Bonds and Misery

Linton

Being retired I see no point in holding medium or long dated gilts at all.

 -Their return is too low for longer term growth or for medium term income. Their lower volatility compared with equity is irrelevant for both.
 - Short term gilts are more appropriate for covering short term expenses.
 

If you cannot find a convincing objective based reason to hold them, then I suggest don’t.

OldScientist

masonic said:

It would be very informative to overlay inflation and the base rate on the annualised return time series plot.

Looking at the tables and charts below really puts the data in context.

https://www.bankofengland.co.uk/boeapps/database/bank-rate.asp

https://www.ons.gov.uk/economy/inflationandpriceindices/timeseries/czbi

A longer view of RPI is available here

I'm not old enough to remember what savings rates were like during those double digit bank rate years, but if similar to today, then I wouldn't discount cash as equivalent to the short end of the bond yield curve.

I deliberately left off inflation because it affects the different maturities equally (which was what I was interested in for this thread).

There are a few sources for the interest rates on savings accounts - 20 years or so of Martin's emails/newsletter (archived on this site) that usually give the best rates for savings account or 1 year accounts.

And back to 1999 at https://www.bsa.org.uk/statistics/savings-statistics

I have some other info about retail interest rates on savings account that goes back to about 1948, but can't currently find it or remember where I had it from.

For easy access accounts, SONIA/LIBOR/3 month bill rates are probably close enough, but I've not looked in great detail.
For N year fixed: Before tax, the interest rate may be greater than the equivalent on an N year gilt more often than not since the banks are lending at higher rates.

The question I'd ask (but don't have the answer to) is would using savings accounts have made up the 1 percentage point difference? My guess is no.

For full disclosure: Currently just over half of our fixed income holdings are in a ladder of a 1 year fixed rate accounts with the aim of having a mean maturity (or duration) over all fixed income of between 1 and 2 years (currently close to the lower end of that range, but once the yield curve becomes non-inverted, it will gradually head towards the top end). The bonds are global so diversifying away from the UK.

edit: ps in terms of context, it is important to remember that the annualised returns I calculated are over a 30 year period beginning with the year on the x-axis. The values of inflation and bank rates in the links you've cited are instantaneous ones and, in the latter case, do not indicate the returns you would have got over the following 30 years (although it is easy enough to calculate that from the BoE data).

OldScientist

GazzaBloom said:

Analysing bond duration along the yield curve versus holding cash assumes that growth is the priority. That's not why I hold cash. I hold it for certainty, liquidity and stability.

Certainty: I assume you are referring to certainty in terms of the (nominal) principle is going to be there. In the absence of a banking crisis that overwhelms the FSCS, there is a good deal of certainty (although an individual bank going under would delay payment for a few weeks) for cash, but no more certain than holding gilts to maturity (in the absence of government default).

Liquidity. Gilts are highly liquid (5 days per week!) and therefore similar to cash (any crisis affecting one is likely to affect the other). Bond funds would probably only become illiquid in circumstances where individual gilts (and possibly cash) have also become illiquid.

Stability. I assume this is referring to 'price'. For short term gilts (maturities of a few years), the price will not fluctuate that much with changes in yields (and not at all if held to maturity). Even short term bond funds are fairly stable in price, although less so than cash. Long term bond funds (currently including those of the 'all stocks' type) are potentially very volatile as we have seen over the last couple of years.

It is also clear from my second graph, that the maturity that was best over historical 30 year periods varied. At times short duration (and cash) were best, while at other time longer duration was best. The idea of holding bonds/bond funds of intermediate maturities together with an emergency cash fund appears to be a tenet widely held (e.g., bogleheads) since it will generally not be the worst solution (or the best for that matter).

As I've said in my previous post, we do hold a significant proportion (over half) of our fixed income in 1 year fixed rate accounts, but we also hold global bond funds to lengthen the duration and, IMV more importantly, diversify away from UK holdings. Like yours, this is a personal decision, but not necessarily an optimal one in terms of returns.

MK62

GazzaBloom said:

Analysing bond duration along the yield curve versus holding cash assumes that growth is the priority. That's not why I hold cash. I hold it for certainty, liquidity and stability.

You can get that by holding gilts too.........but not with gilt funds........as long as you hold the gilts to maturity.

masonic

OldScientist said:

There are a few sources for the interest rates on savings accounts - 20 years or so of Martin's emails/newsletter (archived on this site) that usually give the best rates for savings account or 1 year accounts.

And back to 1999 at https://www.bsa.org.uk/statistics/savings-statistics

Building a data set from the MSE tips archive is something that has been on my to-do list, so perhaps I should pull my finger out and do it.

I was interested in updating the Paul Lewis (Money Box) "research" on 1 year fixed rates vs the FTSE, but with more appropriate assets.

Annualised rates over longer time horizons can obviously be generated as the geometric mean of a series of 1 year rates, or using the ratio of inflation index values covering the period of interest.

The BSA data looks useful, so thanks for that. The challenge with such data sets is they often give average rates that are dragged down by the volume of low rate accounts on offer.

thunderroad88

BlisteringBarnacles said:

Gazzabloom : I am seeing several threads like this on bogleheads and reddit --> almost implying bonds no longer have a place in a portfolio. I am guilty of starting the thread in a very negative note as well, full of panic.

GazzaBloom said:

I have toyed with the bonds funds puzzle for a few years after not holding any and still don't. I decided to accumulate cash instead to go alongside equity index funds.

My conclusion, which feels right for me is that cash (currently earning 5% BOE base rate) is my retirement “risk-off” money alongside my small DB pension. 

Money I want to invest and take risk with goes into equity index funds. I don't really see any point placing “risk” money in bonds funds. If I want risk it's equities, if I want low or no risk it's cash. Simple.

I thought the point of bonds is not to add risk. That's why one is supposed to go for "quality" i.e govt bonds of developed economies and avoid corporate bonds. And when equities tank, there is supposedly a "flight to quality" so these bonds would benefit, giving your portfolio a cushioning effect. By adding bonds in your portfolio, you are not adding risk, but you have another positive return asset class which is not correlated with equities, and that's the free lunch. This is the theory anyway. Occaminvesting's classic bond article recommends intermediate term govt bonds.

If you hold cash, you wont get the portfolio cushioning effect when equities tank, right ? yes cash returns 5% now but that's why we need to get into bonds now, right, so that if interest rates fall, our bonds would go up in value ? If not now when would you buy bonds then ?

I agree with your second post about VAGS, the one stop solution for bonds, although it has corporates.

The mistake I made was never buying bonds at all. I kept buying equities but I didn't go gung-ho on equities either. But I contributed heavily into pension and had it in the default company fund of 70% stocks, 30% bonds, but when it reached £1 million I moved to cash last year since I wanted to take the 25% out. Bad decision as the market went up a lot after that.

But at any rate, I now have a very large portfolio of which close to 50% is in equities and the rest is in cash. The cash part can easily fund 25 years of expenses at today's UK expenses for a comfy life. I am hesitant to move all of this into bonds (VAGS) in one shot. I am 56 and considering FIRE. Ideally I should have glided, owning both stocks and bonds and gradually easing from stocks to bonds.

Thanks

25 years of cash is way too much. I’d immediately reduce that to 10 years, putting it into fixed term bank deposit bonds of varying duration and something like RL mmfs. That will see you through any equity market downturn. The other 15 years I’d put into equities. 

OldScientist

masonic said:

It would be very informative to overlay inflation and the base rate on the annualised return time series plot.

Looking at the tables and charts below really puts the data in context.

https://www.bankofengland.co.uk/boeapps/database/bank-rate.asp

https://www.ons.gov.uk/economy/inflationandpriceindices/timeseries/czbi

A longer view of RPI is available here

I'm not old enough to remember what savings rates were like during those double digit bank rate years, but if similar to today, then I wouldn't discount cash as equivalent to the short end of the bond yield curve.

Since I had most of the data sets to hand, it has only taken a couple of hours to put the following updated graph together. This shows the 30 year rolling  annualised returns for 2, 10, and 20 year gilts (as in the previous version of the graph) together with those for 3 month bills, the bank rate (assuming there was an investable security that tracked the bank rate) and inflation (sources for the various data sets at the end of the post). (edited for clarification) The year on the x-axis indicates the start of the 30 year rolling period.



A few points:
1) In real terms, fixed income was a poor investment for 30 year periods starting between about 1930 and 1950. However, for 30 year periods starting after the beginning of the 1960s fixed income of all durations has at least given a positive real return.

2) For 30 year periods after the beginning of the 1950s, the returns on 3 month bills and the bank rate have been fairly close to each other (I'll just refer to this as 'bills' from now on). However, there was a sizeable difference between them for periods starting before this - I need to dig around a bit to understand why this might be.

3) For 30 year periods since the early 1960s, the returns on 2 year gilts have consistently exceeded that for bills.

4) Even with the effects of the bond crash, over the last 30 years (i.e., the last data point on the graph) anyone holding longer maturity bonds (10 year maturities captured most of the return) will have seen a return of about 2 percentage points higher than bills or short maturity bonds. This advantage will not necessarily continue over the next 30 years.

5) As an aside, for those trying to estimate inflation going forward, the UK geometric average from 1915 was about 4.6% (only 4.2% from 1900 indicating the importance of start and end dates), but as can be seen from the graph, for 30 year rolling periods, highly dependent on the start year. Anyone wanting to test a worst case inflation might choose 8% (although this would ignore 'sequence of inflation').

Data Sources:
Gilts: as in previous post
Bills: Except for the last few years from A Millennium of Macroeconomic data (available from https://www.bankofengland.co.uk/statistics/research-datasets ). I note that this is annual data, so I think there is a 6 month shift earlier than the other datasets.
Bank Rate: Except for last few years, also from A Millennium of Macroeconomic data
Inflation: Long run series at https://www.ons.gov.uk/economy/inflationandpriceindices/timeseries/cdko/mm23 (annual data goes back to 1800)

masonic

Very interesting and useful, thanks. I am starting to pull together some data to explore the relationship between "best buy" savings rates and bank rate / inflation, although I think I'll be limited to going back as far as 1995, with the best data coming from ~2005 onward. But your analysis above clearly challenges the assumption that long-term cash holdings do not hold their value due to inflation.

Linton

masonic said:

Very interesting and useful, thanks. I am starting to pull together some data to explore the relationship between "best buy" savings rates and bank rate / inflation, although I think I'll be limited to going back as far as 1995, with the best data coming from ~2005 onward. But your analysis above clearly challenges the assumption that long-term cash holdings do not hold their value due to inflation.

To really decide on whether long term cash holds it's value against inflation you need to show the data cumulatively.  Like all such graphs I would guess that it very much depends on when you start.

Also data post 2000 would be of interest.

masonic

Linton said:

masonic said:

Very interesting and useful, thanks. I am starting to pull together some data to explore the relationship between "best buy" savings rates and bank rate / inflation, although I think I'll be limited to going back as far as 1995, with the best data coming from ~2005 onward. But your analysis above clearly challenges the assumption that long-term cash holdings do not hold their value due to inflation.

To really decide on whether long term cash holds it's value against inflation you need to show the data cumulatively.  Like all such graphs I would guess that it very much depends on when you start.

Also data post 2000 would be of interest.

The data OldScientist has shown is for rolling 30 year periods, so 1994-2024 is shown on the right of the graph. There's about a 2% real return for the 30 year periods starting in the 1970s and 1980s, but the gap closes in the 1990s.