Given current economic conditions, are bonds a good investment right now? The dilemma was summed up by one Monevator reader like this:
“I’m led to believe that current bond returns are likely very poor, and they could quite conceivably crash too with interest rate rises. So, are they a good idea…?”
I believe bonds are a good investment still, and the following thought experiment helps illustrate why.
Imagine you’re 100% in equities with a portfolio of £200,000. The economy takes a dark turn, and the stock market falls 25%.
Your all-stock portfolio is now worth £150,000.
It’s a blow but you can handle it. You’ve seen bear markets before and recovery soon follows.
But the market doesn’t recover; it falls another 35%.
Your portfolio is now worth £97,500.
And the economic news is dreadful. Everyone thinks worse is to come. Your plans have been set back years.
Meanwhile your job is at risk. Your industry is convulsing. Waves of redundancy sweep through your firm. You could be next.
An economic depression looms. Your portfolio is your family’s lifeline.
So you panic. You sell. Now your portfolio really is worth £97,500.
The loss is locked-in. It’s real.
Avoiding that is why you own bonds.
Are bonds a good investment? Depends how you judge them…
Bonds are a good investment mainly because they’re a shock absorber that can stop you hitting the panic button.
We all know that equity declines can inflict savage losses on a portfolio.
The UK stock market fell 72% from 1972 to 1974. Some 57% was wiped off US stocks from 2007 to 2009. And near-90% was erased from the Dow Jones during the Great Depression.
Few of us really know how we’ll react, though. A crisis feels like a crisis because we can’t see the bottom of it.
The overriding point of government bonds is to protect you from the urge to panic.
The reason minimal-risk bonds can do that is encapsulated in the phrase: the flight-to-quality.
When the economy implodes, a money-tsunami flows out of equities and into government bonds.
Investors assume that, when the smoke clears, bond-issuing countries like the UK and the US will still be standing and paying their bills – even if their companies are in dire straits.
That time(s) when bonds saved the day
We can see evidence of the flight-to-quality at the low point of many historical downturns.
When the dotcom bubble burst, gilts gained 14% from 2000 to 2002 as UK equities sank 41%.1
In the following additional case studies, ETF returns data and charts come from the portfolio-building service JustETF. The blue line is UK government bonds and the red line is world equities.
Coronavirus crash 2020
The developed world stock market (represented by iShares MSCI World ETF) fell -26% by 23 March 2020.
UK government bonds (represented by iShares Core UK Gilts ETF) rose 4% by the same date, as investors bailed on equities and took refuge in bonds.
Global Financial Crisis 2008-09
The World ETF crashed -37% by 6 March 2009. Gilts softened the blow by moving 14% in the opposite direction.
European Sovereign Debt Crisis 2010
Equities slumped -15% but UK gilts were up 5% by 2 July 2010 and 9% by 31 August.
Global Stock Market Downturn 2018
World equities delivered a lump of coal on Christmas Eve 2018 as they slid 16%. Christmas cheer came from gilts – up over 1%, and near 5% in March.
August Stock Market Downturn 2011
Equities dropped 19% while gilts countered with a 5% gain by 19 August 2011 and 11% by 29 November.
Guilty pleasures
Time and again we can see the flight-to-quality effect, as investors sell off equities and flee into government bonds.
Gilts have improved portfolio returns – thereby suppressing the panic reflex – in the majority of UK downturns for over a century.
They don’t work every time (and we’ve talked before on what improves your chances) but the fact is high-quality government bonds usually limit the damage when equities plunge.
Other assets don’t come close. Not emerging markets, tech stocks, infrastructure, dividend aristocrats, commodities, property, cash, corporate bonds, gold, bitcoin, or even index-linked bonds.
A diversified equity portfolio does not help in a crisis.
Stock markets buckle like carriages in a train crash in a globally interdependent economy. You can see this using any stock market charting tool.
You’re not diversified if your holdings all behave the same in adversity.
That’s why true diversification means spreading your bets across asset classes, rather than just among the many shades of equity.
One last time: high-quality (developed world) conventional (not index-linked) government bonds are the best asset class to own when recession hits.
For UK investors that means including gilts or a global government bond fund hedged to the pound when choosing your asset allocation.
Expected bond returns
It’s true the outlook for bonds is poor.
The expected return of your bond fund is its current yield-to-maturity (YTM).
To take one example: the YTM of the SPDR UK Gilt ETF is 0.85% as I write.
Subtract 2% for inflation and an intermediate gilt fund is liable to return a small loss in real terms2 for the foreseeable.
The historical real return of UK government bonds is around 1.4% a year, so it may seem unacceptable to lose, say, 1% per year now.
But that nick is much more acceptable set against the risk of locking-in a double-digit loss during a stock market panic.
Call it an insurance payment.
Bond market crash
What are the chances bonds could inflict a more grievous loss?
- When market interest rates fall your bonds make a capital gain.
- But when market interest rates rise, your existing bonds take a capital loss.
The fear of a bond market crash resonates because interest rates are lower than my chances of winning Olympic Gold on an all-chips diet.
Pundits predict the only way is up for rates. That implies capital losses for bonds.
There’s two important things to know, however.
1. Pundits have been predicting a bond market crash for over a decade.
Stopped clocks may strike it lucky but financial markets can defy received wisdom for decades – and indefinitely after a so-called ‘regime change’.
There’s no law that says interest rates must return to the old normal. Indeed there is evidence that rates are in secular decline. And there’s plenty of runway yet for negative yield bonds.
2. Bond market crashes are a mild after-dinner burp in comparison to the Krakatoa-esque implosions of the stock market.
Compare these two charts showing 120 years of UK equities and gilt returns from the Barclays Equity Gilt Study 2020.
Equity returns are dispersed like a shotgun blast: anywhere from over -50% to over 60%. Note the frequency of corrections (-10%) and bear markets (-20%). Major slumps occur regularly.
Gilt returns are overwhelmingly distributed between -6% to 6%. There are only three annual losses greater than 20%.
Point being the outlier bond crashes marked on the agony-to-ecstasy chart above do not match the equity market’s extremes.
Gilt bear markets and corrections happen much less often than more familiar equity woes.
The worst return for gilts was -33% in 1916 amid the bloodbaths of World War One. That’s grim but still, the stock market would sniff “Hold my Beer”.
Note too, the worst three negative results are amplified because the bonds being measured were undated (1916 and 1920) and 20-year long (1974) bonds.
The longer your bond takes to mature, the more vulnerable it is to inflation.
Rampant inflation hammered all bonds during these periods. But the results look especially horrendous if you track the long varieties, or worse, undated bonds that never mature.
In today’s environment, you can lower risk by using intermediate or short duration bonds. They will react much less violently to interest rate rises.
The worst annual loss since Barclays began tracking 15-year gilts is -14% in 1994.
Not all bonds are alike
Outcomes can vary dramatically depending on the average maturity date of your bonds. The farther away the maturity date (that is, the ‘longer’ the bonds), the harder they could fall (all things being equal).
Bond funds use a metric called duration that indicates how they’ll perform if interest rates rise or fall 1%.
For example:
- A bond with a duration of seven years will lose around 7% of its market value for every 1% rise in its interest rate.
- A bond’s price will similarly shoot up by about 7% if its rate falls by 1%.
Whatever your bond’s duration number, that’s how big a gain or loss you can expect for every 1% change in interest rates.
You can control your exposure to downside risk by choosing a bond fund with a lower duration.
For example, SPDR ETFs offer short, intermediate and long gilt funds with very different durations:
| Bond fund type | Bond fund name | Duration |
| Short | SPDR UK 1-5 Year Gilt (GLTS) | 3 |
| Intermediate | SPDR UK Gilt (GLTY) | 13 |
| Long | SPDR UK 15+ Year Gilt (GLTL) | 21 |
Source: SPDR
Imagine interest rates rose by 1%. With these funds, you’d brush off the 3% loss inflicted on the short-dated GLTS much more readily than the -21% gouge out of GLTL’s long bonds.
On the flipside, eagle-eyed readers will pounce on the fact that if interest rates drop 1% during a recession, you’d gain 21% from GLTL versus a limpwristed 3% hoist from GLTS.
That’s the Damoclean choice we make:
- Short bonds are more resistant to rising interest rates but they offer less stock market crash protection.
- Long bonds can inflict equity-like losses if interest rates hike significantly, and equity-like gains if rates drop a percentage point or two.
Thankfully, intermediate bond funds offer a third way. They blend bonds across the maturity spectrum into a single fun-pack.
Remember too that rising interest rates mean higher future yields for bond investors as your fund manager sells off the old and buys the new. If your interest payments are being reinvested then they’ll buy new bonds at cheaper prices than before the rate rise.
The maths shows that a bond fund will eventually recover its initial capital loss and end up ahead due to a rate rise.
The duration metric reveals how many years that will take (approximately).
For example, a duration 7 fund will breakeven around seven years after the rise.
Note: I’ve had to simplify some of the above for sanity’s sake. Use duration as a rule-of-thumb rather than a written guarantee from the Queen.
Interest rates are forever fluctuating, and this doesn’t affect all bonds evenly at the same time.
See our previous deep-dive into how bond prices react to interest rate changes for more (and check out the appendix below this piece).
If you’re especially keen, you can also glean interesting information on how rapidly interest rates can change in the modern era from Monevator reader and professional bond manager ZXSpectrum48K.
Good bond investments
Choose low-cost index trackers that hold gilts, or high-quality global government bonds (developed world) hedged to the pound.
Hedging to the pound negates exposure to currency risk, which would otherwise add volatility to your defensive allocation.
A total bond market fund that includes high-quality corporate bonds is okay. It’s likely to perform worse than straight government bonds in a recession though.
Accumulators
Think intermediate bond funds. Long bonds could mean a world of pain and short bonds barely spike in a recession.
An intermediate bond buffer should leave you with plenty of dry powder to buy cheap equities during the next crash.
A 5% to 10% slug in cash and gold further diversifies your defences in a downturn.
Some people hold no bonds in favour of cash. But cash won’t counter-rise in a crash. Cash yields are also currently negative, after-inflation. Some cash is fine but don’t go crazy.
Similarly, don’t go overboard on gold. It’s performed very well during the last two crises but its long-term track record is patchy. Gold is like a funny drunk – erratic and highly unreliable, but occasionally brilliant.
For good inflation protection you can put about half your bond allocation into index linked bonds. This is the best asset class to defend against runaway inflation. However it doesn’t do much during standard recessions.
The young and new to investing assume they’ll be fine with 100% equities, but that isn’t necessarily true. Some people just can’t stand losing money.
Read this piece on estimating your risk tolerance to know thyself.
Decumulators
There aren’t good alternatives to bonds for those spending down a portfolio. The same forces that previously boosted their portfolios now leave decumulators bereft of defensive assets that can generate a return.
It’s best to suck it up. Think of bonds as an insurance policy and thank your lucky stars for the bull markets that got you this far.
Given elevated stock market valuations, now isn’t the time to ramp up portfolio risk, though it’s a tricky balancing decision.
Personally I’m holding no more than 40% of my portfolio in cash and bonds, in this climate. But at the same time I’m not deluding myself that I’d be safer by going more overboard on shares, either.
Bonds help you manage your cashflow. Short bonds are for near-term bill-paying needs. Intermediate and index-linked bonds help create an all-weather portfolio that should see you through the decades ahead.
What if someone pushes you a product that supposedly has the defensive capabilities of government bonds based on their proprietary back-testing?
Well, they’re probably trying to sell you something, aren’t they?
If you’re tempted to reach for yield then read the wise words of Warren Buffett. (He thinks it’s a bad idea.)
High-yield bonds, low volatility equities, and dividend aristocrats offer the mirage of a circle squared – but they belong in your risky asset bucket.
Are bonds a good investment? The short answer
In a crash situation, bonds can be priceless because they can cushion the losses that turn a crisis into a full-blown panic.
Yes, conventional bonds are vulnerable to an escalation in interest rates, especially if the economy overheats.
That’s ‘if’.
Equities are vulnerable to a massive stock market crash.
That’s a matter of when.
Take it steady,
The Accumulator
P.S. The lived experience of others can help us imagine the unimaginable. The Great Depression: A Diary is a contemporary account of an economic cataclysm that scarred a generation. I had a new respect for bonds after I read this book. Bonds kept people afloat as their equity positions disintegrated.
Bonus appendix: Bond market interest rates
‘Interest rates’ in the context of bonds does not refer to the central bank interest rates we’re used to.
Instead, we’re talking about the interest rates that apply on the bond market.
Each and every bond is subject to its own interest rate that’s a function of its supply and demand.
This ‘market’ interest rate is the return investors require to invest in that bond.
Bond interest rates fluctuate constantly as the market’s view adjusts in line with new economic data including inflation, the bond’s credit rating and maturity date, and – yes, those all pervasive central bank interest rates.
It’s an incredibly deep and liquid market. Are bonds a good investment in 2021? As ever think very hard before deciding you know better than the sum of the world’s investors!
Comments on this entry are closed.
The history of stock/bond correlation is not as straightfoward as “when stocks go down bonds go up” and there have been periods where both move in the same direction.
The 60/40 model works so long as a crash is a deflationary event, however in an inflationary recession all stock/bond portfolios get crushed. It is sensible to hold some amount of real assets as a hedge against this scenario.
Interesting and useful article. So are bonds contained in Vanguard Lifestrategy products high-quality, conventional government bonds?
> Stock markets buckle like carriages in a train crash
What’s your source for this? Media reports over recent decades make me think train carriages are typically strong and get thrown about rather than buckled.
Could there be an argument that holding cash instead of bonds is a justifiable approach? If the main benefit of bonds is to “cushion the losses”, then wouldn’t that capital be better put to use when equities crash by dollar cost averaging?
I know there’s an element of timing the market here, but rather than investing 60:40, having that :40 ready to buy equities could pay off very well in the future. If you know you’re tough enough to not panic sell and just see it as a buying opportunity, surely having a sizeable amount in cash rather than bonds would be advantageous. Or am I missing another benefit of bonds?
@Gaz
The point made in the article is that cash will not increase in value during a crash, but bonds may and often do. Rebalancing to your predefined asset allocations would then cause you to sell some of the (increased value) bonds and buy some (reduced price) equities. This is exactly what happened to me in March 2020.
It is just as easy to liquidate some of your bond ETFs as it is to add cash to your broker account – possibly easier I would say.
If you have an allocation of cash instead of bonds the same rebalancing can occur but the effect is not likely to be as significant because the cash hasn’t gone up in value.
Much is made of the flight-to-quality effect here but the expectations of a rate cut account for the vast majority of historic bond rallies upon a market crash. The cross asset class flows are nothing compared to bond market liquidity. Since rates are about zero, this is much less likely next time. Not impossible, plus there are asset purchases to think of, but this is the crux of the current bond fears
Bonds? not for me. Very limited up side at the moment and they are very vulnerable to inflation.
I don’t go 100% all in on equities, but I prefer to hold 10-20% cash, rather than say 30-40% bonds. Cash is simple and liquid.
And if equities drop more than 20-25% from the peak, and i think the world has gone mad then I’ll move some of the cash into equities. Should be a decent bet. And I am doing my bit, helping to provide some liquidity to the markets …
Would be interesting to see the back testing results over a few decades, of 80 EQ/ 20 Cash, vs 60 EQ / 40 Bonds, after all dealing costs, tax, etc. Do bonds really stack up, even in a low inflation epoch? They have the odd good year or two but they get left for dead in the boom times.
Would agree that holding some tangible assets might work better than bonds, or just cash as the only hedge against equities. Gold, if you must. And sad (if we care about inequality etc, and I think we should) to say, UK rental properties also look like a much better long term bet than UK bonds. About a 3% yield after tax and costs, plus an asset that in the long run should roughly rise in value with inflation.
btw are those charts including distributions – dividends / bond interest payments / etc? the ‘Dist’ bit got me wondering.
re market trends – in the UK a lot of govt bonds are held by funded DB pension schemes and annuity providers. the former (and possibly the latter) are likely to decline as a share of investable capital, as funded DB schemes wind down. A steady stream of willing sellers. Barring a huge increase in life expectancies, I cant see a corresponding stream of willing buyers… which seems to tilt the odds towards bond values staying flat or declining.
Shout if I am missing something!
D
I get the point of the bond protection in a market crash and agree it’s absolutely the right thing for a lot of people. But it’s not the only solution that it’s often presented as and I think it’s helpful to go through what that ‘insurance premium’ is actually buying you, like any other product before deciding if you need it.
E.g. I’ve always liked Ermine’s analogy of his final DB pension in effect being a big bond holding – so why hold more. If you have the cash-flow to ride out the crash and have been through a couple to know you aren’t a panic seller, then it’s not as obvious they are worth it.
I’d be curious to see an overall piece of opportunity lost cost of holding bonds through all the non-market crash periods. I.e. would the gains made on investing the money differently outweigh the eventual market crash/bond gains?
@ Vano – yes, I made that point in the piece – index-linked bonds are your best bet in an inflationary scenario, other assets have worked at times but have a patchy track record.
@ Ali – for the most part that’s right. IIRC LifeStrategy also have a smattering of investment grade corporate bonds. That wouldn’t bother me in exchange for the convenience of LifeStrategy.
@ BBB – Poetic licence 😉
@ David R – the classic recommendation of equities:bonds, or equities:bonds:cash exists because bonds have beaten cash historically. If they hadn’t then people wouldn’t take the risk of holding them. It’s the usual risk:reward story.
Bonds get left for dead in the boom times. Yes, they should as should cash. Although if you pick your dates, you can create back-tests showing bonds beating equities because of the losses inflicted during stock market crashes like the dotcom bust and the Global Financial Crisis.
Gold is no replacement for bonds but I agree it’s a useful part of the mix, as is cash.
re: willing buyers for pension fund offloads – I’m no expert but I can think of a few candidates:
The next-generation of accumulators, automatically enrolled in workplace pensions default funds that buy government bonds.
Overseas buyers from middle income countries that previously didn’t have the income to invest.
Our own government keeping the lid on interest rates through asset purchases.
Re: charts – they include income payments.
@BeardyBillionaireBloke #3
I agree, the carriages are strong. It’s the train as a whole that buckles, at the weak points between the carriages.
@ Michelle – that’s a really good point. I didn’t put my mum in bonds because her State Pension and annuities perform exactly the same job – a steady stream of income payments that enable her to ignore the volatile equity side of her portfolio.
If you model 100% equities vs any equity:bond portfolio then *eventually* equities win.
What that ignores is:
The irrevocable damage caused to those who panicked and sold out. Think of those articles titled, “If you bought $1000 of Amazon shares in 1997 then you’d be worth $$$$$$$$ now!”
Those pieces gloss over the fact that you had to hold on through several unholy cave-ins of the stock including losses of over 90%. The untold story is that very few have the stomach for that kind of volatility.
Also, think about all the sustainable withdrawal rate simulations of 100% equities vs portfolios with bonds.
Those sims show that you can leave this Earth richer than ever with 100% equities or bust within 25 years. Again, the reason is volatility. A good sequence means you’re minted, a bad sequence means you’re broke.
We diversify because we don’t know what we’re going to get.
@Marcus
Good point, rebalancing is of course an option. I guess my fear is more related to negative bond performance resulting in you having less capital to then buy equities… although the same argument could be made for cash being eroded by inflation.
@TA – thanks for yet another really good and well timed article.
For me, Bonds are my steady(ish!) store of value. If they beat inflation, great. If they don’t get flattened by interest rate rises, even better. They’re not to earn returns. That’s what Equities are for.
Nobody’s mentioned Bitcoin yet
Sorry, couldn’t resist.
Thanks for this article. I do have a serious question: how reliable/useful are “rules of thumb” such as 60/40 or Equity percentage = 100 – age ?
As somebody approaching retirement,I hold my age in non-equity products i.e. half in cash ISA’s,and the other half in intermediate term gilts. It may not be very scientific, but I can sleep at night with this allocation.
If our motive to buy longer dated gilts (ie GLTL) is that rates might fall – and yes, there remains runway, but would less runway be percieved to remain? – If there is ultimately a limit on how low it can drop. And that might mean the capital gains from a drop at a percieved bottom might be less than a drop from a high place?
Or is it the opposite? If a drop from a low place (i.e from 2% to 1%) meant that people seeking a set amount of income have to put a much higher amount than if it dropped from a high place (say 10% to 9% for arguements sake) then the capital gain might be more if it drops from a low place?
Also could we consider cash to be zero duration gilts? The arguements between cash vs bonds seems similar to the arguement between short vs long – bonds are simply like adding duration and default risk & rewards to cash.
@Marcus, @Gaz, I too was surprised @TA didn’t write about rebalancing. I seem to remember an article somewhere (very likely on Monevator) showing that an 80:20 portfolio with rebalancing fell very little short of 100:0 but with a significant reduction in volatility. Due to rebalancing meaning that effectively you sell on a high and buy on a low.
The “right” share:bond ratio is a conundrum. It is obviously important to move to low volatility bonds/cash in preparation for an encashment (like buying an annuity with a DC pension fund). However I can’t see that compelling a case for changing the potential for growth with age – unless you are within sight of running your savings down to nothing which amounts to the encashment situation. Nevertheless unless you are putting money away for the really long term, the interest rate risk of bonds and the volatility risk of equities mean that for me it would feel risky for either to exceed two thirds of the total.
Perhaps this a prompt for Monevator to re-visit portfolio allocations. Does the latest analysis give any good basis for choosing between 40:60, 50:50 and 60:40?
@ John – It’s reasonable to think that interest rates may well rise, but much of the fear is based on the assumption that they *will* return to the old normal – that expectation has been defied for over a decade now. Or the fear is based on ideas like low interest rates can’t go lower.
The ten year yield is 0.82% at time of writing and over 1% for longer bonds. We don’t know how much negative yield the market would accept in a crisis but we already know investors are prepared to go below zero.
The index-linked bond market has accepted yields of -3%.
Zero is not the baseline and that provides headroom for capital gain.
Take a look at this brilliant article about how bond capital gains are amped as yields turn negative:
https://portfoliocharts.com/2019/05/27/high-profits-at-low-rates-the-benefits-of-bond-convexity/
@ Tony – even 15 years ago some were advising 110 or 120 minus your age to account for longer life expectancies. Like any rule of thumb, it’s not a mathematical proof but its heart is in the right place.
@David R (#8):
Re pension fund offloads:
AFAICT, DB funds hold bonds to match their liabilities, that is their bond holdings maturities are shaped in time to match their foreseen liabilities – ie pension payments to members. Short of a major change in DB funding strategy and/or very significant revisions [downwards] in life expectancy and/or legislation I am not sure that DB pension fund offloading of bonds is really a thing.
IIRC, it is a bit of a modern myth that annuity companies favourite tool is gilts.
@Michelle (#9):
I too like the analogy of a DB and/or state pension and/or annuity to a bond holding.
However, I seriously question the usefulness of an overall traditional asset allocation – as used in accumulation – for de-accumulation.
Always good to have my ideas as a gilt-sceptic challenged by these articles.
What has flagged up another safe haven alternative is the following statement – “I didn’t put my mum in bonds because her State Pension and annuities perform exactly the same job – a steady stream of income payments that enable her to ignore the volatile equity side of her portfolio”.
This has got me thinking about de-risking during retirement by using a % of my portfolio to buy an annuity (instead of holding gilts) to both temper market volatility and help enable an overall 4% withdrawal rate to be sustained, even if the annuity part does get hit by inflation over the longer term.
And if I’m concerned about the impact of inflation on a level annuity then it is going to be even worse on a gilt fund yielding 1%.
As in pre-pension freedom days I guess you’d still hold gilts in the run up to retirement in order to earmark for the annuity, so you don’t have to sell equities in a bear market.
@Al Cam – I’m sure you’re aware of the excellent work Michael McClung/ERN has done on reverse glide paths. It’s what I’m doing.
Any views on the merits of so-called “wealth preserving” investment trusts like Capital Gearing and Personal Assets vs pure bond-based investments, at a time like this?
Just my penny’s worth as I am further down the road than some investors here
Aged 75 retd 18 years
Walked the walk
65% of portfolio in a Vanguard Global Bond Index Fund hedged to the Pound ((VIGBBD)
(Has some corporate bonds but mostly Govt bonds etc)
Made a big enough pile so no pressure on me
This fund reduces the volatility of my portfolio -it’s main purpose
Held it this percentage of my Asset Allocation for many years
NAV has increased at 4% pa-nice bonus
Simple cheap and easy to understand
So far so good……….
It is however all very personal-how much risk you are willing to take?
How much you need to withdraw from the portfolio etc?
xxd09
“To take one example: the current yield of the SPDR UK Gilt ETF is 1.93% as I write.”
That’s the running yield, which is a bit misleading as it depends on the coupons on the gilts. For a given yield to maturity, the higher the coupon, the higher the running yield, but also the lower the capital gain if held to maturity, or at present, the bigger the capital loss! Quoting the yield to maturity is better and that is 0.85%, although that does not account for the slope of the yield curve. Then there are the management charges of 0.15%, bring the expected return down to 0.70% (ish).
I came out of gilts and US Treasuries a year ago and flipped to a new decumulation strategy involving equities and cash only. Had we held the bonds we would have lost about 5% on the gilts and about 15% on the US Treasuries ETF.
We only keep around 10% in cash, but if we held more I would be tempted back into US Treasuries before I bought gilts again. Treasuries carry exchange rate risk, but in a market panic there is often a flight to the dollar as well as a flight to bonds, so they are better diversifiers in that scenario, IMHO. I would never hold gold.
@Jonathan B, In decumlation, history indicates that a high allocation to bonds helps get you a higher SWR in those cases when stock markets do badly, and/or you get a poor sequence of returns. Most of the time though bonds are a hinderence and lower bond allocations would give rise to higher SWRs, particularly for long drawdown periods. As you only know what would have been best with hindsight, it is prudent to err towards the higher bond allocation in case you end up needing them!
60:40 Equity:bond appears near optimal in most worse case scenarios and for a typical 30 year retirement. Going much higher with bonds runs the risk of not getting sufficient growth to make it through. If the horizon is lower, then you need less growth and can take on more bonds for safety, but if you have a longer retirement in mind, or want to leave a legacy, it is better to lower the proportion of bonds, which also means lowering the withdrawal rate.
We held 60:40 for many years in both accumulation and decumulation. With the benefit of hindsight, it was sub-optimal. Last year we flipped to 90:10, which is aggressive, but I am currently reckoning on an “indefinite” retirement period for the 2 of us, using a low withdrawal rate. Historically, that points to needing a low bond allocation.
We now hold cash rather than bonds, which is definitely not great, but I don’t want to take on the risk of intermediate to long bonds falling, even if that means missing out when they spike upwards. Short dated bonds are not worth holding as better returns are available on FSCS protected cash deposits for those prepared to shop around.
Thank you for this one @TA I still struggle withe Defensive part of my portfolio and it will get even more important as I move to drawdown an income.
The distribution of bond returns in the Barclays study is interesting, but do think it’s worth noting that intermediate Uk Gilt ETFs went down about 8% last year.
I personally have decided to stick with short term gilts and index linked bonds for a bit.
It’s a very pertinent article as ever. If you are slightly concerned by todays decline in markets, or last year’s volatility or the fact it’s very feasible for equities to return zero over a decade then bonds need to be given consideration by you. Equally if you are just starting out or still earning considerably with years to go they likely have less of a role.
@David R (7) – ‘Bonds? not for me – very limited upside at the moment’. I believe that’s potentially incorrect. 10 year gilts / us treasuries yield very little at the moment. But at these levels convexity has a substantial effect on bond returns and if those yields were to drop to zero you would see significant upside. Equally there is significant downside at this level – if yields rose to say 4% there would be very substantial losses. Which is what can also understandably put people off. The article linked by the accumulator explains it well. I say potentially incorrect as I just illustrate a scenario where you can still make substantial returns from bonds – not a prediction.
If your state pension (combined) covers necessary expenses, then your bond allocation can reduce although it’s worth considering the implications if one of you passes away.
FWIW (nothing) I hold cash, $TIPS and gold (buried….yes its crackers!) as liquidity as I remain accumulating. Were I to be de-accumulating, I guess I would hold my nose and hold some £ intermediate duration bonds.
Mike Johnson @24 – Personal Assets Trust holds 13% of their assets in cash and UK T-Bills. They don’t hold any GILTS of any duration with the rest of their portfolio in Gold, $TIPS and stocks. So it’s not really comparable. You could infer (I think correctly) they do not favour conventional bonds currently probably due to the downside risk amplified by low rates (convexity). Remember they are active investors though. Think Capital Gearing are of the same ilk although from memory they are deep in ILG and have a low gold holding.
XXd09 @25 – You seem the classic case on this forum for someone who has benefitted from +ve sequence of returns across bonds & stocks due to declining interest rates. Sounds like from your previous comments across various subjects that you have deserved your luck and I am very pleased for you! Obviously if you are retired in 1970 it would have been a total shocker ten years later though given rising interest rates!
Naeclue – I quite like this strategy which would have been great for say 2008 and 2020 volatility but wouldn’t have saved you in the 1970’s. There’s no magic button as no doubt you know. Anyway a lot to like about this generally at the moment – your comment in 27 is spot on but if the next ten years are a disaster then intermediate bonds will have a place. Who knows.
There’s something I don’t follow – we are told/sold the concept that the change in bond price equals the duration x the % change – i.e. for a 12yr duration bond, a 1% drop in interest rates will increase the price by 12%. However if I look at the above examples, let’s take the Financial Crisis – the BoE base rate dropped from around 5% in April 2008 to 0.5% in March 2009 – the effective duration for the iShares Core UK Gilts ETF is 12 years, so that should have increased in price by 54% but in the above plots it only increased by 14%. Why is there such a big difference? or had it already gone up 30% in anticipation of the interest rate drops prior to April 2008?
@David The Bank of England rate does not determine the bond yield (as explained in the excellent article above). It’s the bond market which determines the yield, based on buyers vs sellers (supply and demand).
David – BoE base rate is not gilt yields. Yes BoE base rate fell from 5% to 0.5% but the below link (random google) shows the 10 year gilt yield (not a great proxy for ishares core UK gilt given the duration of 12 years but okay enough I think) fell from just under 5% to between 3 and 4% during your time period. This largely explains your q I believe but if others disagree do shout.
https://touchstoneblog.org.uk/wp-content/uploads/2012/07/10yr-gilt-1998-2012.jpg
Thanks @Naeclue, good to know that 60:40 is somewhere near the mark. I assume a switch to 90:10 or similar in your case was made once you knew you were in a position that a large equity drop (?50%) would not undermine your ability to support yourselves.
While I see the case for substituting bonds with cash in the near term, I am not clear it would do the same job longer term. The other issue I struggle with – I see why DB pensions are described as “bond-like” but how should you adapt your investment weighting as a result?
@Seeking Fire, my strategy would have worked through the 1970s with UK shares (FTSE allshare). Last year I started with 5 years spending in cash and 50 years in equities. The strategy I tested was to spend from the cash and top up from the equities each year when the RPI adjusted amount exceeded the starting amount, but to sell one years worth of equities each year if/when the cash buffer ran out. Starting at the end of 1968, the worst year in my backtest, my strategy would have exhausted the cash in the first 5 years and I would only have had a little over 15 years in equities left by the end of 1974. But there was a huge jump the following year and the value of the equities would have exceeded the RPI adjusted starting amount (50 years spending) by 1984, allowing the topping up of the cash buffer. From then on it would have been uphill most of the way.
A conventional 60:40 equities/intermediate gilts portfolio, rebalanced annually would also have worked, but it would have taken until 1998 for the RPI adjusted starting amount of equities to be exceeded, by which time I would have had 174 years worth of equities with my 50:5 portfolio.
Testing with UK average earnings instead of RPI was tougher, but still would have worked. By the end of 1974 there would only have been 12 years of equities left and it would have taken until 1993 to recover the earnings adjusted starting amount.
As of the start of this year, the portfolio was sitting at 6 years cash and 60 years equities, so we should be even more resilient, but the 1970s experience would have been painful no matter what anyone did and in reality I would likely have done some early belt tightening.
@David, As @Gizzard and @Seeking Fire have said, the BoE base rate droped by 4.5%, but the yield of the gilts ETF did not. If it had though, the price of the ETF would have risen by more than 54%. The reason for this is that the price change is only approximately equal to the duration times yield change. As @TA said in the article, “A bond with a duration of seven years will lose AROUND 7% of its market value for every 1% rise in its interest rate.” This “AROUND” is correct to first order, but the relationship between the price of a bond and its yield is not a linear one. The bigger the change, the more inaccurate the approximation.
Just to get a sense of the issue, consider a zero coupon bond maturing in 12 years instead of the ETF and assume it is yielding 5%. The price of £1000 nominal of that bond will be about £556.84. This is because £556.84 compounded at 5% per year for 12 years is £1000 (556.84*(1+0.05)^12). By similar logic if the yield fell to 0.5%, the price would rise to £941.91 as 941.91*(1+0.005)^12 = 1000. But according to the duration approximation, the price should have gone up 54% to 556.84*1.54 = 857.54.
The second order term (convexity) can be used to calculate a more accurate answer, but again the result would still only be an approximation.
ps, should have mentioned that for a zero coupon bond, the duration is the same as the maturity.
@Brod (#23), @Jonathan B (#34):
I am familiar with the rising glidepath (or bond tent) idea.
FWIW, I think de-accumulation – especially for those with DB pensions – is better framed as Floor and Upside. Some recent related chatter on this is at: https://simplelivingsomerset.wordpress.com/2021/04/16/coiled-spring-and-missing-claw/
@Mike (24) I’ve considered Personal Assets Trust a lot for the defensive part of my ISA but always struggled with the idea of @TA’s disapproving look when the charges are considered. I’ve ended up using HSBC Global Cautious as a cheap alternative for it (basically the same thing as VLS20). Despite all the brilliant reasoning on here for bonds I still struggle with it. Knowing how my brain works and the danger of ongoing tinkering, paying for something like PAT isn’t the worst idea – I’m basically buying some piece of mind (at a material cost I know). It’s also potentially better than what I suspect many people do which is file it in the too hard column and just have 100% equities and hope for the best.
@Andy. Disagree with regard to Personal Assets Trust (PNL LN). It’s just a complete mismash of asset classes. It’s about 15% cash/T-bills, 10% Gold, 30% US linkers, 45% equities in quite concentrated holdings. Not clear what the FX hedging policy is.
Just replicate it with a by holding some cash, a Gold ETF, US linker ETF and a equity ETF. Save on fees but more importantly you actually know what you’ve got.
John Bogle of Vanguard fame,s rule/guideline is still a useful touchstone once the principal of using bonds in your portfolio is accepted
ie Your age or your age minus 10 should be the % asset allocation of bonds in your portfolio
100% equities or 100% bonds is to be avoided
Served me me well through my accumulating phase with growth rate and ability to sleep at night,stand still through downturns etc
At 75 -30% equities suits me -may increase % slightly as I age if I can be bothered-probably not
Simple rule and easy to understand and implement
Good starting point for most investors
xxd09
As ever it’s somewhat befuddling to see a thread of dedicated passive investors noticing bonds are expensive and making strong pronouncements about what this means for returns. 🙂
You guys are smart and so you know it’s a banality (for 10 years…) to say that bonds are expensive.
You also know that this is one of the deepest and most liquid markets in the world?
So none of this is edge or brilliant insight.
I’m not being snarky. I’m wondering why you think you can’t have a view about whether a smaller illiquid company like, say, Hotel Chocolat (I currently hold) is under/overpriced, but you believe you can bet against the multi-trillion bond market?
Don’t get me wrong, I agree bonds look expensive, too, and I’ve been very underweight for years. More fool me. 🙂 But I’m an active investor so that’s on me, and I’d hope other active decisions I make will compensate.
Someone mentioned a particular bond tracker had fallen last year. This is pretty irrelevant. Bonds are not cash, they will go up and down. It’s utterly expected that they will. The graph of a bond ETF will have peaks and troughs, like a share graph, albeit less extremely peaky and trough-y. And there will also be long-term secular trends (such as the general march downward for bond yields, until very recently).
This is why it doesn’t matter what bonds did last year, from a passive active asset allocation perspective. You might as well say the stock market fell yesterday, so that’s why you shouldn’t own stocks. No, it’s why you should remember stocks are volatile, and invest accordingly. Similar (but different) with bonds.
I think a mix of bonds and cash is the solution. As always, very rarely if ever should it be all or nothing. I’m very sympathetic to the view that government bonds have less firepower than they used to, both in return terms and in market-reaction counter-balance potential. But that is how things are today, wishing it was different doesn’t help. It’s also a reality that sits alongside generally expensive equity markets.
If you’d rather be very overweight equities and are sure/happy to sit through a 30-50% crash when it comes then that’s fine. I’ve been like that myself several times in the past, and even today I’m basically 70% in shares, which I judge high-ish for various reasons for me right now.
Just be sure you know that’s what you’re signing up for, and the recovery can take years (decades). Don’t get grumpy when (not if) it happens. And yes, hopefully the extra equity allocation might have made you sufficient gains to offset some of that drawdown versus if you’d held a heavy weighting of bonds/cash. (Hopefully. 🙂 )
(All these comments are not directed at anyone in particular. Fully understand most of you get this, just trying to summarize. 🙂 )
As for Personal Assets, I like keeping an eye on it and that it exists but only as an active play on a manager’s skill. (Because they are very clear who they are and what they’re about. You can of course say you don’t like that or want it! Most shouldn’t.)
If you’re a passive investor you shouldn’t go near it with a bargepole, not least because as @ZXSpectrum says you can do more the same more cheaply. It’s not like it’s giving you access to assets you can’t get easily elsewhere. It doesn’t even have the uber-diversification of a Capital Gearing Trust, say. With PNL you are paying for management skill in juggling the assets faster/to-a-better-return than you can copy how they juggle the assets. That doesn’t make much sense for a passive investor at all. 🙂
Naeclue – sorry you are correct – yes – it would have worked – by not working I meant not having very significant volatility, which as you say it would have done. I think with your strategy (any most others…..) by 1974 – I would have been significantly cutting back expenditure and having a few sleepless nights knowing my nature (not knowing the future). To avoid those sleepless nights one would have to have such a large amount of cash / short dated bonds that you would have missed out on much of the upside into the 1980’s and the portfolio would have been ravaged by inflation – deeply sub-optimal! No silver bullet – index linked annuity anyone?!?
Andy J – Nothing particularly wrong with PAT for a particular type of person – I also agree with ZX – feels easy to replicate without the fees.
Something I read on the ERN site related to not holding bonds while also having mortgages based on current returns/interest costs (deaccumulation phase in the article) got me thinking should I apply the same logic during the accumulation phase? I currently have 65% global equity, 30% BTL, 5% PTP and 15% net worth in mortgages (home loan included) – 5 years off earliest possible retirement. I’ve been paying down the mortgage with any new cash for a little while now rather than hold bonds. I’m still wondering if there is a case for bonds in this situation – the ‘flight to quality’ bounce that often but not always happens? Grateful for any thoughts…
@TI. Cognitive dissonance on bonds is a great constant of passive investing.
To be fair, as someone who is a typically been a bond bull, I’ve never owned so few bonds. Partially that’s taking an active view. Partially it’s that as someone who is “wealthy” (vs. two decades ago when I was “poor”), I now need to hedge the risk of falling discount factors (rising yields) that might negatively impact asset prices. That requires me to be short bonds but I’m not that brave, so close to zero is the compromise.
Nonetheless, if you actually look at the US yield curve it hides an important fact. Take the 10-year US yield and decompose it into two parts, a 3-year rate and a 7-year rate, starting in 3-years (you can decompound any spot rate like 10-years into a series of forward starting rates). The 3-year rate is 0.4% compared to 1.6% in at the start of 2020, before the crisis hit, and 3% in late 2018 at the top of the Fed tightening cycle. So munch lower because Fed policy rates are at zero.
By comparison, the 7-year, 3 years forward is currently 2.1%, compared to the start of 2020 when it was 1.9%, and late 2018 when it was around 3%. So that rate is already higher than it was in early 2020. It’s fully priced in the normalization in the economy back to pre-COVID levels.
It’s still 90bp of the recent top seen in 2018 when the economy was on the “Trump pump”. If it goes back to 3%, the 10-year yield goes up another 65bp or so (7/10*90bp) to 2.3% say. That’s another 6% or so in negative returns (a few days in NASDAQ terms or perhaps even an hour or so in Bitcoin).
To think yields go higher than that, you’d have to make a call that the economy is doing better than any time in the last decade, that inflation will be persistently higher or real yields higher (or some combo). That’s a perfectly legitimate call but it’s an active view and surely would have no place in a passive portfolio.
For readers intrigued by the talk of US treasuries as an alternative to gilts, take a look at this piece:
https://monevator.com/do-us-treasury-bonds-protect-uk-investors-better-than-gilts/
US treasuries can work well in a crisis but can also backfire. From a passive investing perspective, it also means betting on the currency markets on top of everything else.
@ Whettam – intermediate gilts were up near 9% in 2020. But as TI says, one year’s performance doesn’t matter. Strategic asset allocation means each asset has a role to play beyond this year’s return:
Equities for long-term growth.
High quality conventional government bonds for recession defence.
Index-linked government bonds for rampant inflation defence.
Cash for liquidity.
Gold because it can work when nothing else does.
@ Naeclue – thanks for spotting that on the SPDR ETF. You’re right, I should have used the YTM. Fixed now.
Great article. Thanks @Accumulator
Still get so confused with Bonds, that for my simplified allocation pie-chart, I’ve resorted to lumping together with Cash and labelling it ‘Fiat’.
With regards to PNL, I can’t find another ‘all-in-one’ fund / IT that has a significant allocation to Gold.. and also they have their own bullion, rather than in an ETF. Plus I get the impression that they would be better at preparing / reacting to a financial shock than I would be.
@Jonathon B, cash can sometimes work out better than intermediate bonds. In the 1968 scenario, here are the maximum SWRs for a 30 year drawdown and various equity:bond allocations. Bonds in the first column, cash in the second. RPI is used as for inflation:
100:0 3.66% 3.66%
90:10 3.82% 3.90%
80:20 3.95% 4.09%
70:30 4.03% 4.21%
60:40 4.07% 4.28%
50:50 4.07% 4.28%
40:60 4.03% 4.22%
30:70 3.95% 4.09%
20:80 3.82% 3.90%
10:90 3.65% 3.66%
0:100 3.44% 3.35%
Discovering this result was one of the reasons I opted for cash deposits over gilts and maybe the reason some experts recommend short dated bonds over intermediate or long bonds. If I went back to 60:40 I think I would probably go for half in cash, half in intermediate gilts, as @TI has said.
As you can see, anything between 80:20 and 30:70 was respectable. To see the tradeoff between comfort and return, here is the size of the pot (number of years remaining) at the end of 1974 and 1998 for a 3.5% SWR, cash used for safe asset:
100:0 6.8 8.0
90:10 8.1 18.1
80:20 9.4 23.8
70:30 10.8 25.8
60:40 12.3 25.0
50:50 13.8 22.2
40:60 15.3 18.0
30:70 16.9 13.1
20:80 18.5 7.8
10:90 20.0 2.7
0:100 21.6 -2.1
All but 100% cash worked out, but the more bonds/cash the easier it would have been to sleep at nights in 1974, but after 70:30, the lower the residual pot. As you go to lower SWRs, the better off you eventually are with less in cash/bonds, but you are still get a more comfortable ride with bonds. Here are the figures with a 2% SWR:
100:0 13.5 144.5
90:10 16.0 149.9
80:20 18.6 147.9
70:30 21.3 140.3
60:40 24.1 128.4
50:50 27.0 113.8
40:60 30.0 97.6
30:70 33.0 80.9
20:80 36.0 64.4
10:90 39.0 48.9
0:100 42.0 34.8
When people say DB pensions or annuities are “bond like” I think it is mainly this aspect of bonds to which they are referring. Having a higher allocation of bonds leads to a more comfortable ride in extreme market conditions. Otherwise DB pensions are very unlike bonds. DB pensions cannot run out as bonds can, but neither can they be rebalanced with equities and there is no option to spend more of the DB pension when stock markets are down. If you can use a DB pension to reduce the SWR on your equity/bond portfolio, then that does mean you can more comfortably take a higher equity allocation, so in that respect a DB pension or annuity might be thought of as a bond substitute.
@ Naeclue – that’s great stuff. Thank you for sharing. Would you be prepared to share your spreadsheet? There’s so little in the public domain using UK returns.
1972-74 is a brilliant example of when neither equities nor bonds work during a recession. Cash was hurt too but in comparison it looked amazing.
My conclusion was:
Make cash part of my asset allocation
Add 5-10% gold
Index-linked bonds would have been a lifesaver in this situation
Because I don’t know whether the next ‘big one’ will be like 1972-74 or the Great Depression, I don’t write off any of the major asset classes. They can all ride to the rescue therefore I hold them all.
That level of diversification naturally means you’re holding less in conventional bonds too.
Re: optimal allocations. I don’t think there is one. There’s a great table in this piece by Wade Pfau that shows how the equity allocation can vary between 33% and 72% and still lie within 0.1% of the best SWR for a 40-year period with a 90% success rate. See table 3.
https://www.financialplanningassociation.org/article/journal/JAN12-capital-market-expectations-asset-allocation-and-safe-withdrawal-rates
@ZX, @TA, @TI – thanks for great responses – I think this discussion is hugely helpful to play out on here and is why we love the site. So much experience and thoughtful insight. VLS20 type thing plus cash is cheaper and working for me to achieve a level of diversification from my world equities tracker.
@TA Apologies I worded that badly I should have said UK Gilt ETFs (e.g. IGLT, GLTY) are down approx. 8% over last 12 months, not last year and not intermediate. I agree with you and @TI that one years returns are irrelevant, I only made the comment as an observation that its outside the normal “distributed between -6% to 6%”.
As I have said before, although I love Monevator and both your articles, I’m not totally passive, hence my decision to hold cash + short dated gilts currently. I accept that I might regret that decision at some point.
Accumulator-that Table 3 set of Asset Allocations was noticed by me rather late in my investing career
Fortuitously I had arrived at a 30/65/5 equities/bonds/cash in my old age via other paths and this table just confirmed my current stance
It is a great source of comfort to a investor that within reason his/hers asset allocation is not critical ie 70/30 to 30/70 and everything in between will do the job
Staying the course and watching costs are more important and under the investors direct control
xxd09
@TA, these drawdown backtests were done using MATLAB rather than a spreadsheet. Unfortuneately I cannot share or point you towards the data as it comes from a private database. The database owners would have to make a request to their feed and data provider for me to put the data into the public domain and they would almost certainly refuse (the providers in turn license the data from FTSE Russell, etc.). The lack of public domain UK data is incredibly frustrating.
Agree with you about the impossibility of coming up with an optimal allocation without the benefit of hindsight, but it is interesting and heartening how a wide variety of diversified portfolios can all end up having roughly the same maximum SWR in the toughest of market conditions.
Apart from the unlucky few, the prudent SWRs that people start off with should be successful, with declining SWRs as the portflios outperform the worst case scenarios that the portfolios were designed to cope with. That success then allows choices to be made between taking more income, going for more safety/lower risk, or going for more growth. xxd09 has used the flexibility provided by good returns to go for increased safety, whereas I have gone the other way for more growth.
A 1970s style bloodbath in bonds and equities would hurt us both, but both portfolios would survive as the SWRs are low. The difference is that I would experience more far more pain.
@Naeclue, many thanks for your incredibly detailed and illuminating response. So many of my questions are answered in your response!
You have a great way of seeing and analysing things. I hope you are one of the brave volunteers who will in due course contribute to Monevator; there must be many people who don’t routinely trawl through the comments who would benefit from your insight.
@Naeclue:
Fascinating data and analysis.
Your point about data [availability] is well made and was echoed loud and clear in an interview I recently heard with Jacob Lund Fisker of Early Retirement Extreme.
A few Q’s follow:
I think your equities is the FTSE all share, in which case:
Q1: is it not clear that a higher SWR should be achievable by globally diversifying; and
Q2: have you seen this post https://www.retirementace.co.uk/2021/04/even-nugget-of-gold-can-dramatically.html which IMO (when compared with say ERN’s SWR series part 34 work) seems to suggest that the addition of gold has a larger impact (a bigger SWR uplift if you prefer) on a UK portfolio than a US (and possibly therefore global) portfolio – I am not sure if this is real or just an artefact of the method used; what do you think?
Q3 your point about choices/options for the majority after a period of initial success (#54) is captured in somewhat gory technicolour detail by ERN at part 38 of his SWR series – when can we stop worrying about sequence risk – noting his observation that “In retirement, you’ll always face risks. The biggest concern for me personally and a lot of other retirees I know would be spending uncertainty because we’ll never know our old-age expenses, especially health and nursing home expenses until we actually get there.” and his overall pithy summary “Sorry: Sequence. Risk. Will. Not. Go. Away!”
Well…. I still don’t really get bonds at the moment. If I follow the advice to buy Gilts to match my ‘investment time horizon’ that implies holding to maturity, probably. In which case if gilts go up when equities crash – who cares? I know exactly what I’m getting back, and when. It’s completely decoupled from the equities part of my portfolio and hasn’t provided any real benefit in a crash, unless I need to sell everything in a hurry – not a good thing to do ..
If the portfolio is rebalanced occasionally it’s different
@ Al Cam – that’s an interesting piece. IMO bases it’s findings on a small sample size and overstates the case:
Amazing performance of gold in early 70s is connected to the end of Bretton Woods system and legalisation of private gold holdings. The article acknowledges this but then goes on to include gold’s performance during 73-74.
Doesn’t show bear markets and corrections where gold’s performance was less rosy.
Doesn’t mention the -80% drawdown lasting 30 years from 1980 to 2010.
This piece is really good and offers a more complete picture:
https://occaminvesting.co.uk/should-you-invest-in-gold-as-a-uk-investor/
The two articles numbers are quite different for GBP gold returns during the GFC and dotcom bust so go careful.
ERN’s piece on gold and SWRs effectively concluded with a no-score draw on gold. US returns.
I’ve checked out gold in Timeline before and not seen a dramatic difference. I need to take another look at that and write a post. We could double-check it in Portfolio Charts too.
There’s certainly a case for 5-10% gold or even a bit more. Uncorrelated asset, works when other assets don’t (sometimes). It’s not a slam-dunk though. Could just be a drag on returns for years. Oh, damn you risk!
@ Naeclue – ah, I see – well thank you for sharing that snippet all the same – it’s really interesting. I’ll see if I can take a bash at it using Timeline.
Agreed that a low SWR is the best strategy of all, with the option of pressing on the gas later.
I hope we don’t see anything like the 70s again but at least it prompted the launch of linkers.
@ Al Cam, TA – The protective and return boost effect of gold in a portfolio is well explained by some simple maths. It’s what Bernstein calls the rebalance bonus. The more volatile and the less correlated an asset is to the rest of the portfolio, the larger is the rebalance bonus. Gold is more volatile than equities and is very weakly correlated to both equities and conventional high quality government bonds.
To get the rebalance bonus it is, of course, very important to rebalance the portfolio regularly. Here is the link to Bernstein’s paper:
http://www.efficientfrontier.com/ef/996/rebal.htm
@TA (#58); T-B Dr Who (#60):
I can just about buy Au (or other precious metals?) being protective in awful scenarios – but it is often a drag not a booster.
IMO an awful lot of SWR-type results are down to the modelling assumptions/approach/data, etc. Hence my Q2 to Naeclue above. We have discussed this topic here at Monevator before – see e.g chatter to TA’s post about back up plans for living off a portfolio and I always remember this paper too: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2548651
@ Al Cam – The main point in Bernstein’s paper is that having something very volatile and weakly correlated boosts your return when you rebalance. He looks in detail into the effect of gold miner stocks. The correlation is not as low as gold itself but the volatility is significantly higher. He finds a sizable increase in return but stresses that most people would find very hard to keep a portfolio like this and rebalance in the middle of a crisis.
@Berkshire Pat (#57):
IMO you are correct if you hold [index linked] bonds to maturity then -provided they are not callable – you know the outcome in advance and a lot of the chatter here is not applicable.
However, it is my understanding that building a ladder of such bonds by an individual (to pay your Floor income annually/semi-annually if you like) is not easily achievable/cost-effective in the UK but is (was?) in the US. This approach is completely feasible and cost-effective for DB pension funds in the UK.
In general, there are many analogies that can be drawn between DB funds and personal retirement planning but a couple of things are inescapable for the individual, namely scale (often many tens of thousands of scemem members vs one) and probabilities (meaningful for a DB fund vs meaningless for an individual – in so far as you are either alive or dead).
I also agree with your re-balancing comment, hence my post at #38 above re Floor and Upside (F&U).
In de-accumulation F&U and SWR approach re-balancing rather differently. Simply put, in F&U – once in de-accumulation – you never re-balance away from the Floor!
People appear to react to loss differently with the stock market than they would with gambling – if you lose in a slot machine you get the false feeling that “this machine will pay out soon because it owes me a win” or “I must bet on another horse to make up my losses” – whereas with investing – the thing that really does store up a reversion to mean, people have less faith
@ Berkshire Pat – when equities are down in a crash and gilts are up – this is the classic ‘buy low, sell high’ scenario everyone dreams about. Primetime for rebalancing.
In retirement, your time horizon is your death / your partner’s death / never because you wish to provide a legacy. In this situation, your bond holdings are a bond fund whether you hold them as individual securities or not.
“if gilts go up when equities crash – who cares?” – everyone who would have panicked if they couldn’t see at least one asset class capable of paying their bills during a major downturn.
Panic is real. I know people who’ve done it.
This is a general comment – take a 60:40 portfolio. Divide the defensive allocation like this:
Conventional government bonds – 10%
Cash – 10%
Gold – 10%
Index-linked gov bonds – 10%
How nuts is that? Does it overcommit to a ‘doomed’ asset class? Does it rely on a particular prediction coming true? Or does it use different diversifiers that have proven useful against a range of downturns from depression to stagflation?
@Al Cam, Q2 World index better than FTSE all-share?. Beware of recency bias! In 1968 the outcome would have been worse.
Here are the max SWRs, compared with the FTSE (cash as safe asset):
100:0 3.26% 3.66%
90:10 3.36% 3.90%
80:20 3.45% 4.09%
70:30 3.51% 4.21%
60:40 3.56% 4.28%
50:50 3.58% 4.28%
40:60 3.58% 4.22%
30:70 3.56% 4.09%
20:80 3.51% 3.90%
10:90 3.45% 3.66%
0:100 3.35% 3.35%
The first column is for the MSCI World index (converted to pounds in case anyone asks). That only started in 1969, so I have cheated and used the FTSE for the first year. What improved a little was the large meltdown in 1974. For a 50:50 portfolio in the FTSE allshare, there were 13.8 years of income left, but with MSCI World there were 16.0. The sequence of returns matters, but so to does long term performance. The FTSE allshare produced a real return of 7.26% from the end of 1969 to 1998, but the MSCI World only managed 5.40%.
Regarding the rebalancing bonus, it is well worthwhile in the worse case scenarios of drawdown. From the numbers in a previous comment, here are the remaining pot sizes drawing down at 3.5% from 1968-1998 for various asset allocations:
100:0 8.0
90:10 18.1
80:20 23.8
70:30 25.8
60:40 25.0
50:50 22.2
40:60 18.0
30:70 13.1
20:80 7.8
10:90 2.7
0:100 -2.1
In the 50:50 case, rebalanced, there would have had 22.2 years remaining, but without rebalancing and instead drawing 1.75% from each asset class there would only have been the average of the 100:0 and 0:100 pots left. That works out at about 3 years.
Rebalancing does not have to be done symmetrically and the McClung approach of only rebalancing from risk assets into bonds can work as well as conventional balancing. McClung would say it works better, but I am not convinced there is enough data to show that one way or the other. I prefer McClung’s technique though as I just think it is simpler and avoids the psychologically challenging business of buying after a market crash.
PS I would not want to give the impression that the FTSE allshare is a better index to invest in just because it performed better in the 1968-1998 case. That really does not make any sense and I would still go for global diversification. Some people like some home bias. I don’t, but I would not condemn it and if someone wanted to simply drawdown from the daily rebalanced 60:40 LifeStrategy with its bonds and UK bias I would not condemn that either. It is likely to give a very satisfactory outcome.
If you believe in the passive investing theory you just keep your equity and bond allocation and ignore the news. Doing otherwise is suggestive you do not believe in your theory.
The UK did better than the US for the ’67-68 cohort that famously tripped the 4% rule in the US. So I’d guess that could be one of the factors dragging down the MSCI World in this case.
Across the broader historical record, the US wins but the UK beats the Rest of the World because the ROW includes Germany, Austria, Japan and other countries that were hammered in WWI or WWII or both.
Would that make me think I should diversify less in the future? Absolutely no-way. It only shows what a crapshoot the historical record is.
@Naeclue:
Thanks, I trust you meant my Q1.
More fascinating numbers!
I am intrigued how exactly this data and analysis maps to this set of international SWR’s published by Pfau (which includes a global entry):
https://retirementresearcher.com/4-rule-work-around-world/
I think a key clue is that according to the Pfau paper the ‘killer’ year(s) all somewhat pre-date your sample – 1937 being quite common. Also the difference between global vs UK – whilst admittedly in the noise – is the reverse of your results.
As I have said many times before, I am no fan of the SWR approach – and am not, and probably never would – use it. Primarily because I am fortunate enough to have a DB pension. However, I do still enjoy its attendant mental gymnastics!
In that spirit I think McClungs re-balancing approach is similar to that used by F&U in de-accumulation, but I still cannot see how McClung can ever be compatible with Vanguard Lifestyle funds.
@TA, The returns of the 70s were much worse for MSCI World than for FTSE all-share. -4.2% real vs. -1.8%. Returns during the 80s were closer and very high at 15.4% real vs. 15.2%. This meant that for 60:40 portfolios with SWR 3.5%, The FTSE portfolio had 14 years remaining at the end of 79, but the MSCI portfolio only had 10. That made the SWRs 7.1% for the FTSE, but 10.0% for the MSCI. The slightly higher returns in the 80s for the MSCI could not compensate for this and by the end of the 80s there was only 9 years left in the MSCI compared to 20 in the FTSE. The damage from the 70s essentially sealed in the lower maximum SWRs for the MSCI.
@TA, regarding this:
Conventional government bonds – 10%
Cash – 10%
Gold – 10%
Index-linked gov bonds – 10%
Cash and bonds held to maturity produce positive nominal returns (until recently anyway!). As you indicated previously, gold can go nowhere for decades (1980-2005), or have almost bitcoin-like volatility as it has over the last 14 years, so I don’t really think it belongs in with “defensives”. If I held it I think I would put it in with the risk assets.
@SemiPassive (22) – sorry I’m a bit behind with these comments. Wade Pfau wrote a paper some years ago “An Efficient Frontier for Retirement Income” comparing equity:bond portfolios with equity:annuity portfolios. Searching for this exact title will find it. With the assumptions he used, and of course with rates available in the US at that time, he concluded that equity:annuity was more efficient.
@Al Cam, “I think a key clue is that according to the Pfau paper the ‘killer’ year(s) all somewhat pre-date your sample – 1937 being quite common.”. I would agree with that and would go with the World Index. It is more diversified, and historically has lower volatility. Probably better, but probably is not the same as always for all retirement periods, as I have illustrated for the period starting in 1968.
Another point about the World result in Pfau’s document is that I strongly suspect that his testing was done with a World portfolio drawn in US dollars. The result may be better or worse when converted to pounds. I tested using a World Index in pounds.
I doubt many people actually use SWR. It is just a useful calculational tool.
I am sure that most people would naturally reduce spending if returns are poor and spend more if available. Since we are planning on avoiding bad outcomes instead of focusing on the average, it is more likely that retirees will have the opportunity to spend more as time passes. They may not necessarily have worked out in advance a systematic way of varying the amout they draw, but they will still do it.
The need and desire to spend can fluctuate quite a bit in retirement in any case and nobody needs to spend the exact amount as dictated by SWR even if they were religiously drawing that amount. Stuff happens, sometimes good money saving stuff, other times very bad and expensive stuff.
@Naeclue (#75):
Thanks for your helpful thoughts; no major disagreements from me.
The £ vs $ is an interesting angle – over my lifetime, to date at least, the £ has been a relatively poor store of value.
My own view is that – bar a catastrophe – folks who are prepared, before pulling the plug, to do the most basic of sensible calcs will be fine in retirement; and a lot will end up with much more than they ever thought they would. Beyond basic calcs it is each to their own and, at best, the resulting answers are a guide for the next few years and beyond that timescale are relatively meaningless.
One of the problems with an SWR like approach is that on average it means a rising income during retirement because the planning has to cope with below average bad outcomes which, on average, don’t happen. Whereas most people would like the ability to spend more earlier on in retirement, perhaps with a bump up at the end to cover care costs. McClung and others try to deal with this, but it is a difficult circle to square.
Might be different for US investors who have to buy monotonically increasing health insurance.
@Tom-Baker. Gold is best thought of as a currency, not an asset. We define currencies in terms of bilaterals, GBP/USD, EUR/USD etc. So Gold is just XAU/USD or XAU/GBP. When we get a risk-off period or a period of high inflation, people say that the value of Gold is rising. It’s more likely that the value of USD or GBP is falling but we never really know. It’s just a bilateral.
Gold does what most other developed market currencies do – they are volatile short term, mean revert over the long-term, and have low pairwise correlations to assets. So for buy-and-hold investors, currencies are just uncompensated volatility and should always be hedged. Gold has no real place in their portfolio.
For those, however, that are willing to engage in rebalancing (an active concave strategy) a high volatility currency with low pair-wise covariance, that mean reverts over the longer term is pretty much ideal. But you have to be willing to rebalance frequently to capture that. You also have to have something else in the portfolio that is convex or you suffer in fast or persistent trending moves with high correlation.
@TA. The justification for those that don’t like owning bonds is easy. Have some bonds, not because you like them. Have bonds because you don’t like them but you might just be wrong … again.
@Naeclue (#77) I agree the spend profile of most retiree’s is that they wish to spend most early, then settle into more sustainable spending patterns. I’ve long thought a straightforward solution for this would be to simply budget a higher cash portion of the portfolio with the intention of spending this down over the first few year of retirement. Then settling back into the longer term portfolio. Or perhaps if you can’t wait, increase spending for the last few years of work and get the splurge out of the way prior to retirement?
@ ZXSpectrum48k – Thanks for another insightful comment. Thinking of Gold as currency never occurred to me previously but makes total sense now you’ve pointed that out.
I suppose for those employing a rebalancing/passive approach, and looking for a simple allocation, TA’s allocation (as below) seems to tick all the boxes?
Equities – 60%
Conventional government bonds – 10%
Cash – 10%
Gold – 10%
Index-linked gov bonds – 10%
@Rosario:
Re front loaded spend profile:
Your extra cash suggestion should work.
Your idea (albeit called a “travel bucket” in this document) and some other ideas are discussed in Section 6 of this Report from Stanford, authored by Pfau, et al:
https://longevity.stanford.edu/wp-content/uploads/2019/07/Viability%20SSiRS%20Final%20SCL.pdf
@The Accumulator #65
Investment time horizon is a bit vague -I’m thinking about a portfolio to provide for retirement. The point was -if one buys a bundle of Gilts to mature at different points during retirement to provide a foundation income, it doesn’t matter what wobbles they undergo during market crashes – you’ll get exactly the same income, and exactly the same cash at maturity. Unless you re-balance it doesn’t matter if they are at 140 or 80 at any point – they will pull back to par and the coupons will roll in just the same. Take your point re the psychological aspect – but cash savings in FSCS protected accounts would have the same effect
@Berkshire Pat (#81):
FWIW, I think it is conceptually easier to use “bullets” to build such a foundation income – this approach eliminates the intermediate pay outs. IIRC, stripped versions of Gilts do exist – but they are relatively rare and there are substantial gaps in available maturity dates. Have you looked into the feasibility and costs of doing this – see #63 above.
@ BerkshirePat – sure, a bond ladder that you hold to maturity – you know how much cash you’ll have in nominal terms per year for as long as it lasts. You don’t know how much it’ll buy in the shops years from now after inflation – unless it’s a linker ladder.
Imagine it like cash. If you put £10K in a two-year savings bond today at 1% then you know exactly what you’re going to get back. But if another bank offers a two-year savings bond tomorrow at 2%, then the opportunity cost to you is that lost interest. That is a real loss.
Is rebalancing not part of your plan?
Cash – yes, I would take comfort from cash savings but it’s not the same. Cash doesn’t have the same capacity for gain in a recession. One way to think about cash is as a very short term bond. Why not hold both?
@ Rosario – everything I’ve read on the topic suggests the early splurge increases the chances of a cut-back later (I’m talking about someone who is decumulating rather than your spending spree before ending work scenario).
Which is fine if someone decides it’s worth it, believes they’ll spend less later in life, has health issues that may curtail retirement, has the flexibility to spend less later on and not mind… which is my roundabout way of agreeing with you… I’ve built a separate treasure chest for a wee splurge myself.
@ ZXSpectrum48k 🙂
In practice……..
I used a rolling bond ladder of 5 year gilts for some years but eventually got tired of the mechanics
Plus I wanted exposure to the US bond market
I opted for a bond fund (VIGBBD)
Simple cheap and easy to understand and no work needed from me!
Last fact more important as you get older
xxd09
@ xxd09 – agreed, convenience is a godsend in investing – it reduces opportunities for shooting one’s own feet off. Setting the right course then putting on the auto-pilot is a super-power.
@TA: (#83)
Re: Is rebalancing not part of your plan?
Once in de-accumulation, if you think about such a bond ladder as being a Floor, then the money for this is “locked in” (albeit reversible – with possibly some costs/delays) so should not be available (for want of a better phrase) to rebalance with. Functionally in a Floor and Upside (F&U) approach you should never rebalance to take money away from the floor; you may, of course, add flooring – i.e. re-balancing to add more Floor is OK. In F&U parlance this is known as “raising the floor” as opposed to lengthening (in time) the floor – which can be a dangerous approach, as it means at the outset of de-accumulation there were holes in the floor!
In the F&U approach you probably also hold an Upside Pot and that should be re-balanced as you would in accumulation to maintain risk profile, etc. Hope this clarifies things for you?
Re your comments about inflation – it is perfectly feasible to take a view on what future inflation might be and build that into your nominal flooring allocation. Personally, I found this quite easy to implement for a relatively short duration ladder. And, as xxd09 notes, the mechanics of ladders really are a bit of a pain in the axxx!
@ZXSpectrum48k – Thanks for your insights.
I also like to think of gold as a currency that Central Banks can’t interfere with. It’s a very good approximation most of the time, except when something extraordinary happens (e.g., last year when the pandemic reduced air travel affecting the physical transfer of gold across the pond and the gold price in London deviated signicantly from that in New York).
Your mention of having something convex in the portfolio reminded me of Chris Cole’s Dragon Portfolio (https://www.bnnbloomberg.ca/how-to-build-a-portfolio-that-outperforms-for-a-century-1.1595829). This is one of the main differences between their attempt to develop a ‘permanent portfolio’ and the original Harry Browne or even the Golden Butterfly. What do you think about the Dragon Portfolio?
@Berkshire Pat, @Al Cam, It is reasonably straightforward to build a bond ladder to match cashflows if that is what you want. The trick is to start at the long end and work backwards. For example, let’s say you want £10,000 on the 1 Jan every year up until 2031. Starting with 1/1/31, the closest gilt is Treasury 4.75% 2030, maturing on 7/12/30. This will pay interest of 2.375% on 7/6/30 and at maturity, so you only need £10000/(1+0.0475) = £9,546 nominal. For 1/1/30 the best gilt is Treasury 0.875% maturing 22/10/29. You will be getting interest on the 2030 gilt during the year totaling 9546*0.0475= £453, so you only need £9,565. As before, that means you need to buy 9565/1.00875 = £9,482 nominal. And so up down the ladder. It can get fiddly as sometimes there are wide gaps between maturities, so best to do it all on a spreadsheet.
Once you have built the ladder you might want to look at fixed rate deposits for some of the rungs instead, provided you don’t mind being locked in. For example, Hodge Bank are offering a 1.35% fixed rate deposit. In total £5,000 in that will get you £5,346 in 5 years time, with interest reinvested. The closest gilt maturing before 14/5/26 pays 0.125% and matures on 30/1/26. I just got a quote for £5,000 of that on Hargreaves Lansdowne and it works out at £5,046 nominal (after accrued interest and HL’s brokerage). The return on that would only be £5046 capital plus £32 interest, total £5,078 (assuming no interest on interest payments). Little point in holding that gilt IMHO, unless you value the liquidity.
In fact even with Treasury 4.75% 2030, £5000 will get you only £3624 nominal which will pay £1,721 interest over 10 years for a total of £5,345, a pound less than the Hodge Bank 5 year deposit. Again though that ignores any interest on the coupon payments. So it is questionable whether it is worthwhile retail investors building a cashflow matching gilt ladder at all at the moment with any rungs under than 10 years.
Our cash is a mixture of instant access and fixed rates, but no gilts for this very reason.
An article on the mechanics of bond ladders would be interesting, to at least understand the complexities and whether it’s a viable option.
I think the idea of holding bonds for their full duration is interesting. My simple approach has been to hold cash and short terms bonds for the defensive part of my portfolio, because I intend to retire in approx. 4 – 5 years and will want the TFC.
I’ll look into it myself when I get a bit nearer, if I can work out the the DMO website, but any guidance / links appreciated.
@Whettam (89) Two articles from Wade Pfau on building a TIPS ladder may help. As has been commented by others, building a similar ladder using UK Inflation-Linked Gilts is not really practical because of the lack of a suitable range of maturities. Nevertheless, the mechanics of building a conventional Gilts ladder should be broadly similar.
https://retirementresearcher.com/building-retirement-income-tips-ladder/
https://retirementresearcher.com/how-do-i-build-a-tips-bond-ladder-for-retirement-income/
@Naeclue (#88); @Whettam (#89)
Good explanation, and agree entirely with your conclusions.
May be worth adding that if you use linkers it is a real cash flow and if not it is probably worth taking a punt on what inflation might do to your cash flow needs.
IMO, fixed rate deposits (FRD) (vs Gilts, etc) are – and have been for some years – the way to go for durations of up to ten years for the reasons you give plus FSCS protection too.
IMO bullets simplify the process as the eliminate all the pesky coupons – but they are a bit like hens teeth!
Have you ever considered using commercial bond alternatives to gilts?
I was not aware that you could acquire gilts via HL – is this new? I never found the DMO terribly helpful. They are not really there to serve individual punters.
I have been using this logic and conclusions for a while now to bridge the Gap until I turn on my DB.
One practical issue is that I have never found any FRD longer than 5 years. I have effectively used a sort of stack and roll approach – and therefore had to estimate future interest rates too.
Reversibility options are rarer now too.
Also, you should be prepared to be pretty flexible around maturity dates – and whilst working an annual cashflow is doable, anything else (like monthly) is just far too much effort IMO.
Keeping cash in instant access a/c’s helps smooth things and is also at hand to deal with any emergencies.
To date, I have been fortunate for various reasons including:
a) I over estimated inflation;
b) I under-estimated interest; and
c) plans change and my Gap will now be shorter than originally foreseen, courtesy of a number of things – not least of which is the last budget.
My conclusion is that this is the one area where a pragmatic approximation to a ladder can serve you well – but do not get too hung up on the mechanics.
One last thought for Whettam – the longer you leave it the shorter the duration [of bonds] you will require. As mentioned above, currently around ten years duration is the cut-over from treasuries to FRD’s. He who buys the longest earliest often gets the best deal in this game!
@Al Cam, HL have offered gilts for many years. What is relatively new (4 years?) is the ability to buy them online. Previously you had to deal by phone and pay the higher dealing fee, although they often let me buy for the online fee as I complained about the lack of an online option every time I traded. They offer corporate bonds as well, many online, but I prefer not to mix credit risk into my bonds (that’s what equities are for) and if I did I would use a fund/ETF.
Another thing to be wary of with bond/deposit ladders is taxation. The best fixed deposit rates are rarely available with cash ISAs and even at 20% tax this is an important consideration for investors who have used up their annual interest allowance. At least gilts can be placed in ISAs and SIPPs and that might tip the balance towards gilts in some instances for those who don’t mind using their SIPPs/ISAs for gilts.
Holding gilts outside ISAs/SIPPs can be really tax inefficient. In the example I gave earlier, £5000 of the Treasury 4.75% 2030 would pay £1,721 in coupon payments if held to maturity. Ignoring accrued interest tax relief, that’s £344 at 20% tax. That pretty much makes the net nominal return on that gilt 0% if held to maturity and 20% tax is paid on the interest. That compares to just £69 tax on the £346 of interest from the Hodge Bank 5 year deposit. Any gilts bought above par value will suffer from “unfair” amounts of income tax like this. There are a few gilts though that are trading below par and they are tax efficient as there is no CGT to pay on gilts.
@Al Cam, you can occasionally get longer fixed rate deposits. Shawbrook have a 7 year one at the moment. They are not very popular as most people (me included) don’t want to tie their money up for that long.
https://www.shawbrook.co.uk/direct/savings/personal-savings/fixed-term-accounts/
Not that long (OK so it probably longer than I remember) I had a 5 year fixed rate cash ISA expire at a rate of 5%, wish they were still available!
@Tom-Baker. Re: Dragon portfolio. I’d agree that there have been clear economic regimes (typically lasting 20-40 years). The most recent regimes of Secular Stagnation (1964-1983) and Secular Boom (1984-2007) and again Secular Stagnation (2008-2020) have been generally good for the simple Modern Portfolio Theory type bond/equity mix but too to assume that continues looks hard at current levels. In that sense recency bias is a risk.
I’d also agree you want more diversification than MPT provides. But is the solution adding in long volatility, CTA trend-following and Gold as the Dragon Portfolio does? They seem to help but the Dragon Portfolio is subject to the same risks of MPT, just with a longer lookback period. It’s still trying to find that magic portfolio that is both “economically rational” but also performs over a specific historical period. It’s vulnerable to those biases.
My biggest single investment has taken the idea of mass diversification to extremes. Diversification over 200+ independent investment strategies (teams of portfolio managers), not just a few asset classes or countries. A vast array of products. Time-horizon is a key form of diversification with everything from HFT to buy-and-hold PE. Concave short-term mean-reversion strategies vs. convex option-like structures. The results are truly phenomenal and even somewhat scalable.
They are not, however, replicable for the average retail or even institutional investor. Evidence suggests it’s the exception, not the norm. It’s not even clear to me how the average retail investor gets good exposure to 2 of the Dragon strategies (long vol and CTA trend following). I’m convinced though they can easily get bad exposure to them.
@Naeclue (#92 & 93):
Your points re taxation, wrappers, and so-called phantom income are well made.
I think you have been more “aggressive” than me in pursuing good rates – I well remember a comment you made about changing for an uplift of 0.1%. I wouldn’t. Having said that, I have done OK and as long as I can achieve my objectives (which with hindsight I set rather low) I am happy. My best current FRD rate is 2.25% and it has c. 2 years to maturity and is reversible – albeit on payment of a penalty. Along the way, I did get access to a few ‘loyal customer’ deals – but these seem to have dried up now.
My remaining Gap is now somewhat less than 5 years – but thanks for the tip re Shawbrook; seven years locked up seems a bit of an ask to me too! But then again I initially avoided the longest term FRD’s at the start of my journey across the Gap as I just could not believe ZIRP could persist – got that well wrong too!
Also, what I do with cash once I cross the Gap and my DB comes on stream remains to be seen. I suspect I will gradually wind down any excess (to plan) cash to boost my Upside – and this is where the ‘convertibility’ of ISA’s can/could be handy again.
Then, there is also the issue of ‘optimizing’ the precise start date of my DB – but that is a whole other ball park!
Great chatting to you about this!
Bonds are actually “positive debt” so there’s no point holding gilts if you’re also in debt. Paying a mortgage rate @2% and holding gilts @0.5% is dumb!
So the ‘bond insurance’ argument is valid, but only if you’re totally debt free. But then you should be able to tolerate a bit more portfolio volatility, too.
Although I agree in general, for 2021 and maybe 2022, I think that investing in bonds is like having cash but with extra & possible high risk. They are low, even negative yielding, and if inflation takes off the price will tank and possibly even crash if there is a rush for the door, which could happen when a ‘safe’ assest fails to be actually safe.
@chrisssB – what about if you’re contributing to your pension?
There is an interesting article from Wade Pfau about using annuities as bond substitutes:- https://www.advisorperspectives.com/articles/2015/08/04/why-bond-funds-don-t-belong-in-retirement-portfolios
If you do the sums on a combination of fixed-rate annuities with inflation-linked income products such as the state pension and drawdown you achieve a higher income during the first decade or so of retirement and you have to live a pretty long time to suffer significantly from inflation.
@chrisssB – I think you’ve confirmed the sentiment of my previous post #44. I’ve been trying to max out the annual pension allowance and hold only a global equity tracker in the pension. This currently just about meets my target portfolio allocations. Any ‘spare’ taxed savings are now directed at paying down the mortgages before buying bonds (or anything else tbh). It has the potential to move the portfolio out of line over time (but having illiquid BTL property makes it harder to perfect anyhow). I guess I could buy bonds to balance, but it feels ‘dumb’ right now like you say. I can’t think of a better plan at the moment with the limited knowledge I have.
@Ade – yes we make the same point. There’s no better ‘insurance’ than not having debt to pay.
@Accumulator – are you suggesting that the basic maths in my comment of ‘pay 2%, earn 0.5%’ is altered by the tax wrapper of a pension? I don’t see it. A pension is, after all, just deferred tax until retirement. The old argument that a retiree’s tax rate will be lower than during his/her accumulation phase now looks a bit shaky (with tax rises likely).
@ chrisssB – tax relief and potentially employers match means it makes sense to contribute to a pension while paying a mortgage.
The reason to hold bonds in that pension is to ensure you don’t panic in a crisis.
That’s particularly important when you’ve built up a large pension portfolio.
If the market drops 50% then doesn’t recover for a decade you’ll be glad of the bonds.
You make a good point about paying down the mortgage outside of pension holdings. I agree with you and did the same thing myself.
@Accumulator – your logic is fair enough but in the real world I wonder what provides the best ‘sleep at night’ insurance in your 50% market-drop scenario?
A 50% drop probably means recession, depression, unemployment etc. and lots of bad news.
In that scenario: would I prefer a pension – that I can access in 15 years – having ‘only’ dropped 20% because of my higher bond weighting? Or that I’m totally debt / mortgage free and the roof over my head, and of my wife & kids, being bullet proof?
To me it’s a clear choice. Bonds have no place in the portfolio if you have debt. But I guess what ‘insurance’ we all have is a personal choice.
@ chrisssyB – 100% agree that the decision has to be calibrated to our personal circumstances i.e. our need, willingness and capacity for risk.
You’ve got me thinking now about how my risk profile changed over time:
During the Credit Crunch I didn’t care about my portfolio – it was weeny. My concern was losing my job. Like you, I prioritised paying down the mortgage for several years after.
When the near bear market of 2018 hit I felt a stab of discomfort – my portfolio was much larger by then but I was underweight bonds. Hat-tip to The Investor who prompted me to de-risk about a month before the downturn. (Mortgage sorted by this point).
I’d just reached FI when the corona crash hit. I was completely sanguine though as my strong bond allocation cushioned the blow. Noticeably as the market slid to 40% down I couldn’t bring my self to rebalance – a sign that I was nearing the limit of my risk tolerance.
@TA (#105):
You are/were probably experiencing two things, namely:
a) increasing financial capital (FC) – if you like the value of your portfolio; and, at the same time
b) decreasing human capital (HC).
A lot of people overlook the HC aspect.
Incidentally, this model also explains the recommendation that most people should hold the highest proportion of equities whilst young.
Are you familiar with the book: Are You a Stock or a Bond, by Milevsky and/or the 2007 paper: Lifetime Financial Advice: Human Capital, Assett Allocation, and Insurance, by Ibbotson, et al?
@ Al Cam – bang on. I’ve read various human capital theories and Milevsky is brilliant. Haven’t read that book though and can’t remember if I read that particular paper – much is buried beneath the silty mud of my mind.
@TA (#107):
The premise of that Milevsky book is that some jobs/professions are more risky than others – hence more equity-like than bond-like. I seem to recall at some point he contrasts a tenured economics professor with a stock trader.
In my experience the mud just seems to gets thicker!
Another classic paper from around the same time is The Theory of Life‐Cycle Saving and Investing, by Bodie, et al. This paper provides a framework to throw some light on personal finance perennials, such as: rent vs buy; clearing a mortgage early; how much to save; etc.
@TA:
FWIW, I like the treatment of “A lifetime of Risks” in Section 1 of this report
by Tom Idzorek: http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=CD2E1934221E785860D04F9F1706AD99?doi=10.1.1.183.5994&rep=rep1&type=pdf
Hey, thanks for that Al Cam. Perfect timing. I was just thinking about glidepaths today. I’ll take a look at that paper tomorrow. Cheers!
Great article thanks.
What are peoples thoughts on a flexible bond product like TwentyFour dynamic. Bond where (for a fee) you let hopefully experts navigate the short intermediate long bond options. Other strategic bonds as well i guess. Less interested in their return against say Gilts, but more whether they expect to adapt quick enough in a crisis to give the Longer Bond protection but at other times perform more like short bonds.
Any thoughts
MrBatch