Inflation-linked or regular gilts?

masonic

aroominyork said:

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.

I've come to the opposite conclusion: UK only, or global and unhedged.

Hedging does not work very well in practice and exposes you to magnified local currency inflation risk due to the link between currency devaluation and inflation. During times when Sterling is weakening (and you'd get a better return from not hedging), the hedged Vanguard fund was declaring profits due to ineffective hedging in its annual report. It follows that during a time when Sterling is strengthening (and hedging would deliver a superior return to unhedged), you will only get a partial benefit and that benefit will come at a price (the cost and spread of the derivative employed). I was originally weighing up whether to switch to the Vanguard hedged product from an unhedged equivalent, but what I've read in the annual reports has put me off completely, and instead I am increasing my UK bias to fixed interest instead.

Since the main hedged fixed interest products haven't been around for very long, they've only experienced periods where sterling has broadly weakened, but there are a number of hedged equities funds that have a bit more history. In general these have tended to underperform to a much greater extent than currency movements would explain, and this may be an indicator of the drag on returns from the derivatives employed, and also the ease at which such a strategy can over- or under-shoot the desired result.

UK and global bond indices have pretty much the same nominal YTM, so there really doesn't seem to be an advantage to the long term holder in holding non-local currency debt hedged back to GBP vs holding local currency debt. While the UK doesn't hold its former AAA credit rating, it is of high enough quality not to look elsewhere if its yield is comparable (as it is now).

Global bonds do seem to have an advantage in being lower volatility and global bond indices tend to have a slightly shorter duration. The unhedged variety can presumably also be useful is in a scenario where there is local inflation that isn't reflected in other countries. For example, if there are political or structural reasons why goods and services might get more expensive in the UK specifically. In that case, the spending power of Sterling would generally be reduced and this would be reflected in its exchange rate with other currencies such as the dollar. In this case, unhedged exposure to US treasuries would be beneficial, as the income stream and principal would become more valuable over time. There is of course a risk of the converse, which if you did not want to take on it would make most sense at the moment to invest in UK bonds rather than hedged global bonds.

The above is from recent reading around the subject and my own research, so I'm happy for any errors to be pointed out.

aroominyork

Many thanks, masonic and FeralHog, though it is a level of detail above my pay grade. I will read it slowly several times and maybe come back with questions, but in the meantime I look at a UK gilt index against a hedged global govt ETF (XGSG) and am drawn to the global fund's lower volatility. Is that a false comfort? (I ran the same chart over ten years on Trustnet and the gilts outperformed, but that was solely down to a rise in gilts after the 2016 referendum.)

masonic

aroominyork said:

Many thanks, masonic and FeralHog, though it is a level of detail above my pay grade. I will read it slowly several times and maybe come back with questions, but in the meantime I look at a UK gilt index against a hedged global govt ETF (XGSG) and am drawn to the global fund's lower volatility. Is that a false comfort? (I ran the same chart over ten years on Trustnet and the gilts outperformed, but that was solely down to a rise in gilts after the 2016 referendum.)

The dampening of daily volatility is a genuine benefit of hedged global govt bonds. However, I think the comparable performance is coincidental and may not hold up in circumstances where exchange rates move differently. Ultimately bonds generate income only and that income can be pulled forward (or pushed back), but it can't be increased over the life of the bond. If there is no disparity in YTM at the time you invest (after allowing for forex), then over a long holding period returns (before costs and inefficiencies) will be about the same.

If you are aiming for a target volatility as your objective, then that is where I think they have value (your alternative would be short dated gilts, which will likely be a bigger drag on returns). This is perhaps why they are a popular component of multi-asset funds. If you just want a diversifier that is negatively correlated with equities and don't mind a bit more short-term volatility, then gilts look more attractive to me from the perspective of hedging costs and risk (which are unknowns given the return history isn't sufficient to see their performance over a range of conditions).

I do agree with FeralHog's points around unhedged global bonds. I've noted some of the capital preservation funds have an allocation to (for example) US treasuries/TIPS, which doesn't appear to be hedged, but I could be wrong on this, and they may be making currency bets that would be ill advised for us to do.

Prism

masonic said:

aroominyork said:

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.

I've come to the opposite conclusion: UK only, or global and unhedged.

Hedging does not work very well in practice and exposes you to magnified local currency inflation risk due to the link between currency devaluation and inflation. During times when Sterling is weakening (and you'd get a better return from not hedging), the hedged Vanguard fund was declaring profits due to ineffective hedging in its annual report. It follows that during a time when Sterling is strengthening (and hedging would deliver a superior return to unhedged), you will only get a partial benefit and that benefit will come at a price (the cost and spread of the derivative employed). I was originally weighing up whether to switch to the Vanguard hedged product from an unhedged equivalent, but what I've read in the annual reports has put me off completely, and instead I am increasing my UK bias to fixed interest instead.

This is an interesting point and one which I had not fully considered. I use hedged global gov bonds for the diversification and slightly higher yield while the hedging keeps the volatility down. However I had not actually considered if the hedging actually works or not in extreme conditions and it seems that it might struggle to keep currency movements neutral.

It is also convenient that the global bond funds tend to have a lower duration so less affected by interest rate changes but nothing that can't be solved with two gilt funds - a 0-5 year short duration combined with a standard duration.

The OCF of the gilt index funds is typically lower too.

So now I am reconsidering my choices... 

JohnWinder

Yes, interesting.

However I had not actually considering of the hedging actually works in extreme conditions and it seems that it might struggle.

1. The syntax has thrown me, so if it can be edited for clarity so much the better.

2. Vanguard says: 

'Currency hedging techniques are used to minimise the risks associated with movements in currency exchange rates but these risks cannot be eliminated entirely.' So we might have overlooked currency hedging being imperfect, but we were warned.

3. I am interested, but can't be sure of the inflation/hedging issue that would make unhedged bonds better. Not saying it's wrong, I just don't follow the argument in that quote. It seems to be: when sterling falls hedging is bad, but we know it's not perfect in dealing with currency movements so not as bad as it might have been, AND we get inflation when sterling falls. Do we?

Prism

JohnWinder said:

Yes, interesting.

However I had not actually considering of the hedging actually works in extreme conditions and it seems that it might struggle.

1. The syntax has thrown me, so if it can be edited for clarity so much the better.

2. Vanguard says: 

'Currency hedging techniques are used to minimise the risks associated with movements in currency exchange rates but these risks cannot be eliminated entirely.' So we might have overlooked currency hedging being imperfect, but we were warned.

3. I am interested, but can't be sure of the inflation/hedging issue that would make unhedged bonds better. Not saying it's wrong, I just don't follow the argument in that quote. It seems to be: when sterling falls hedging is bad, but we know it's not perfect in dealing with currency movements so not as bad as it might have been, AND we get inflation when sterling falls. Do we?

I think for me its just about government bonds being as predictable as possible although of course that would likely come with a pretty low return. If the hedging works it would seem that hedged global government bonds would edge it on diversification and performance as it seems that other governments on average pay a slightly higher yield than UK Gilts. However the hedging doesn't work as expected then maybe the more unpredictable factor introduced by currency movements would be less desirable.

I currently use a small allocation of hedged global government bonds but would likely want a bit more at retirement so will need to decide which route to go over the next few years.

masonic

I could previously see a case for having unhedged exposure to international bonds (and did so) during the secular devaluation of GBP, which was driven by factors that have now largely played out. It will generally cause inflation in the price of goods and services that are imported or otherwise sold in other currencies that the pound is weakening against (primarily USD). For a tongue in cheek guide to this phenomenon read up on the Big Mac index and burgernomics.

If you don't want to take a play on exchange rate movements, then the choice is between hedged global vs local currency. I would be more comfortable with hedged global if there was more transparency/data from which to assess what could happen under a range of different scenarios. The hedging aspect currently sits in the category of things I don't adequately understand, so it is something I am reluctant to invest in. Given that there is no appreciable advantage (for me as someone who doesn't mind day to day volatility), I'm happy enough with a Gilt ETF. I'm holding this as the most negatively correlated asset vs equities (allowing me to hold proportionally less than other options) and for its protection against negative interest rates. Due to inflation and interest rate risk (i.e. rates rising), I also hold some linkers and cash (in consumer savings accounts).

tebbins

Prism said:

JohnWinder said:

Yes, interesting.

However I had not actually considering of the hedging actually works in extreme conditions and it seems that it might struggle.

1. The syntax has thrown me, so if it can be edited for clarity so much the better.

2. Vanguard says: 

'Currency hedging techniques are used to minimise the risks associated with movements in currency exchange rates but these risks cannot be eliminated entirely.' So we might have overlooked currency hedging being imperfect, but we were warned.

3. I am interested, but can't be sure of the inflation/hedging issue that would make unhedged bonds better. Not saying it's wrong, I just don't follow the argument in that quote. It seems to be: when sterling falls hedging is bad, but we know it's not perfect in dealing with currency movements so not as bad as it might have been, AND we get inflation when sterling falls. Do we?

I think for me its just about government bonds being as predictable as possible although of course that would likely come with a pretty low return. If the hedging works it would seem that hedged global government bonds would edge it on diversification and performance as it seems that other governments on average pay a slightly higher yield than UK Gilts. However the hedging doesn't work as expected then maybe the more unpredictable factor introduced by currency movements would be less desirable.

I currently use a small allocation of hedged global government bonds but would likely want a bit more at retirement so will need to decide which route to go over the next few years.

Aside from emergnng markets government bonds, I think the only major issuer paying a higher yield than the UK is the US whicb brings up the average yield of a global government bond fund, hence as I see it the choice is between hedged US government bonds and gilts.

masonic said:

I could previously see a case for having unhedged exposure to international bonds (and did so) during the secular devaluation of GBP, which was driven by factors that have now largely played out. It will generally cause inflation in the price of goods and services that are imported or otherwise sold in other currencies that the pound is weakening against (primarily USD). For a tongue in cheek guide to this phenomenon read up on the Big Mac index and burgernomics.

If you don't want to take a play on exchange rate movements, then the choice is between hedged global vs local currency. I would be more comfortable with hedged global if there was more transparency/data from which to assess what could happen under a range of different scenarios. The hedging aspect currently sits in the category of things I don't adequately understand, so it is something I am reluctant to invest in. Given that there is no appreciable advantage (for me as someone who doesn't mind day to day volatility), I'm happy enough with a Gilt ETF. I'm holding this as the most negatively correlated asset vs equities (allowing me to hold proportionally less than other options) and for its protection against negative interest rates. Due to inflation and interest rate risk (i.e. rates rising), I also hold some linkers and cash (in consumer savings accounts).

Gilts are the most negatively correlated asset with UK equities, but you can't count on that inverse relationship. Comparing the FTSE All-Share vs 15 year gilts (since that's what the only decent data I have access to in the Barclays Equity Gilts Study is), in the 1990s it was actually +0.59, moderately positive, with only 1 year (1990) with inverse returns. In the 2000s this flipped to -0.64 with 6/10 years showing inverse returns, and has remained negative but less strongly correlated at -0.29 for the years 2010-2020 inclusive, with 4/11 showing inverse returns.

masonic

tebbins said:

Prism said:

JohnWinder said:

Yes, interesting.

However I had not actually considering of the hedging actually works in extreme conditions and it seems that it might struggle.

1. The syntax has thrown me, so if it can be edited for clarity so much the better.

2. Vanguard says: 

'Currency hedging techniques are used to minimise the risks associated with movements in currency exchange rates but these risks cannot be eliminated entirely.' So we might have overlooked currency hedging being imperfect, but we were warned.

3. I am interested, but can't be sure of the inflation/hedging issue that would make unhedged bonds better. Not saying it's wrong, I just don't follow the argument in that quote. It seems to be: when sterling falls hedging is bad, but we know it's not perfect in dealing with currency movements so not as bad as it might have been, AND we get inflation when sterling falls. Do we?

I think for me its just about government bonds being as predictable as possible although of course that would likely come with a pretty low return. If the hedging works it would seem that hedged global government bonds would edge it on diversification and performance as it seems that other governments on average pay a slightly higher yield than UK Gilts. However the hedging doesn't work as expected then maybe the more unpredictable factor introduced by currency movements would be less desirable.

I currently use a small allocation of hedged global government bonds but would likely want a bit more at retirement so will need to decide which route to go over the next few years.

Aside from emergnng markets government bonds, I think the only major issuer paying a higher yield than the UK is the US whicb brings up the average yield of a global government bond fund, hence as I see it the choice is between hedged US government bonds and gilts.

Thanks, this looks interesting, so for example GOVP, with YTM of 0.85% vs 0.7% for a Gilts tracker and effective duration of 7 years vs 14. Higher return and less interest rate risk. Just the unknown hedging cost to try to factor in, but presumably it wouldn't outweigh those benefits. Shame it's been around less than 2 years (perhaps there are longer running equivalents?). As expected for a single country, volatility is higher from what data is available.

It does, however, lead me to question why there is this pricing anomaly of US debt being higher yielding, better credit quality and shorter duration than the global weighted average... Is there something we're missing? Is it perhaps based on projections for interest rates in the US (which a hedged foreign investor would be sheltered from)? There is, after all, a clear explanation for why EM bonds are higher yielding.

tebbins said:

masonic said:

I could previously see a case for having unhedged exposure to international bonds (and did so) during the secular devaluation of GBP, which was driven by factors that have now largely played out. It will generally cause inflation in the price of goods and services that are imported or otherwise sold in other currencies that the pound is weakening against (primarily USD). For a tongue in cheek guide to this phenomenon read up on the Big Mac index and burgernomics.

If you don't want to take a play on exchange rate movements, then the choice is between hedged global vs local currency. I would be more comfortable with hedged global if there was more transparency/data from which to assess what could happen under a range of different scenarios. The hedging aspect currently sits in the category of things I don't adequately understand, so it is something I am reluctant to invest in. Given that there is no appreciable advantage (for me as someone who doesn't mind day to day volatility), I'm happy enough with a Gilt ETF. I'm holding this as the most negatively correlated asset vs equities (allowing me to hold proportionally less than other options) and for its protection against negative interest rates. Due to inflation and interest rate risk (i.e. rates rising), I also hold some linkers and cash (in consumer savings accounts).

Gilts are the most negatively correlated asset with UK equities, but you can't count on that inverse relationship. Comparing the FTSE All-Share vs 15 year gilts (since that's what the only decent data I have access to in the Barclays Equity Gilts Study is), in the 1990s it was actually +0.59, moderately positive, with only 1 year (1990) with inverse returns. In the 2000s this flipped to -0.64 with 6/10 years showing inverse returns, and has remained negative but less strongly correlated at -0.29 for the years 2010-2020 inclusive, with 4/11 showing inverse returns.

Yes, nothing is perfect. In fact a -1 correlation would be undesirable as it would be equivalent to having less money invested overall with the diversifier having negative returns over the long term.

tebbins

To answer both of the above, the extra yield minus the hedging cost, and any risks of a hedging failure which is negligible and would simply mean temporary currency exposure on the proportion of the fund that hedging contract hedged covered (I asked Vanguard, they generally use 15 counterparties) until a new hedge could be found, is "free" if you believe the US federal government is no more of a credit risk than other developed markets governments.
The reasons this apparent arbitrage is not exploited away by the market are:-
- currency speculation as suggested previously
- hedging requires a counterparty, it is a mutual gamble and there are only so many willing counterparties. Otherwise all developed markets government bond yields would be distinguished only by the differing credit risks.
- Potentially US investors are a more equity-oriented

- European and Japanese large/institutional investors are often required, whether explicitly in law/rules or implicitly to hold their own government bonds. Active Japan fund managers have commented that in Japan, industrial strength is considered part of national security, and companies are expected to prioritise resilience over pure growth, as opposed to the Western attitude of national security being a means to the end of economic welfare. I don't know as much about the US but I suspect their rules are perhaps more lax?