Day to Day of Investment Grade Credit Trading

You need to start with a very specific market/security sector, and then work your way from there. So, let's start at the beginning...US Treasuries.

The US yield curve has 6 liquid cash points (let's ignore Bills, 2yr note, 3yr, 5yr, 7yr, 10yr, 30yr - mostly traded in BrokerTec and eSpeed), and 5 liquid futures points (ZT, ZF, ZN, ZB, UB - all traded at the CBOT). Let's also ignore fed funds futures for now.

Typically, each liquid cash point has a marker maker trader that specializes in that sector of the yield curve. For example, the "bond trader" trades the 30yr, and all off the runs down to the 10yr (and also some old 30yr bonds with high coupons that have rolled down into the curve).

Part of being the bond trader is understanding how to fairly value off the run bonds as they roll down the curve. If you plot the yields of all the off the run securities, should they form a straight line, a smooth curve, a smooth curve with kinks, etc...? Exactly what shape should the curve look like? If the shape of the curve changes, should you try to hammer it back to how it was, or are you always just accepting how other traders price the curve? These are the sorts of questions that you need to answer.

You never know how long it will take for the imbalance to be fixed...it depends on the liquidity of the market..ideally its the next day, but sometimes it takes months or gulps years. "Relative value" means that you are not outright long or short on the direction of the rates market...but rather, you are taking positions on the minor shape of small sectors of the curve.

Quants come up with yield curve models to help traders understand what the curve should look like without any special supply-demand factors. Some popular models are the [cubic spline, tension spline, 2+, Affine, linear interp, cash blow based spread to swap curve, the zero curve, par yields, SPB, etc...).

Then the trader takes that as a baseline and adds to it the knowledge of what has been trading in the market (supply and demand for specific issues) to determine if there is a trade to do or not.