Talk:Yield curve

From Wikipedia, the free encyclopedia

This is a great article. Congratulations to all the editors involved.

This article is a former featured article candidate. Please view its sub-page to see why the nomination failed. For older candidates, please check the archive.

Reasons for failure were:

[edit] Yield curve

Second try for this nomination. I believe all the concerns that were expressed the last time around have been fairly addressed. Its a central concept in finance, clearly expounded, with valuable references for those who would like to study the matter further. I've contributed to it some (although the bulk of the credit goes to others) so I guess I should call this a self nomination. --Christofurio 17:02, Mar 31, 2005 (UTC)

• Object. The article is very confusing and flows poorly. Some initial suggestions:

1. 1. Use the normal yield curve as the top picture, and show the inverted one later (you wouldn't show a picture of a three legged dog in dog in the top section and then caption it "most dogs have four legs").
   2. Work in technical jargon slowly, and define, define, define. Why is it called a yield curve? What is yield? How is that relevant to the cost of money?
   3. The article needs to be reorganized -- maybe laid out to first establish a basic understanding of existing theory, then an example, then history. It's very confusing to introduce examples or history of development before the reader is made to understand the basic concepts.
   4. Consistency -- for example, on the table showing construction of the yield curve, it switches from % rate to (1 - % rate).

It's market convention to quote futures prices as 100-100*rate, so the article is correct in this respect. Pcb21| Pete 23:32, 31 Mar 2005 (UTC)

1. 1. Could use a thorough copy edit for spelling, grammar, and sentence structure too.

That should be enough to work on for now. - Bantman 19:08, Mar 31, 2005 (UTC)

• object although there are some references, it's not clear which reference to use to check which material and there could be more. You could, for example, use a footnoting system such as Wikipedia:Footnote3 or one of the other ones to make this clear. Mozzerati 09:41, 2005 Apr 2 (UTC)
  • The article is appropriately referenced - it is a synthesis of material in those five books. You use footnotes to provide a cite for specific facts that otherwise the reader may not have confidence in. Which facts fall into this category? P.s. is five books really too few for an FA these days? Pcb21| Pete 09:49, 2 Apr 2005 (UTC)
    • five books is approximately a basic minimum (though it's impossible to be absolute; sometimes there just aren't that many sources). There are many many facts which I have no reason to be sure about in the article. The nomenclature "curve" is used rather than "yield function" because when plotted on a graph, the function is a curve. - maybe this is true; maybe it isn't. A good reference telling me which of those five books to look at and preferably which page to look on would really make it much more practical to check. This theory perfectly explains perfectly? Really? Please at least give a reference for that. "Liquidity preference theory...is also the most accepted theory of the three" seems total common sense, but again, what are the proportions of people accepting them? or do you mean economists? etc. I don't want to pick too many nits, since I don't think the "perfect" article is what we are asking for. Just a level which stands out above other normal articles. 10 most interesting/important/surprising points would be great.
      • Excuse me, but these do seem to be rather small nits you're hunting. The explanation of the term "curve" will seem quite familiar to anyone with any background in economics, not to say finance -- even a single undergraduate course using Samuelson's textbook will render this familiar. I don't believe specific references of the sort you seem to want are necessary or appropriate where a point is notorious within the pertinent field, available virtually anywhere. --Christofurio 00:38, Apr 5, 2005 (UTC)
      • From a mathematical perspective, a curve is the result of plotting the graph of a function. I don't see the need for any more explanation of why the "yield curve" is called a curve. --Carnildo 21:15, 5 Apr 2005 (UTC)
        • Sorry, but it's exactly this kind of misunderstanding which needs to be cleared up. A curve is not just any function, it must be a continuous function. There is no obvious (to me) reason why the yield curve cannot be discontinous. For example, maybe there is a known night in the future when it is advantagous to have cash. Yields just before that day may be higher and those just after may be lower. Another point: Samuelson isn't in the reference list. Finally, if this really needs an undergraduate course to understand then it qualifies as too technical. As I mentioned, these are specific minor symptoms of the difficulty of following the references. If you rested more on your sources then less would have to be done in the article.. Mozzerati 19:45, 2005 Apr 7 (UTC)

• • • • • • "no obvious (to me) reason"... I'll lend you $1m for 1+delta years at 6% and you lend me $1m for (1-delta) years at 5%, for any delta of your choosing less than 1 second. You'll learn pretty quick :) Pcb21| Pete 10:25, 8 Apr 2005 (UTC)

• Comment - the terms USD and GBP are used in the figures without definition. While I assume that they are US dollar and GB pound, this isn't immediately apparent to the reader. Assuming that they are, could you are least link your captions to the appropriate articles? Guettarda 14:48, 5 Apr 2005 (UTC)
  • I changed the terms to US dollar and British pound in the image captions. - Marcika 22:58, 6 Apr 2005 (UTC)

• Question - I have seen systems that plot yield curves with logarithmic X axes and ones that don't. Is there any advantage to interpreting the curve by using a log scale for the X axis? If so, would it be worth saying why in the article? Thanks. --Simon 03:14, 10 August 2005 (UTC)

• • Curves are typically (in my experience), plotted with logarithmic X axes for readability. The important points on the curve are closer together on the short end (3 and 6 mo treasuries, vs 10 and 20 yrs on the long end). --Dave 19:58, 14 October 2005 (UTC)

• What are "stylized facts regarding a normal yield curve"? What is "stylized fact" in general?

(I say we wait about a month, and resubmit in late March.) --Christofurio 20:59, Feb 25, 2005 (UTC)

[edit] Let's get rid of "stylized facts"

"Stylized facts" are stylistically horrible. Facts must be true, otherwise a statement is either a falsehood or an opinion. Which make "stylized facts" ... mere economic jargon.

I'm not totally disagreeing, but you could also argue that the term 'yield curve' is 'mere economic jargon'. The use of the term 'stylized fact' is widespread in economic literature, because so many behaviours are statistically inferred. It is very cumbersome to say something like "...under the conditions for which data exists thusfar, which may not be enough to make a robust statistical inference, bond yields of different maturities have tended to move together a statistically significant number of times, which we will now posit is a 'normal' market condition, and henceforth will refer to as a 'fact' and use as the basis for a model of the market under real conditions in conjunction with other similarly created suppositions that together will aid our understanding of how the complete model will react to various changing environs, with other additional caveats, many of which can be found in various studies of the matter." Tristanreid 22:15, 19 February 2006 (UTC)

[edit] Implicit Future Rates

I deleted the paragraph quoted below because it is inaccurate/oversimplified. I wanted to explain before I got in a fight. You cannot merely assume, as the original author did, that a 5% 1yr rate and a 5.5% 2 year rate implies a 6% end of year 2 rate. It could imply a market assumption that rates will skyrocket at the end of the year (thus a market predicted rate in excess of 5.5%) or it could assume a January first jump to 5.5% and thus a 5.5% market predicted rate. Additionally, the author forgot that long term investments are inherently more risky and thus require a higher risk premium. If the fed GUARANTEED rates would remain 4% for 10 years, you'd still see a curve in the yield curve. --Laxrulz777 21:52, 28 August 2006 (UTC)

Yield curves carry an implicit forecast of future short-term interest rates: for example if the annual yield on a 1-year bond is 5%, and on a 2-year bond is 5.5%, then the implicit yield (forward rate) in year 2 is