Evaluate Portfolio Performance of FI Securities following, statements, management, understand

LPoulin133

Reading 46, CFAI text Vol 6 P169,
I don't understand following statements under the title of "Interest rate management effect" :

...., each security in the portfolio is "priced" as if it were a default-free security. The interest rate management contribution is calculted by subtracting the return of the entire Treasury universe from the aggregate return of these REPRICED scurities.

Can anyone explain by raising a tangible example ? What does it mean by "REPRICED" ?

Valores

Thank you for your explanations.

wake2000

It's basically trying to explain yield curve positioning.

You're trying to answer the question "how well did the manager predict changes in the yield curve?".

To do this, you'd take your portfolio at the beginning of the period and price it AS IF ALL SECURITIES were treasuries, and then you'd price the portfolio at the end of the period AS IF ALL SECURITIES were treasuries. Then, using the beginning and ending prices, you'd calculate a return, and then subtract the treasury benchmark return from this. Effectively, you're trying to isolate just the interest rate management effects in the portfolio - so you'd need to remove the impact of sector allocation, security selection, and trading.

For example:
Say you're holding a portfolio of treasury securities (for simplicity's sake). You think that the yield curve is going to flatten with the long end falling, so you'd extend the duration of your portfolio to say 8.5, while your treasury benchmark duration is 6.1. If you're right, and the yields do fall as you expected, you may end up with a return of 3% while the benchmark only returned 2.5%. Consequently, since we're using treasuries in our portfolio, we can ignore sector allocation and security selection (and just say there's no trading impact), so we've contributed 50 basis points through yield curve positioning.

Now, if we were using corporates, we'd also have to incorporate the effects of yield spreads into the analysis as well. But in the case of interest rate management, we're just concerned with the yield curve positioning, so we don't care about spreads right now, which is why we need to reprice everything using treasury yields to take the spread effects out of the equation.

Does that help?

Windjam

Any other further advice ?

skycfa

A manager may chose 2 or 3 bonds with different maturity and credit quality. How would you compare 2 such managers , if you don't want to study credit guality choices?

Choose default free i.e. treasury securities of similar maturities as each manager's choice. Rank the returns of the particular combination of each set of treasuries.

Thus you're selecting managers with good prediction capabilities on yield curve positioning, and excluding spread differences due to credit quality.

Darien

The noise in the manager's portfolio due to differences in pricing between otherwise identical issues is eliminated.

Say Corp A has credit rating BBB while Corp B has rating A , then for an identical maturity the two bonds would have different prices owing to different premium.

The only way to completely eliminate a premium due to credit rating would be to go to a default free instrument of identical maturity / duration .

The set of instruments/maturities chosen by the manager is repriced this way , i.e. we would substitute with a set of treasuries . Then compare to the "market" of default-free bonds ( proxied by entire treasury universe ) which serves as benchmark.

If the manager's returns on this basis outperform , then he is making good judgments on positioning his choices

Iginla2011

janakisri Wrote:
-------------------------------------------------------
> Say Corp A has credit rating BBB while Corp B has rating A , then for an identical maturity the > two bonds would have different prices owing to different premium.
>
> The only way to completely eliminate a premium due to credit rating would be to go to a
> default free instrument of identical maturity / duration .

Sorry, I don't get it. The yield of default free entire treasury universe shall be known, why bother inferring the yield by "repricing" the bonds of both Corp A & Corp B ?