Local Expectations Theory... wtf?

I’ve read this 3 times. I don’t get it. I’ve gone back and looked over sharpshooters question (http://www.analystforum.com/phorums/read.php?12,957330) and I still don’t understand it. Anyone actually understand this? FYI: This is ONLY in CFAI, Book 5, p. 236

I think that first paragraph sums up what they want you to know: that your total return on any bond will be the same over the short-term. It doesn’t matter whether you buy a 1-year, 1% coupon bond or a 30-year, 20% coupon bond…within say 6 months, the price will adjust so your return on either investment will be the exact same Remember, it’s just a theory…Does that help at all?

i don’t recall the theory, but it is not true that in the short term your total return is the same (1%, 1-yr bond, versus 20% 10-yr bond), because after one year you get back your principal with the 1-yr bond, but with the 30-yr bond, you get the 20% return, but the price of the bond could by anyhere, i.e., you may get 20% but lose 70% on the principal.

If you have Schweser’s notes, you can look at the section on Pure Expectation theories and it’s the same as the second interpretation given in the notes (lock-in rate). The “local” interpretation of the pure expectation theory says this. Suppose the i-years forward rate j years from now (which is k%) will be realized and there is no arbitrage. If you want to lock into a k% rate today, you can buy a(n) (i+j) year zero-coupon bond rather than investing in a j-year zero-coupon bond and when it matures, reinvest into a zero-coupon bond that matures in i-year. It’s in-different if you invest into several shorter terms zero-coupon bonds or a longer term zero-coupon bonds as long as the forward rate realizes and the no arbitrage assumption holds. P.S. There are a few questions in CFAI text that specifically address these interpretation. I found it helpful to work through them since I was pretty clueless too after reading the relevant section in SchweserNotes.