SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
rcohen
Topic Author
Posts: 252
Joined: November 15th, 2001, 12:06 pm

model for yield curve

February 25th, 2006, 9:24 pm

Although paper accepted, would still appreciate comments on the approach.Thx!
Attachments
YieldCurve.zip
(140.74 KiB) Downloaded 26 times
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 25th, 2006, 10:25 pm

I didn't read it too much, because I expect to get stumped by the math anyway. But I am curious what this model can be used for. Option pricing?If I understand right, the reason you need to come up with some sort of economic dynamic is so that you can describe the volatility at different spots on the curve, and their relationship to interest rates, in as few numbers a possible? Put differently there's correlation between movement in the different variables, so things don't move as freely as if they were all independent, but instead move at the same time?At the end, you listed some disadvantages. What sort of problems will you run into as a result of these disadvantages?Also, you mentioned there's a lot of money to be made with a good model. I keep hearing that. How?Thanks for posting the paper.
Last edited by farmer on February 24th, 2006, 11:00 pm, edited 1 time in total.
 
User avatar
SanFranCA2002
Posts: 649
Joined: October 3rd, 2002, 5:05 pm

model for yield curve

February 27th, 2006, 5:09 pm

Sorry. I had a really really bad weekend so I'll take it out on you. I think these sorts of papers can be bought by the pound and I don't see anything there that adds value. What is there that allows one to understand the dynamics of the yield curve? Here in the US about a trillion dollars walked in the door last year (mostly from Asia) to gobble up our debt, driving down long term rates and inverting the curve. Where is that in your model? What volatility number are you referring to? Mine? Fund managers in Boston or China? Are we all the same? You have some kind of linear function (nothing in finance is linear and though you don't get into corporates for some bizzare reason the risk premium as you go down the rating spectrum is incredibly non-linear) and use the Sharpe ratio, but even Mr. Sharpe has disowned that measure. And where the hell is inflation? I skimmed this fast but where is good old Fisher and dividing up the real rate of interest and inflation expectations? And when the stock market goes to hell and people put their money into bonds, where is that in your model? That happens every 5 to 10 years or so. And papers that mix arbitrage theory and economics really really gets on my nerves. Arb theory is true by mathmatically proof and you talk about going long this and short that or whatever. And then authors glide into economics without acknowleging that none of the remainder of what they are saying is true by any arb arguments. OK. I'm done and feel better now. I'm sure people will read your paper and nod their heads as if this helps them understand anything.
 
User avatar
bigslick
Posts: 101
Joined: October 4th, 2005, 2:01 pm

model for yield curve

February 27th, 2006, 6:27 pm

SanFran, i believe rcohen adds a disclaimer right at the end of his paper..."We conclude here by reiterating that this is only a model, which, like any other,has its advantages and disadvantages. One disadvantage is that that it is not dynamic,although it has the potential for serving as a platform for a more sophisticateddynamic or stochastic setting."..which means he has acknowledged the limitations of his model with the examples you described.Keep posting your ideas people and don't be discouraged by the thought your paper may not be appreciated by the few who have had a bad weekend
 
User avatar
SanFranCA2002
Posts: 649
Joined: October 3rd, 2002, 5:05 pm

model for yield curve

February 27th, 2006, 6:40 pm

Oh, I did appreciate it. The paper had incredibly strict constraints, and given those constraints it was wonderful. Unfortunately, the same (or a little more) effort could have addressed the current DYNAMIC situation in the markets. This right now is about as interesting as it gets with the Fed doing one thing, Asian investors another, the interest rate impact on the slowing of wealth generation leading to a possible recession starting in the next year or two, and those are just a few issues. The model just frustrates me for that. As I've heard quoted, "The worst hell on earth is neglected opportunity."
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 27th, 2006, 6:40 pm

QuoteOriginally posted by: SanFranCA2002And when the stock market goes to hell and people put their money into bonds, where is that in your model?This is why I'm curious whether this model is for pricing options, or what. Because I don't think I can find what you described in Heston's model either.
 
User avatar
SanFranCA2002
Posts: 649
Joined: October 3rd, 2002, 5:05 pm

model for yield curve

February 27th, 2006, 6:43 pm

The purpose of the paper was to write a nice safe "get your feet wet" kind of thing to move in the direction of more ambitious projects in the future. His web site list his papers are more about physics, and now he is getting interested in economics.
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 27th, 2006, 7:43 pm

So he's just trying to propose factors X and Y, and see how moves in these factors manifest in different yield curves?
 
User avatar
cosmologist
Posts: 640
Joined: January 24th, 2005, 8:08 am

model for yield curve

February 28th, 2006, 12:55 pm

QuoteOriginally posted by: SanFranCA2002The purpose of the paper was to write a nice safe "get your feet wet" kind of thing to move in the direction of more ambitious projects in the future. His web site list his papers are more about physics, and now he is getting interested in economics.I wonder why should physicists be interested in economics? Cohen woks for citigroup,must be well-paid.he should have taken permission from this sanfrancisco2002 fellow before getting this paper published.I think he should have taken permission from economists and finance guys before changing his field. Why should he be better paid. he should have been satisfied with whatever pea-nuts he would have got in physics.sanfranc2002, i think you must have published some 'cutting corner' research . may we know one of them,please?farmer is also participating in the drama.hehe.cheers
 
User avatar
rcohen
Topic Author
Posts: 252
Joined: November 15th, 2001, 12:06 pm

model for yield curve

February 28th, 2006, 6:12 pm

SanFran,My conclusions after reading your comments are that: (1) you either didn’t read the paper, or (2) you just plain didn’t understand it! Why? To quote you: “And when the stock market goes to hell and people put their money into bonds, where is that in your model?”This is covered in scenarios 2 and 3, both of which involve getting rid of the underlying security (stock) and investing in bonds (short or long term).I can suggest three things before I will even consider replying to your comments next time: (1) find some weekend that suits your moods, (2) read the paper and, most important, (3) understand it before you start spitting out any nonsense!As for your last comment, which I quote:“His web site list his papers are more about physics, and now he is getting interested in economics.”You’ve got a problem with this?! I challenge you, or any other economist for that matter, where ever you think you sit on the intellectual ladder, to come up with an original, publishable idea in physics – and I don’t mean “econo-physics”!farmer,This is not a factor model with factors X & Y and was not intended to be one.You’ve made a point, however, on the applicability of this to option pricing. I thought about this while writing the paper and I concluded that it is not applicable. The reason is explained in Section 3, titled “Limitations”. The model basically works for maturity time scales longer than 1 year, which is over the practical range of most typical vanilla options, such as calls and puts. You’ll find something on this in an earlier Wilmott thread discussing the tenure limits of an equity option: http://www.wilmott.com/messageview.cfm? ... lick,Where would the world be without people like you?! Also, thx for pointing out the “disclaimer”. Another indication that SanFran hurled out his comments without reading the paper!
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 28th, 2006, 6:27 pm

If you can't use it to trade options, how do you make money with it? Some kind of arbitrage between stock prices, bond prices, risk premiums, leverage, and implied volatility?
 
User avatar
rcohen
Topic Author
Posts: 252
Joined: November 15th, 2001, 12:06 pm

model for yield curve

February 28th, 2006, 6:38 pm

QuoteOriginally posted by: farmerIf you can't use it to trade options, how do you make money with it? Some kind of arbitrage between stock prices, bond prices, risk premiums, leverage, and implied volatility?The financial markets are not made up only of equity options. They involve fixed income also, where the yield curve plays a dominant role. If you could come up with a decent yield or swap curve model - any model - that increases your chances of winning by even 0.01%, you can become rich!
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 28th, 2006, 6:41 pm

I'm confused. You can't use it to trade bond options, right? So what does your model enable you to do? What inputs do you use to come up with prices, and for what?Do you use the price of a three-year treasury note to figure out what price to pay for a two-year treasury note?
 
User avatar
rcohen
Topic Author
Posts: 252
Joined: November 15th, 2001, 12:06 pm

model for yield curve

February 28th, 2006, 7:12 pm

QuoteOriginally posted by: farmerI'm confused. You can't use it to trade bond options, right? So what does your model enable you to do? What inputs do you use to come up with prices, and for what?Do you use the price of a three-year treasury note to figure out what price to pay for a two-year treasury note?Don't economists like to talk about their projections of the gdp growth rate, uncertainties, unemployment, etc? A model such as this could potentially, after some sort of "callibration", allow one to convert these projections (of, let's say, gdp growth rate and uncertainties) into a (projected) yield curve, which can be used to trade futures in bonds, swap rates, etc. That's all! In its current state, the model is obviously only a framework, which means that it's in its most basic form. If there's any potential in it, it could be expanded and made more sophisticated to account for further ideas and inputs.And you're right - it allows the price of a two-year treasury to be calculated from that of a three-year note and vice versa.
 
User avatar
farmer
Posts: 13477
Joined: December 16th, 2002, 7:09 am

model for yield curve

February 28th, 2006, 7:15 pm

Thanks for the explanation.
Last edited by farmer on February 27th, 2006, 11:00 pm, edited 1 time in total.
ABOUT WILMOTT

PW by JB

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...


Twitter LinkedIn Instagram

JOBS BOARD

JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...


GZIP: On