Chapter 22 Relationship between returns on asset classes

[mugono]

Hi guys,

I have just seen a comment on page 6
2.4 Index-linked bonds:

The real return on index-linked bonds is known at outset, if they are held to redemption....

Surely this cannot be the case because the coupon and redemption value will move in line with an index (that will be determined as inflation becomes known, albeit with a lag)

Presumably, this is precisely why purchasing index-linked bonds provides a hedge against unanticipated inflation.

I thought that it is only conventional bonds that are held to maturity that will have a known nominal return at outset.

I would really appreciate insight into what I am missing.

Thanks in advance.

 

[Calum]

Both you and the text are correct, just you're at slightly cross purposes, I think. It's confusing because real return sounds like the actual numerical figure you would end up with, which seems sensible, but is unfortunately dead wrong

The real return from a bond is the return *after* inflation is taken into account. So, if you buy a normal (non-indexed) bond today that pays fixed coupons of, say, 2% pa for the next fifteen years, and you want to know the present value of the real return, it's impossible, because you need to know inflation figures for the next fifteen years.

With an index linked bond, it's the opposite: you've no idea what the actual final amount of the payments and hence the present value are. However, since you know that the payments will all increase in line with inflation, you can work out their real value now and hence you know what the real return is.

 

[mugono]

Hi Calum,

Thanks for your comments.

My understanding with regards to conventional fixed interest bonds is that future inflation EXPECTATIONS ARE priced in. One explanation as to why bond prices change is as a result of actual inflation turning out to be different to what was initially expected.

Index-linked bonds have been designed precisely to adjust for any differences between the expected inflation originally priced in and the actual inflation observed.

This is how I am able to justify why investors do not need an inflation risk premium for index-linked bonds as compensation for inflation uncertainty.

Maybe I've missed the point but I am still not convinced we can ever know the real return for an index-linked bond if held to maturity.

Any thoughts on my rationale?

Thanks.

 

[didster]

The real return is return after stripping out changes in inflation.
So if you apply standard bond calculations all based on a index of 100, you'll get the real return.
Eg bond price is 100* index. In one year you get 100*index plus 2*index in coupon, return is 2% real.

If inflation is 10% or 2%, it doesn't matter, the index will increase and your cash payments will increase. But after stripping out the effect of inflation, you earn 2% real regardless. You can still only buy 102 of whatever basket was represented by the index.

As with all bonds this is dependant on holding to maturity since otherwise your return is dependant on the price you sell it for.

 

[Cardano]

The real return of inflation proof bonds is known if you hold to redemption and believe the government is not going to fiddle the cpi to suit their own purposes!

But then I am a cynic

 

[Calum]

My understanding with regards to conventional fixed interest bonds is that future inflation EXPECTATIONS ARE priced in. One explanation as to why bond prices change is as a result of actual inflation turning out to be different to what was initially expected.

Index-linked bonds have been designed precisely to adjust for any differences between the expected inflation originally priced in and the actual inflation observed.

No. In general, anything you invest your money in pays interest to compensate you for not having your money while it is invested and to pay for the risk that the investee may lose it (ie blow it on a tranche of sub-prime mortgages). Assuming "reasonable" inflation, then most bonds will indeed exceed inflation (or they would be a bad investment and nobody would buy them), but conventional bonds are specifically priced with inflation in mind.

This is how I am able to justify why investors do not need an inflation risk premium for index-linked bonds as compensation for inflation uncertainty.

Maybe I've missed the point but I am still not convinced we can ever know the real return for an index-linked bond if held to maturity.

Check your CT1 notes Real return is the return after the inflation is accounted for: so, since we know an index linked bond will account for inflation, we know exactly what the real return is going to be.

As I said in my first post, I think you're just misinterpreting what is meant by real return: there is a logic to it but it's not immediately obvious.

Cardano - yes, the government can and does change the basis for measuring CPI, but in terms of index linkage what practical difference does it make? I still know that whatever inflation turns out to be, I will be covered.

 

[didster]

Cardano - yes, the government can and does change the basis for measuring CPI, but in terms of index linkage what practical difference does it make? I still know that whatever inflation turns out to be, I will be covered.

There's a lag, so presumably possible to fiddle to their advantage. I suspect however, there's a lot more advantage to be gained than just the coupons on index linked bonds.

 

[Cardano]

Ever since Governments have existed they have found ways to fleece their bond holders. They control the rules and so they can pretty much decide what CPI number they use. Of course this isn't without consequences as many governments have found when the bond vigilantes turn up demanding much higher nominal and real returns.
My original post was meant in terms of "hardnosed practical politics", not in terms of the self serving nonsense most actuaries believe and the sort of stuff you have to turn out to pass these exams.