Thursday, November 1, 2012

OIS Discounting – A lot to Consider

A lot has changed in the banking world over the last 5 years, so it is not surprising that the way banks value future cash has come under a lot of consideration. The discounting of future cashflows is fundamental to understanding the value of any financial product. After all, the discounted value is the price at the end of the day. The old cash and carry arbitrage argument tells us the right rate to use when discounting is the cost of funds.

Under Libor discounting, there was the assumption that banks can fund at Libor which is not unreasonable if you fund exclusively at say 3m Libor, but we know that this is just not the case. Banks these days should be trying as much as possible to their assets and liabilities, and that means the true cost of funding is somewhere down their own curve. There is an argument that unsecured counterparties should be discounted at the firms cost of funds, although in practice I don’t know how widely this idea has been adopted. If however a Credit Support Annexe (CSA) is present in an ISDA, and collateral is posted against an exposure then story is different. Typically in the case where cash is the collateral choice, the reference rate used to pay interest on the collateral is the Overnight Index Swap (OIS). It follows then that it should be the OIS curve that is to be used for discounting future cashflows.

Another way of thinking about this issue is to consider the credit implications of posting collateral. CSA’s are an excellent way of mitigating credit risk and in its purest form where net exposures are margined daily, no residual credit risk remains. To a financial institution this implies the future cashflows should be values at the risk free rate of return, and that exactly is what the OIS rate implies.
OIS Discounting presents many challenges to a bank. The idea that you have a 40+ year OIS curve in some markets is difficult to imagine, even for developed markets. The Australian market there are Bills OIS Basis (BOBs) quotes to 30 years, but to my knowledge there have only been a handful of trades beyond 3 years. For emerging markets, the idea of a long dated OIS curve is a difficult question to ponder as some markets have no OIS market at all, or the quoted market is only for a few months.

Many CSAs specify collateral to be posted in a currency that is not the same trade in question. For example we have a Sterling swap, but collateral will be posted in EONIA (EUR OIS). The curve needed to price such a swap is the EUR OIS in GBP curve. The implication of doing this however is we will end up with risk to GBP and EUR currency basis and EONIA. From a traditional risk system we would see sensitivity to EONIA.GBP, but a trader however cannot directly hedge against this curve. From a traditional risk system we would see sensitivity to EONIA in GBP, but a trader however cannot directly hedge against this curve. The risk needs to be decomposed in to the underlying risks of GBP basis, EUR basis and EONIA. This is not just a systems issue; it requires a change in thinking from the trader, risk management and the other support functions like finance and middle office. Management also needs to be on-board as well as it is likely there will be a P&L impact.

There has been quite a bit of talk about standardisation of CSA terms. One of the ideas is to adopt the London Clearing House (LCH) approach where you fund in in the same currency as the trade. Certainly posting collateral in the same currency as the underlying transaction has benefits as the trader would not have any cross currency basis risk or FRA/OIS risk in another currency as described above. While this may have some operational benefits, it might not be desirable for minor currencies. The LCH approach has a concentration of risk in specific currencies, and this creates an issue currently not properly comprehended. We saw during the global financial crisis there was a squeeze in USD and US banks had to raise USD via the cross currency market. Now what would happen if LCH clients get margined in a currency they cannot raise? OK, having a squeeze that ended up in a complete freeze in USD may be unrealistic, but it could happen to a minor currency. The LCH approach is assuming all parties have complete access to cash in each currency.  So what is the alternative? Multi-Currency CSA prevents this type of scenario from happening. This is where you have a choice of different currencies to post. There are a number of issues still outstanding in respect of pricing with these types of agreements, such as optionality at the point where one might switch collateral. However such agreements do provide organisations the flexibility to fund in a way that suits them.

The challenges for any financial institution faces with an OIS migration are many. For banks, understanding the true cost of funds for each client, creating long dated OIS curves, changing the discounting on every cash flow, translating the risk back into driver risk and making sure traders, risk managers, accounts and business management understand what is happening are going to be issues. How the organisation decides to manage the risk will be telling as well. A smart bank should be thinking how they can make money out of this. The transition to OIS discounting will leave the market less than perfect and that provides opportunity. Those who understand it the best will make money.

For price takers such as fund managers, there is opportunity as well as sensible selection of counterparties will make a difference. Also understanding how banks value your trades will put you in a position where you can have a sensible conversation about them. In some cases the best thing to do might be to renegotiate your CSA terms.

I think OIS discount is the biggest change to the swaps market in 30 years. Banks in Asia and Australia are yet to fully embrace OIS discounting, but that does not mean there is time for contemplation. Slow adopters are going to be the ones that miss out on the opportunities it presents.  My white paper goes into more details on these issues, but it also looks at the risk implications of cross currency swaps and the risk aspect under OIS discounting. There are many issues to consider and Maroon Analytics are able to help institution that is being confronted with OIS discounting.

No comments:

Post a Comment