Risk Factors for Historical VaR on Currency NDFs

Summary

The question asks which market variables drive the value of a currency non-deliverable forward (NDF) and therefore belong in a historical simulation VaR model. It expresses the NDF value using the forward exchange rate, strike, domestic discount factor, and notional, then considers representing the forward rate through spot and swap points. The proposed risk factors are spot, forward FX swap points, and the domestic risk-free rate, with correlations among their changes also raised.

The answer says NDFs use the same risk factors as vanilla FX forwards. It explains that an NDF settles a cash amount, commonly in US dollars, based on the exchange-rate difference rather than delivering the restricted currency. It points to a separate discussion of FX-forward risk factors but does not reproduce its details. The post raises a question about whether the currency-pair convention should be switched for USDCNY; the answer does not directly address that convention, specify a VaR construction, or provide data or empirical evidence. Readers must consult the referenced FX-forward discussion and confirm how their pricing and settlement conventions define the relevant exposures.

Key ideas

• NDF valuation depends on the forward exchange rate, strike, domestic discounting, and notional.
• The answer treats NDF risk factors as the same as those for a vanilla FX forward.
• An NDF settles a cash amount based on exchange-rate differences rather than delivering the restricted currency.
• The post raises, but does not resolve, how currency-pair quotation conventions affect risk-factor setup.

Tags

# fx : HistSim VaR for currency NDFs


# fx : HistSim VaR for currency NDFs












I might be asking a very simple question for the FX experts...

I would like to check the HistSim VaR process for a currency NDF.

Given the PV for the product is defined as:- (f(T) - K)*discFactor(DOM,T)*N

```
f is the forward outright for the NDF ccy pair
k is the strike
T is the maturity
discFactor(DOM,T) is the domestic discount factor
N is the foreign ccy
```

This can be re-written as = (Spot + Swap_point(T) - k)* discFactor(DOM,T)

Am I then right in saying that the risk_factors are:- 1) Spot 2) the forward fxSwap point 3) domestic risk-free rate ?

For VaR, I will also need the correlation between each of these risk-factor changes

Also, for USDCNY where base = USD, I am guessing I need to switch from Base=USD to Base=CNY

Thanks.

## Answer by rupweb (score 0, accepted)

https://quant.stackexchange.com/a/31817

The risk factors for NDF are the same as for vanilla FX forward. The thing about an NDF is that all that changes hands on value date is a cash flow calculated according to the difference in rates since trade date, usually in USD. That means that you can "trade" the instrument offshore and you don't have to comply with onshore trading regulations...

Looks like Matt Wolf already answered the question about risk factors for FX forwards here

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.