Research
Articles on AI and currency exposure
A reading list of editorial pieces on how machine-learning methods are applied to currency exposure inside international portfolios. Each article sets out one mechanism, the data behind it, and the model's limits. Nothing here is investment advice; nothing is sold.
Translation drag · Method
Decomposing a foreign holding into local return and FX return
A model can split the performance of an overseas position into what the asset did in its own currency and what the exchange rate contributed. This piece walks through the decomposition and why a single blended figure hides the part that matters for hedging decisions.
Hedging cost · Forward pricing
Reading the carry that is already in the forward
The interest-rate difference between two currencies is embedded in the forward price, so a hedge is never costless. Here we look at how that embedded carry is measured and when it makes an unhedged position the more honest choice.
Volatility regimes · Clustering
Marking volatility regimes that move a hedge ratio
Clustering foreign-exchange volatility surfaces periods where a static hedge ratio leaves a portfolio under- or over-protected. The article sets out how the clusters are built, what they miss, and where the method tends to fail.
Base currency · Reporting
The base currency as a moving reporting lens
Reported returns change when the currency results are measured against shifts underneath them. We examine how models account for that movement across periods, rather than freezing the base currency in place.