[DcifrThs]

[ QUOTE ]
Anyone have any actual experience in Bond Trading. Preferrably not something like along the lines of buying and holding A or AA securities to diversify a portfolio. Any recommendations for books are welcome as well.

Basically, I pretty much know all there is to know abuot options and can't really learn more until I really start to trade them often and find out what market dynamics are really like. With bonds however, I know the mechanics of course, so I know how they're priced and what affects them (credit risk, yield curve, etc.), but I don't really know what makes bond trading profitable.

Any input/experience is welcome. I made my first bond trade on a countrywide bond that is basically a bet on their bankrupty probability being overstated as well as an expectation in fed rate cuts affecting the yield curve. Equities bore me and I feel like an analytical person can outperform the market using bonds moreso than using equities.

[censored] equities.

[/ QUOTE ]

bond trading is just like anythiing else.

you see what the market is pricing in and then determine whether you agree or not and with what level of conviction you disagree (or agree).

for instance, looking only at nominal bonds, the yield curve is the main determinant of the prices of the bonds. if interest rates are expected to fall, holding bonds will result in a profit, if they are expected to rise, holding bonds will result in a loss.

that may seem obvious but it is the base. since you say you know how to price a bond, i'll assume that you know that back and forth.

so now look at the yield curve and how it has shifted over time. many profitable bond positions come from yield curve adjustment bets rather than just direction bets. these relative trades are sometimes called "steepeners" (for long the short end and short the long end of the curve) or some other name for the reverse bet....betting the curve will become less steep or invert.

knowing about duration and convexity is also important since that affects the overall risk you hold in your portfolio.

nominal bonds have 1 source of that (interest rates). convertible bonds, corporate bonds, inflation linked bonds etc. all have things that make the convexity and duration different from that of nominal bonds. you can make bets that these additions (i.e. the option to prepay in MBS) are overvalued or overvalued by shorting out the default risk (betting that fewer people will prepay mortgages than expected and thus MBS spread above treasuries will fall).

other types of relative value bets (or diff bets) involve either corporations or countries where you think one's rate will move more than the other's beyond what is priced into the market.

for instance, if you think that the US rates are too low and UK rates are too high on the near end of the yield curve, you can short the US tbill and long the UK 1yr bill (i don't think that is called a Gilt but i forget the name).

these positions need to be normalized to the same level of risk via duration and convexity analysis so that one leg doesn't move proportionally far more than the other.

so those are some examples of bond trading off the top of my head.

feel free to ask for more.

now in terms of the position you mentioned, i think you have gone about executing it incorrectly.

you have 2 views:

1) your bond's credit risk is overstated relative to treasuries

2) some theory about the future of interest rates (do you think the fed will cut more or less than priced into the market?)

when you have 2 views, you shouldn't execute them in 1 bet. you should execute them individually and take into account their correlation and individual riskiness in your actively managed portfolio.

for 1), you are betting that the compnay will be more able to pay back debt and less likely to default than priced in. therefore you are long the spread. so you purchase that bond and sell the US treasury security of similar duration (since company bonds typically have longer duration than treasuries you can either purchase more of the 10yr or maybe X*10yr + Y*30yr to arrive at some average between 10 and 30 years depending on yoru duration estimates. i'd think 15 yrs of duration should be ok for corporate bonds but i'm not sure about that)...or as stated, purchase enough of the shorter duration piece to equalize their moves in risk space.

for 2), you are simply directionally long or short the yield curve however you view it. you can execute that view in the US treasury market, the eurodollar market, the fed fund's futures market, options on fed funds futures etc.

this bet may correlate highly with your other bet depending upon the company's current and projected ability to pay down its creditors.

this post is probably getting long so i'll stop there.

hope this helps.
Barron

Barron