Inflation-linked or regular gilts?

masonic

My rather simplistic view is that the market has priced linkers according to what it thinks will happen to the rate of inflation in the future. Price will probably incorporate some element of a discount for the protection against higher inflation, but if you buy a linker and hold it to maturity, then you'll get the YTM of -2.5% plus a boost from the current measure of inflation being RPI vs CPI, which has tended to be an extra ~1%:

https://www.ons.gov.uk/economy/inflationandpriceindices/timeseries/dra2/mm23

So taking that real return of about -1.5%, you'd compare that to the nominal return from other assets, conventional gilts at 0.8% or cash at 0.6 to 1.5% and determine that CPI would need to be above 1.3 to 3% (depending on your comparator) to achieve a better return. If investing within a pension or S&S ISA, then that would be 1.6%, or in a cash ISA 2.1 to 2.6%, or in a conventional savings account 2.1 to 3%. If you want protection from CPI inflation that is higher than that, then you'd see linkers as a worthwhile part of your low risk assets. The cost of including them is that inflation is unlikely to be higher than that (it's projected to stay within the 2-3% range over the rest of this year and next year), but like any insurance, you are worse off having it if you aren't unusually blighted by the risk you are insuring against.

aroominyork

masonic said:

My rather simplistic view is that the market has priced linkers according to what it thinks will happen to the rate of inflation in the future.

I agree - the only logical explanation seems to be that the markets have consistently overestimated future inflation. That begs the question of whether the UK habitually undershoots inflation forecasts - maybe the markets are set by people still nursing hangovers from the 1970s? Edit: have I got this the wrong way round? If the real return from linkers is high, is that because they were cheap to buy (the market thought inflation would be low) but higher than anticipated inflation boosted their real return? Someone help straighten this out please....

Below is a chart with nominal vs. linked data from Trustnet charts (estimated from graphs). Nominal and linked gilts closely matched each other from 1990, when the charts begin, until the early noughties when linkers started to open a gap.

tebbins

aroominyork said:

masonic said:

My rather simplistic view is that the market has priced linkers according to what it thinks will happen to the rate of inflation in the future.

I agree - the only logical explanation seems to be that the markets have consistently overestimated future inflation. That begs the question of whether the UK habitually undershoots inflation forecasts - maybe the markets are set by people still nursing hangovers from the 1970s? Edit: have I got this the wrong way round? If the real return from linkers is high, is that because they were cheap to buy (the market thought inflation would be low) but higher than anticipated inflation boosted their real return? Someone help straighten this out please....
Yep it's the other way round, if inflation turned out lower than expected linkers would underperform. However over shorter periods, bearing in mind they average about a 20 year maturity, inflation speculation drives a lot od the difference so overestimating inflation can bump up the return over shorter periods, only the market underestimating inflation can cause linkers to beat gilts over the longer term (plus the extra average maturity).

Below is a chart with nominal vs. linked data from Trustnet charts (estimated from graphs). Nominal and linked gilts closely matched each other from 1990, when the charts begin, until the early noughties when linkers started to open a gap.

... You can find this information recently in the performance tabs of Vanguard funds, and older data in the Barclays Equity Gilts Study (you can normally find a recentish copy for free on Scribd or some random website asking for a free trial, just be careful obvs).

Nb the gilts index was calculated using undated funds until 1962, then an average 20y maturity til 1990, then 15y.

The linkers index used an average 15y maturity til 1990, 20y since.

The equity index was based on a retrospectively computed FT30 index from 1899-1930, the FT30 from 1930-1962, then the FTSE All Share since.

Gilts data

Page 2

Linkers data

Total return data

Linkers total return data

JohnWinder

Dive in, folks. https://www.bogleheads.org/forum/viewtopic.php?t=265761

tebbins

Deleted_User said:

JohnWinder said:

Dive in, folks. https://www.bogleheads.org/forum/viewtopic.php?t=265761

So according to Larry Swedroe, there are 2 ways of estimating what I thought there was no way to estimate, viz. how the market's expectation of inflation compares to the difference between the YTMs of conventional and index-linked gilts.

However, I have no idea how to dig out equivalent figures for UK gilts in 2021. (His examples are for US treasuries in 2018.)

Swedroe always seems very sharp, but sometimes he just adds complications and I have no idea what to do with the extra layers of complexity

The market's expectation is implied by the difference between the YTM of conventional gilts and linkers of the same maturity . You can get this if you look around a bit on dmo.gov.uk, however as usual the UK is crap at reporting data compared to the US where a simple search of "treasury yields" takes you to a working website with a table that presents the data in a way you can readily understand.
E.g.
UKT 0.25 07/31 yield 0.61%
UKGI (linker) 0.125 08/31 yield -2.95%
Implies breakeven inflation of 3.54% over the next decade-ish

UKT 1.25 07/51 yield 0.95%
UKGI (linker) 03/51 yield -2.43%
Implies breakeven inflation expectation of about 3.38% over 30 years.

This will never be exact because of differences in maturity, the RPI indexation lag, and different levels of accrued interest between dividend payment dates.

tebbins

Deleted_User said:

tebbins said:

Deleted_User said:

JohnWinder said:

Dive in, folks. https://www.bogleheads.org/forum/viewtopic.php?t=265761

So according to Larry Swedroe, there are 2 ways of estimating what I thought there was no way to estimate, viz. how the market's expectation of inflation compares to the difference between the YTMs of conventional and index-linked gilts.

However, I have no idea how to dig out equivalent figures for UK gilts in 2021. (His examples are for US treasuries in 2018.)

Swedroe always seems very sharp, but sometimes he just adds complications and I have no idea what to do with the extra layers of complexity

The market's expectation is literally implied by the difference between the YTM of conventional gilts and . You can get this if you look around a bit on dmo.gov.uk, however as usual the UK is crap at reporting data compared to the US where a simple search of "treasury yields" takes you to a working website with a table that presents the data in a way you can readily understand.

Try reading the Swedroe article (linked to from bogleheads). He gives 2 reasons why it's not "literally implied". Conventional gilts are more liquid than index-linked (he says this about US treasuries, but I believe it's also true of gilts), which should give them a lower yield than IL. And IL provide a known real yield, which is valuable to some investors, and therefore should give them a lower yield than conventional. There 2 factors work in opposite directions, and may or may not cancel one another out.

BTW, I used these pages to look up yield for gilts: https://www.fixedincomeinvestor.co.uk/x/bondtable.html?groupid=3 https://www.fixedincomeinvestor.co.uk/x/bondtable.html?groupid=3530

That site is so much better than the DMO! And valid points but the market will still correct if it sees inefficiencies with linker yields. But then they are different beasts bought and used for different reasons so perhaps the market could expect inflation to be off from the implied breakeven rate, yet not exploit that for other reasons. Also they tend to be held somewhat passively by pension funds for specific portfolio reasons, the illiquidity suggests the market may not care as much to correct inefficiencies, and even if it wanted to the illiquidity and volatility may add risk to any attempt to do so.

masonic

tebbins said:

Also they tend to be held somewhat passively by pension funds for specific portfolio reasons, the illiquidity suggests the market may not care as much to correct inefficiencies, and even if it wanted to the illiquidity and volatility may add risk to any attempt to do so.

If illiquidity was causing the market to be inefficient, then I'd expect to see that reflected in deviations from the yield curve, yet both flavours of gilt seem to be priced according to maturity with little deviation from the curve. There are a few examples where a slightly shorter dated gilt has a slightly higher YTM, but that is to be expected.

aroominyork

Prism said:

Anyway, these sorts of decisions are beyond me so I will stick with global government bonds (hedged)

Me too. If you don't understand something and stick with an index fund, you might not get the best result but you sleep easy knowing you'll never have anything to blame yourself for. So 50% in Vanguard global bond index and the other half as reduced balances in two strategic bond funds I already own.

masonic

aroominyork said:

Prism said:

Anyway, these sorts of decisions are beyond me so I will stick with global government bonds (hedged)

Me too. If you don't understand something and stick with an index fund, you might not get the best result but you sleep easy knowing you'll never have anything to blame yourself for. So 50% in Vanguard global bond index and the other half as reduced balances in two strategic bond funds I already own.

Unfortunately there is no consensus when it comes to which index is most appropriate for UK investors' bond exposure. Do you, for example, go UK vs global, hedged vs unhedged, include or exclude inflation linked debt. If global, do you track the "Bloomberg Global Aggregate Bond Index", or the more contrived "Bloomberg Barclays Global Aggregate Float Adjusted and Scaled Index"? There are several decisions to be made that will lead to different results. It's all far more murky than the situation with equities, where global is a clear thumbs up, hedged is a clear thumbs down, and the only real point for debate is home bias.

aroominyork

Global and hedged seems good enough. Add some linkers if you want. I've concluded that's good enough and you can then add more complication if you want.