Bond element of pension portfolio

BobR64

Thanks Linton. That sounds like an interesting approach, which makes sense at a high level. For me, it raises lot of questions like:

• What do the income and growth portfolios actually look like? Do you tend to hold funds, or ETFs or individual shares?
• What criteria do you use to measure diversification and how do you arrive at the right balance?
• How much work does it take initially to develop a portfolio and then to maintain it?

I am not expecting you to answer these by the way! It's just that while I get the basic idea of your approach, I'm not sure I would know where to start in implementing it.

BobR64

OldScientist said:

Holding individual gilts can take two forms:
1) a collapsing ladder of the type you have already constructed (i.e. where coupons and maturing bonds are taken as income)
2) a rolling ladder where some or all of the maturing bonds and coupons are reinvested in bonds with higher maturities.

I didn't spell it out in my opening post, but whereas the ladder I have in my GIA to bridge to state pension I definitely do regard as a collapsing ladder, the one in my SIPP could potentially be a rolling one. It seems like there is scope to be flexible. If equities are doing well, I could take my income from there and keep the ladder rolling, and if they are not doing so well I could start collapsing the ladder for a few years.

Functionally, there is little difference between a rolling gilt ladder and a gilt fund with the same maturity range and similar duration.

I think this gets to the crux of one of the things I am struggling with. One of the things I have considered is extending what I have already started doing in my pension, maybe both by extending the gilt ladder for a few years as well as doubling up on the years I already have.

But if, as you say, a gilt fund is essentially the same thing it would make sense to keep it simple and just buy the fund. I would love to be persuaded of this, but I keep seeing and reading things that suggest otherwise. That the fact that funds have to sell their holdings at particular times, before they mature, makes them behave in a fundamentally different way from holding gilts to maturity in one's own ladder.

redpete

I  am fortunate in having a DB pension, 2 full state pensions (me and Mrs RP) and bits of part-time work. I've worked out how much I want to take as income each year from my DC pension.
In the DC pension I've got a bond ladder (all direct UK gilts) to provide this income over the next 4-5 years.  I'm happy with gilts because they have little risk with current yields at or above the withdrawal rate I'm using.  About 50% of my DC is in index equity funds and about 50% in bond or money market funds and the bond ladder.  All the income comes from the bond side and the plan is to periodically sell off a portion of the equity funds into bonds whenever the equities grow by 20%.

hara____

BobR64 said:

But if, as you say, a gilt fund is essentially the same thing it would make sense to keep it simple and just buy the fund. I would love to be persuaded of this, but I keep seeing and reading things that suggest otherwise. That the fact that funds have to sell their holdings at particular times, before they mature, makes them behave in a fundamentally different way from holding gilts to maturity in one's own ladder.

To me, the fund route is much more acceptable if we're talking short duration rather than long.

To use two gilt ETFs as examples: I wouldn't want to hold a large amount in VGOV (full range of gilt durations) as recovery from a yield increase could take so long. But something like IGLS (maximum maturity 5 years) could be more convenient than an equivalent manual ladder and you know that the price wouldn't take too long to recover from an increase in yields.

There's also the question of what drives yields: IGLS will relate strongly to base rates whereas VGOV will move with demand for the long-term bonds. Much easier for me, or the average investor, to understand and accept the first of those.

OldScientist

BobR64 said:

OldScientist said:

Holding individual gilts can take two forms:
1) a collapsing ladder of the type you have already constructed (i.e. where coupons and maturing bonds are taken as income)
2) a rolling ladder where some or all of the maturing bonds and coupons are reinvested in bonds with higher maturities.

I didn't spell it out in my opening post, but whereas the ladder I have in my GIA to bridge to state pension I definitely do regard as a collapsing ladder, the one in my SIPP could potentially be a rolling one. It seems like there is scope to be flexible. If equities are doing well, I could take my income from there and keep the ladder rolling, and if they are not doing so well I could start collapsing the ladder for a few years.

Functionally, there is little difference between a rolling gilt ladder and a gilt fund with the same maturity range and similar duration.

I think this gets to the crux of one of the things I am struggling with. One of the things I have considered is extending what I have already started doing in my pension, maybe both by extending the gilt ladder for a few years as well as doubling up on the years I already have.

But if, as you say, a gilt fund is essentially the same thing it would make sense to keep it simple and just buy the fund. I would love to be persuaded of this, but I keep seeing and reading things that suggest otherwise. That the fact that funds have to sell their holdings at particular times, before they mature, makes them behave in a fundamentally different way from holding gilts to maturity in one's own ladder.

Taking IGLS (mentioned elsewhere as an example), by definition it (since it physically replicates the underlying index) will hold all UK gilts between 0 years (i.e., maturity) and 5 years in proportion to their issue weights. At the moment that is 18 gilts with the shortest maturity of just under 3 months (T25) and the longest just under 5 years (T30). When T25 matures, the proceeds will be reinvested in the fund. I note IGLS is distributing, so the coupons (less fees) will be issued as a 'dividend' (IGL5 is the equivalent accumulating version - coupons would be reinvested in gilts).

For convenience, a self built rolling ladder covering the same range of maturities (i.e. up to 5 years) would probably only hold a sample of the gilts available (possibly one per year?) and (probably) in similar amounts rather than weighted by the amount in issue, but overall the weighted maturity and weighted duration would be similar to the fund and therefore the behaviour (returns, volatility, etc.) would also be similar.

However, there are two possible advantages of the ladder over the fund
1) The holding costs may be lower (although the bid-ask spreads for retail investors may be larger which might negate some of this advantage). However, the OCF of IGLS is 7 bp, so that is fairly small.
2) In a taxable account, the capital gains on individual gilts are not taxed whereas those of a fund are (coupons/dividends are taxed for both). However, I note that the demand for low coupon gilts to lower CGT pushes their prices higher, reduces their yields and, hence, their returns.

Likewise, the disadvantage of the ladder
1) Additional complexity (longer ladder = more bonds) (edit) particularly with rebalancing overall asset allocation

Since the above is a bit handwaving, I can post a somewhat more technical comparison of a rolling ladder and fund if anyone is interested.

Finally, I think the most important decision for fixed income is the target duration to hold since this has way more effect on the returns than whether to hold it in a ladder or a fund. For example, the following figure shows the annualised nominal returns for UK gilts as a function of maturity (1915 to 2023).

There was a peak in returns at a maturity of just under 10 year with a steep decline at lower maturities and a shallower one at longer maturities. Of course, this graph does not show the volatility (which would be greater for longer maturities).

Bostonerimus1

OldScientist said:

Finally, I think the most important decision for fixed income is the target duration to hold since this has way more effect on the returns than whether to hold it in a ladder or a fund. For example, the following figure shows the annualised nominal returns for UK gilts as a function of maturity (1915 to 2023).

There was a peak in returns at a maturity of just under 10 year with a steep decline at lower maturities and a shallower one at longer maturities. Of course, this graph does not show the volatility (which would be greater for longer maturities).

The bonds I own are mostly inside Vanguard multi-asset funds and their average maturity is 8.2 years, maybe Vanguard knows something.

The OP obviously understands more about bonds than most people so I don't think it much matters if they take the ladder or fund approach. The vast majority of people will probably be best keeping things simple and using a multi-asset fund or a bond tracker fund. The key is to have the maturity in that decade range which will also keep interest rate volatility reasonable. I also us MMF and an ultra-short maturity bond funds for cash, but they are cash flow tools rather than investments.

Hoenir

Bostonerimus1 said:

OldScientist said:

Finally, I think the most important decision for fixed income is the target duration to hold since this has way more effect on the returns than whether to hold it in a ladder or a fund. For example, the following figure shows the annualised nominal returns for UK gilts as a function of maturity (1915 to 2023).

There was a peak in returns at a maturity of just under 10 year with a steep decline at lower maturities and a shallower one at longer maturities. Of course, this graph does not show the volatility (which would be greater for longer maturities).

The bonds I own are mostly inside Vanguard multi-asset funds and their average maturity is 8.2 years, maybe Vanguard knows something.

Average maturity of US Treasuries is around 72 months. To sell the paper to the markets both Yellen and Bessent  have taken to keeping issuing short term debt. Hence why the problem of refinancing the debt pile grows ever larger and larger at a faster rate. 

Bostonerimus1

Hoenir said:

Bostonerimus1 said:

OldScientist said:

Finally, I think the most important decision for fixed income is the target duration to hold since this has way more effect on the returns than whether to hold it in a ladder or a fund. For example, the following figure shows the annualised nominal returns for UK gilts as a function of maturity (1915 to 2023).

There was a peak in returns at a maturity of just under 10 year with a steep decline at lower maturities and a shallower one at longer maturities. Of course, this graph does not show the volatility (which would be greater for longer maturities).

The bonds I own are mostly inside Vanguard multi-asset funds and their average maturity is 8.2 years, maybe Vanguard knows something.

Average maturity of US Treasuries is around 72 months. To sell the paper to the markets both Yellen and Bessent  have taken to keeping issuing short term debt. Hence why the problem of refinancing the debt pile grows ever larger and larger at a faster rate. 

Most of my bonds are now outside the US having switched over to Vanguard Global Wellesley fund and I'm 50/50 US/Global ex US equities. But I only have about 10% of my portfolio in intermediate bonds, I have more in cash and ultra short term bonds now.

AlanP_2

BobR64 said:

AlanP_2 said:

Personally I would avoid the issue by using a multi-asset fund at the appropriate split that incorporates bonds and let the manager worry about it.

I can't get my head around direct bond investing but it sounds like you have so well done.

From what I can gleam the main difference between a "direct" bond and a fund is that you control the duration with the former and, if held until maturity, you know exactly what your return will be.

With the latter, say it is a fund that holds 10-12 year bonds. Next week it will sell some and replace them as the duration will have gone outside their remit. The price they will sell for is dependent on the market so anybody's guess really.

While this does seem to be the most straightforward approach, the multi-asset funds that I am aware of are OEICs and these are expensive with HL. I have moved everything into ETFs and a few individual ITs and shares in recent years. Of course, I don't need to stay with HL in the long term, but it's where my work pension is and my wife has her own account, so I would probably prefer to stay for the time being at least. 

The other thing is that I did see someone recommend keeping bonds and equities separate so that you retain the flexibility to sell in different proportions.

What you say about funds vs individual gilts pretty much sums up the issue I have in that you lose the predictability you get from holding to maturity. Funds will end up selling at fixed times without any consideration of the price at that time.

Our mult-asset allocation is via Royal London and works out at about 0.15% per annum all in so to us the cost is negligible, and certainly less than I value my time at.

I know this is MSE and the mantra "costs are bad" is bandied about on here and elsewhere but to my mind there is a tipping point when comparing cost v aggro and for me that was reached when I looked into bonds.