EBA Stress Test Arbitrage

Question

The EBA stress test defines specific shocks to yield curves that are applied to positions as at year end. There is no account for cashflows - it is simply an immediate shock.

Suppose the interest rate shock was 50 bps which is a reasonably large value for some interest rate markets, then the purchase of a 50bps wide strangle with an expiry that is soon after the year end would cost very little, yet it might entirely hedge 25bps of the shock in either direction.

Whilst I am sure that regulators would disapprove of the direct specificity of this trade, my question is whether anyone has any experience with similar structured products, in terms of regulatory hedges and if they are indeed permitted?

Answer 1

I would not be surprised that you can perform some regulatory arbitrage by mean of little financial engineering as you suggest, see for example: https://www.risk.net/our-take/7046041/in-stress-test-window-dressing-timing-is-everything.

In principle, why wouldn't you be allowed to take such a market position? It is a default of the regulatory approach. In the above linked articles, it is said that regulators try to prevent this problem by specifying a "random" window for stress tests a posteriori.

Note however that the picture might change with increased and periodical regulatory oversight such as with the Fundamental Review of Trading Book.

My guess is that these issues are indeed existing with any regulation that takes the balance sheet at a specific date. You could also look for example at Solvency II on the insurance side.

As a financial institution, you are taking a serious risks by doing such arbitrage: - What happens if the regulation changes rapidly and you are not able to do the trick anymore and you end up with inadequate capital? - What happends if the same cheap regulatory hedge is no longer that cheap later on and you end up with inadequate capital? (I am more familiar with insurance where your balance sheet can't change that rapidly)