FX portfolios compared to CEM.
• SACCR is sensitive to the Margin Period of Risk for collateralised trades. This can range from 10 business days to much longer (even 40!).
• Clearing substantially increases netting, which should lower SACCR exposures for non-directional portfolios.

Stay informed with our FREE newsletter, subscribe here.

## 4 thoughts on “SACCR vs CEM for FX Products”

1. JC says:

The CEM PFE is incorrect due to the PFE factor applied to the 6m and 4y trades being switched, CEM PFE should be 1.85mm vs 1.45mm shown
10y 10m * 0.075 = 750k
6mo 10m * 0.01 = 100k
4y 20m * 0.05 = 1,000k

1. Chris Barnes says:

You are quite right, sorry for the spreadsheeting error and great spot!

2. Hayri Konuk says:

First of all, thank you very much for all of your efforts, for such clear explanationsof SACCR. However, I couldn’t see any example related with listed derivatives such as futures & options on indices or fx underlyings at anywhere, especially from the side of CCP; to confirm my understanding. So, I think it can be very helpfull if it’s possible to give such an example, from the side of CCP.