No Margin for Error - FX Week

The days when margins could be influenced by the quality of the lunch are long since passed. The credit crunch has provided numerous examples of how important it is to get the margins right and perhaps made it easier to understand why these margins are required. Simple transactions like FX Spot, where the major risk is settlement, have seen major initiatives such as CLS and more talk of central counterparty clearing, so margins are narrowing and not much of an issue. For the Money Markets (MM),

This may be an over simplification of the problem, but nowadays MM margins need to be applied for a host of reasons. They have to allow not only for pure credit risk, but also for market risk, regulatory costs imposed by central banks, return on capital, operational risk and if the banks are honest as a payback for what they lost when the margins weren’t as scientific. Whilst there is outcry from some areas whenever a bank makes a profit, we’re all happy to see our bank shares and the FTSE as a whole benefiting from these profits and increasing. We’re happier seeing their ‘boat coming in’ rather than their ‘northern’ boat being ‘rock’-ed.

In the modern age, where system memory and latency are taken for granted, it is easy to calculate (sometimes using complex algorithms) numerous margins for one deal, our current record is six margins on one deal. In fact it’s not that long ago, when in a meeting with a customer, he mentioned ‘Tennis Court’ margins on loans. I thought I had misheard him, but he went on to explain, “if you have a tennis court in your garden and a neighbour wants to come round and use it, that’s ok, but there has to be some reciprocal benefit for you, otherwise you politely ask him to take his balls elsewhere”.

It is however important to ensure that your systems functions in the correct way. Numerous banks have spent small fortunes on credit and margin systems, to then decide that they can’t be hooked up to trading systems because they’re too slow, leaving the bank to rely on carve out limits and over-simplified margins. Oh, speed isn’t a problem for MM the crowd shouts, but it is. MM prices are available over several ECNs and as we all know, the first half decent price is the one that gets dealt on.

Vendors need to make sure that their systems can calculate margins in real time, fast enough to allow their customers to be able to connect to ECNS, even for spot trading or they risk being left out in the cold. We also think it is important to have independently verified response times in single milliseconds, something which we achieved with our Siena product range when it was verified by Sogetti. Using our business process engine we can combine numerous factors (currency/currency pair, product, tenor, time zone, amount band, credit rating, onshore/offshore, tennis court fee, etc.) and still produce one or several margins in real time.

There is another dilemma for the banks. Where is the real value in the deal? If scientific evaluation of the real cost and risk of lending means bigger margins, will the banker who sticks his finger in the air, quoting a smaller margin, get all the deals. Or more to the point, if he gets more deals, does that bring economies of scale and improved profitability or does it just bring ruin ever closer? One of the major decision factors has to be, if you buy in a corporate’s borrowing what other business comes free on the back of it? This is where relationship and knowing your customer plays an important part in margins.

Then we come to the really difficult issue…deposits. Well, we all know a bank can’t exist without depositors, but it can also lose money if it hasn’t got anyone to lend it to. If someone from Goldmans wants to deposit their bonus with them, they have a dilemma – will that mean they make a profit or a loss? There are occasions when banks want to decline or at least penalise large deposits by charging the depositor for the use of credit lines. Margins on deposits are more important in an environment like the one we have now. If you have to use valuable credit lines to lend out the money, you need to offset the cost of hedging by charging the depositor.

As the penny drops it becomes increasingly important to tackle the chicken and egg question about which is the transaction which bears the brunt of the margin, the deposit or the loan. Traditionally it was always easier and more acceptable to bump up the price of the loan, but is this still the case? Are we closing in on an era when margins will actually be influenced by the bank’s position?

We were recently asked to add in a feature to our Siena system whereby margins for a specific product (e.g. overnight sterling loans) could be suspended manually for a short period. The reason behind this was to help with day to day balancing of the books. It gets later in the day and you still have an overnight balance to settle. In this instance why pay away the spread and just deal out the balance when you could do that deal with one of your customers, simply by offering them a market rate rather than the usual margined one. Surely it brings goodwill and hopefully there is some profit left in that.

To sum it all up, margins are likely to get more flexible and dynamic. They need to reflect the true cost of hedging the deal, be it a loan or a deposit, and they must reflect the credit and other risks to the bank. It must be possible to amend margins easily, even if it’s only for a period of minutes, and it must be possible to split the margins into accountable pieces reflecting the different reasons for taking them and who in the bank benefits from the margin.