Index-Linked Gilts question

zagfles

You can download a spreadsheet showing the index ratio and loads of other useful stuff here Index-linked Gilts 
I track mine using a spreadsheet working out the IR for the next coupon and maturity from the base RPI for the gilt and the RPI index which I update every month or so. Obviously the maturity value is based on current real value as is the coupon until a couple of months before the payment date. 

aroominyork

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

masonic

aroominyork said:

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

Yes, it seems to break down at the short extreme, and probably diverges due to inflation linking of the coupon at the very long end.

I haven't figured out the reason for the discrepancy when there are only one or two distributions left, but I suspect the YTM calculation isn't correct. A 1.75% real return seems unlikely.

zagfles

masonic said:

aroominyork said:

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

Yes, it seems to break down at the short extreme, and probably diverges due to inflation linking of the coupon at the very long end.

I haven't figured out the reason for the discrepancy when there are only one or two distributions left, but I suspect the YTM calculation isn't correct. A 1.75% real return seems unlikely.

For gilts due to mature very soon the indexation lag will have a significant effect. The dirty price won't include the last few months inflation so if eg inflation spiked last month you'd expect the clean price to rise as there's a known rise coming in the dirty price. You need to look at the inflation spot rate rather than the headline figure which is a 12 month rolling figure. For instance the April 2025 spot rate for RPI was 23.1% 

DRS1

zagfles said:

masonic said:

aroominyork said:

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

Yes, it seems to break down at the short extreme, and probably diverges due to inflation linking of the coupon at the very long end.

I haven't figured out the reason for the discrepancy when there are only one or two distributions left, but I suspect the YTM calculation isn't correct. A 1.75% real return seems unlikely.

For gilts due to mature very soon the indexation lag will have a significant effect. The dirty price won't include the last few months inflation so if eg inflation spiked last month you'd expect the clean price to rise as there's a known rise coming in the dirty price. You need to look at the inflation spot rate rather than the headline figure which is a 12 month rolling figure. For instance the April 2025 spot rate for RPI was 23.1% 

I am hoping that was a fat finger and that RPI for April 2025 wasn't really 23.1%

zagfles

DRS1 said:

zagfles said:

masonic said:

aroominyork said:

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

Yes, it seems to break down at the short extreme, and probably diverges due to inflation linking of the coupon at the very long end.

I haven't figured out the reason for the discrepancy when there are only one or two distributions left, but I suspect the YTM calculation isn't correct. A 1.75% real return seems unlikely.

For gilts due to mature very soon the indexation lag will have a significant effect. The dirty price won't include the last few months inflation so if eg inflation spiked last month you'd expect the clean price to rise as there's a known rise coming in the dirty price. You need to look at the inflation spot rate rather than the headline figure which is a 12 month rolling figure. For instance the April 2025 spot rate for RPI was 23.1% 

I am hoping that was a fat finger and that RPI for April 2025 wasn't really 23.1%

The April spot rate (by which I mean the change in the index from March to April annualised) was 23.1%. 

March RPI index 395.3. April index 402.2. Change is 1.7455% over a month which is 23.1% per year. 

It's not unusual to have high spot rates of inflation, or even negative. We're too used to thinking about inflation in terms of rolling 12 month rates which hide these peaks and troughs - but the rolling 12 month rate doesn't tell you how much prices are changing at the moment any more than how many miles you drove in the last hour tells you your current speed. IL gilts don't use rolling 12 month rates, they use the current index value (with a lag of usually around 3 months but 8 months with the older per 2005 gilts). 

DRS1

Thanks.  I can see I am going to have to look at the coupons on my ILGs.  Despite the lags I figure the increase between coupons should still reflect 6 months of RPI increases.

The rolling 12 month RPI rate for March was 2.8% and for April was 4.2% - that seems like a 50% increase (even more worrying than 23.1%).  Anyway I'll just accept that I don't know anything about ILGs  I might have more of a clue when my 2026 one matures.

zagfles

DRS1 said:

Thanks.  I can see I am going to have to look at the coupons on my ILGs.  Despite the lags I figure the increase between coupons should still reflect 6 months of RPI increases.

The rolling 12 month RPI rate for March was 2.8% and for April was 4.2% - that seems like a 50% increase (even more worrying than 23.1%).  Anyway I'll just accept that I don't know anything about ILGs  I might have more of a clue when my 2026 one matures.

The 12-month rolling rate is fine for looking at the longer term picture, but misleading when looking at the short term. For instance the "50% increase" is an increase in the rate of increase (ie acceleration compared to speed in a car analogy), so is pretty meaningless. It also depends heavily on the spot rate in April last year, for instance if the spot rate in April 2024 was low then the April 2025 rate will rise even if the spot rate stays the same because a low inflation month has dropped out of the 12-month rolling figure. So apparent rises in inflation headline rates can be due to low inflation a year ago rather than high inflation now. Basically inflation will rise if the last month's spot inflation is greater than spot inflation 12 months ago, and will fall if vv.  

My spreadsheet also works out a rolling 10 year average rate, and at the moment that's 4.6%, the highest it's been since before 1997 !! (and probably a lot earlier - it only goes back that far as it's based on 100 index in 1987). So for people looking at the longer term that's a figure worth watching. 

OldScientist

aroominyork said:

masonic said:

aroominyork said:

I'm struggling to interpret that, masonic. Does your answer explain the cost of the insurance to index link your gilt? In other words, the level over expected inflation where you will receive no benefit. Or am I wrong in likening this to an insurance policy because no-one is taking a cut for issuing the policy?

Yes, it's approximately the difference between the nominal YTM of a nominal vs index linked gilt, getting less accurate at either extreme of duration. For example, T28 (Aug) @ 0.5% vs TS28/TG28 (Jun/Oct) @ 3.9/3.7 gives a 3 year market forecast of inflation @ 3.3%pa. So if that comes to pass you will receive no benefit. If RPI is lower than that you will be worse off with index linked.

That is not considering tax/level of coupon. YMMV if investing unwrapped.

As RPI inflation has averaged 3.5% over the last 25 years, 4.3% over the last 10 years, and 2.8% over the historically abnormal post-GFC decade, it is not expensive insurance.

So at the very short end, TR26 ILG maturing 22/3/26 yields 1.75%. T26/TG26 maturing 30/1/26 and 22/7/26 yield 3.23% and 3.76% respectively, so call it 3.40%. The ILG beats nominal if inflation exceeds 1.65%... that doesn't seem right. (Yields from dividenddata.co.uk)

T26 and TG26 both have relatively low coupons for nominal gilts which can suppress the yields (because they have favourable tax properties, so the prices are bid up), so this may make a small contribution to the apparently low breakeven inflation. As @zagfles said, the 3-month indexing lag also has some odd effects close to maturity.

aroominyork

They sure are complicated, these ILGs. OK, I get the purpose of linking your returns to UK inflation and matching maturity to your expenditure, but if you want an easier way to go about it - proxying developed world inflation for the UK's and being less time specific - how good a solution is a global index/inflation linked bond fund such as Royal London's or Aberdeen's?