wilmott.com

This should be fairly straightforward for all the fixed income experts at this forum but not for me. Someone described me the following transaction:A) Major IB issues on a behalf of a client a callable 10-year bond linked to euro cms spreads. The payoff has a typical structure, say for example, Year 1: 6%, Year 2: 5.5%, Year3-Year10: 4*(CMS10-CMS2). This structure was issued at a price of 94.B) The IB receives from the client Euribor - 18bp annually and pays the client 100 at the issue date. And of course, the bank undertakes the obligations of the CMS note. What is the trick in these structures? How the bank is willing to pay the client 100 while it collects 94 from the bank at inception? Given that this structure was initiated at the beggining of 2006 it seems that the bank is making some good money under the current state of the euro curve. But how the bank was so confident to set up the structure at first place?

What one needs is to price the CMS spreads structure by modeling directly and hopefully correctly the monetary policy scenarios. I just wrote a note on this and posted it on my web site at http://www.level3finance.com/publications.html. See fig. 10 for this type of analysis from the structuring viewpoint.

Very interesting paper. In my previous job we developed a similar model to exclude unrealistic curve shapes, given the current state of the economy, while preserving no arbitrage.I dont think the world is ready yet for this line of thought...maybe hedge funds may adopt it for better relative value analysis. On-the-desk modelling at sell-side institutions is, unfortunately, dominated by brainy science types who have an aversion against putting "fuzzy" economics into "precise" models.KLEON:To answer your question...the bank (probably) hedged most of the risks away at the beginning (and dynamically rebalances to this day). There is no trick. The client, for whatever reason, believes it is a good idea to buy this structured note. Clients generally do not hedge...this happens for many reasons, most notably because if they had the models, they would just buy the underlying vanillas themselves. Other reasons may be investment constraints (they cannot buy vanilla derivatives...but they can buy funky bonds, as long as they look like bonds). In any case, the bank charges the client a spread for taking on the risk and for having to manage the deal in the future. Without the spread, the structure may well have sold at 92! Now it is up to the exotics desk at the bank to "protect" this 2 point upfront margin throughout the life of the deal, until it matures. To do this, the desk will go into the vanilla market and hedge the rates risk, the vega, and the correlation risk of the 2-10 spread.