This article can be read by selected Monevator members. Please see our membership plans and consider joining! Already a member? Sign in here.
What derisking your portfolio looks like [Members]
Next post: Surviving system meltdowns and cyber attacks
Previous post: Weekend reading: Monevator Christmas sweatshirt competition
Comments on this entry are closed.
Proofreading error: “There are also aren’t any outlier results”.
@Wodger — Cheers, fixed!
Yet more thought provoking analysis TA, thanks!
I’m struck by the split between inflationary and deflationary periods associated with the equity bear markets.
I’ve spent a reasonable amount of time and effort getting to grips with linkers and setting up a linker ladder. Now I’m thinking whether my average 2% real yield is still okay in outright deflation. So used to thinking in inflation adjusted terms and so used to inflation being positive I feel like I have a blind spot if nominal prices start falling.
I’ve not rationalised the cash allocation in this way until now. But now I’m thinking that in outright deflation holding some cash will buy more next year than it did this year.
This is probably why deflation is so insidious for the economy as a whole. If my cash buys more next year why don’t I wait before making that purchase until the price has come down.
@prospector #3 > So used to thinking in inflation adjusted terms and so used to inflation being positive I feel like I have a blind spot if nominal prices start falling.
Observe that the deflationary periods were when the world was on the gold standard. There are some that argue in a gold standard with the number of goods and services expanding in value and a constricted store of wealth tokens you will get deflation. My bar of gold today would buy twice as much in 10 years if there were 2x as much stuff for it to chase in 10 year’s time.
We aren’t in that Bretton Woods world any more. I am happy to bet that I will never see nominal deflation in the rest of my life, because governments have to overpromise and devalue the payback value of their promises by inflation or by cutting back.
There is a presumption in the analysis here that the economic environment is statistically stationary. I don’t think that we will see les trente glorieuses again, ’twas a different world. There is a hidden presumption that we have swept the entire problem space if you are going to infer the general from the particular. OTOH it is interesting that the Permanent Portfolio was not arrived at by backtesting but inferred from Harry Browne’s (possibly limited) survey of the economic situations he could trace historically.
With huge thanks to both @TA and @Alan Stocker. Amazing work. Unique, new insights and really valuable. Very much obliged to you both.
I don’t know what to make of this observation, but I’ll make it anyway:
The Permanent Portfolio is broadly 3 risk off and assets 1 risk on asset.
Maybe coincidentally, maybe not, this is also broadly the optimum mix for a (3x) leveraged portfolio using daily reset LETFs and a once every 19 months rebalance (back tested 1926-2022 : 27% Consumer Nondurables / Staples LETF, 27% Healthcare LETF, 26% Energy LETF, 10% ITT, and 10% Gold: see my comments as @Time like infinity, and in particular @Hydromol’s and @Logan Roy’s replies, from page 283, 26th Nov 2023 @9:31 am onwards, and from page 285, 18 Jan 2024 @7:01 pm onwards; on Bogleheads here:
https://www.bogleheads.org/forum/viewtopic.php?p=7565270#p7565270 )
Somewhat reassuring that what seems to ‘work’ unlevered also appears to ‘work’ (with suitable adaptations to keep the underlying very roughly 2 or 3 parts relatively risk off, to 1 part more risk on, asset mix principle, but instead using utilities and staples in place of bonds etc).
@TA – many thanks, this is all fascinating stuff! It’s interesting sometimes how little variation in returns results from each of the defensive asset classes. I have allocations to medium and short gilts as well as cash, not sure if that added complexity is necessary, though I’m now technically in deaccumulation. (Even if I haven’t actually drawn down yet).
The Permanent Portfolio seems to have been very effective in the past but I’m not sure I’d be wanting to initiate that strategy right now.
I do notice that the tables work best in landscape format on my phone, in portrait the numbers after the decimal points move down below the others. Apologies if that has been pointed out in previous posts, just noticed it today.
We experienced annual deflation from 1921 to 1924 inclusive while off the gold standard. Back on the gold standard in 1925 – 1931 (more deflation), off the gold standard in 1931 – followed by 2 more years of deflation until interest rates were cut.
Japan has notoriously suffered from annual deflation in recent decades as has Switzerland.
Like @Ermine I don’t think deflation is a major concern but it’s possible.
@Ermine – Naturally the past doesn’t contain all there is to know about the future. Would be interesting to debate what reasonable steps someone could take knowing that. I mean from a DIY investing perspective, not a prepper perspective.
Holding a global bond fund hedged to GBP seems like a sensible move. There was also a great convo on another thread – I can’t remember where – about Edward McQuarrie’s paper challenging the ‘Stocks for the Long Run’ thesis.
https://www.tandfonline.com/doi/full/10.1080/0015198X.2023.2268556
McQuarrie argues the equity premium can disappear. On that view, it’s always worth holding bonds and long-dated ones at that.
@Larsen – yes, I was surprised by how little difference the various fixed income indices made at the short end.
I currently hold cash and All Stocks (not including my linkers). Think I might swap out the All Stocks holding for a medium or short holding.
@TA I think the lower probability of deflation is one reasonable inference we can take from today’s highly interconnected world, the West in secular decline possibly culturally possibly due to ageing. There is high government debt. The world was much more compartmentalised when the deflations happened, acknowledged it wasn’t just the gold standard.
I am surprised the PP comes through this well, I personally am in favour, but I do think it’s an old man’s portfolio – one to retain your wealth, not build it, as @DH said, it’s 3x risk-off compared to risk-on. I personally don’t do bonds (I have a DB pension that performs some of the same role though not the swing producer element to networth.
It’s a tougher game to build wealth at today’s high valuations that after the GFC or the dotcom bust. There’s also value in looking at how the various asset classes respond to market stress. 3652 days has a good post on the much lower fuctional liquidity in the bond markets under stress. That’s probably at least worth noting and perhaps some reason why bonds seem to have failed people recently in their time of need.
If the conclusion is you need the PP to hedge serious times of hazard then that has serious adverse implications for building wealth, particularly if you change gear down halfway through a RE aspirant’s working life. The normal assumption that compounding roughly doubles your money over 40 years for an equity-heavy asset allocation just won’t hold if you have to change down to a mix that’s three-quarters risk off.
The long-term returns show the Permanent Portfolio growing at the rate of the 60/40 portfolio. That’s a respectable pace of accumulation and generally considered to be the default portfolio not an old man’s refuge.
Possibly the PP won’t keep that up. Who knows? But if someone thought it too cautious then you can always up the equity quotient a few degrees and slim down the defensives.
Plus we’re talking about glidepaths not gear crunches. So you would gradually shift from a riskier portfolio to reach the PP’s asset allocation say 5 years or less from retirement. Then hold it through peak sequence of returns risk, and perhaps, glide up again from there later.
I think risk tolerance varies so much. One of the reasons I wanted to show each equities allocation step in the tables was so people could pick out the level that’s acceptable to them.
@ermine: You write:
Hmm, now I can see why you’ve had such a downer on compounding all these years.
That’s an extremely pessimistic assumption about future returns (/read of the past). It’s less than a 2% real return CAGR over 40 years. (c. 1.75%)
In @TA’s table above, even a 20% allocation to equities and the rest in cash delivered more than 2% since 1970, and that’s hardly ‘equity heavy’. A 60% allocation to equities did a 4%+ real return.
You could speculate about the future, but as @TA implies you’d certainly need to show your workings. IIRC you could get around 2% real from a 40-year linker not long ago!
Am I missing something or was it a typo? Or too much homemade wine and birdsong? 😉
@ermine #8, @TI #10: Total equity returns with dividend reinvestment, and assuming no frictions, are 5% to 6% real p.a. over multi decades holding periods (30 year plus) based upon post-1900 data for 35 markets looked at by DMS, UBS/Credit Suisse yearbooks (and are drawn upon for most other studies, like Oxera’s in 2011).
Jeremy Siegel since 1998 has tracked US equity returns since 1802 at almost 7% real p.a.
Jordà–Schularick–Taylor in the “Rate of Return on Everything 1870-2015” finds long run average real returns at broadly around 6–8% p.a. across countries, with a sizable equity premium (ERP) over safe assets.
Robeco’s “Long‑term expected returns” study uses relatively new 1800–1914 datasets to build global bond and equity return series back to 1800 for several major markets, complementing DMS, and finding mid single figure long term annualised real total equity returns.
For the US, Edward McQuarrie’s work on 19th C stock and bond returns, summarised in the 2023 Financial Analysts Journal article “Stocks for the Long Run?”, and related papers, reconstructs total returns back to 1792, and challenges the notion that the high 20th C US ERP is representative out of sample.
But, for all of the impact it made in academic circles when it was first published, McQuarrie’s figures still don’t actually make all that much difference to the original DMS figures overall TBH.
David Chambers of CEPR, Elroy Dimson (of the DMS/UBS/CS studies), @ZX’s former colleague Antti Ilmanen (now at AQR) and Paul Rintamä (a Finish academic), together produced a nice comparative overview of the various different studies’ figures in 2024 in the Annual Review of Financial Economic.
And of course MSCI also has data from ~1970 onwards.
(And with apols for the typos/odd missing words in my earlier #7: typing on the move)
“Portfolio contributions: I haven’t tested the impact of continuing to invest throughout a bear market. I’ll run this analysis later in the series.” As this series is aimed at those close to retirement or in the early years of retirement would looking at impact of portfolio withdrawals rather than contributions be worthwhile – I would love to see it.
Great series – very timely for me as probably overdue some de-risking
@Sharkey – Cheers! I’m very glad it helps. On the withdrawal side, I can definitely look at withdrawal rate impacts on different portfolios in the future. Here’s what we’ve got at the mo:
https://monevator.com/safe-withdrawal-rate-health-indicators/
https://monevator.com/whats-the-safe-withdrawal-rate-danger-zone/
https://monevator.com/safe-withdrawal-rate-uk/
@TI #10 sorry the doubling was 30 years not 40, my bad. How did I determine this? I used your very own CI calculator and the default 4.5%, starting from ground zero. So full mea culpa on saying 40 not 30, but in practice people here need 30 years or less. And there is the small issue that people don’t earn at the max from the get go.
30 years is probably the long benchmark here, nobody comes to Monevator because they want to work a traditional retirement to 65, FI/RE and all that. I am a relatively late retiree at early fifties. I didn’t change down until quite late on, which was perhaps barking, but then I am an active chancer and was lucky.
So changing down halfway through one’s FI/RE career sounds like giving up much potential. Possibly viewed differently because I saw most career progression in the second half of my 30 year career, accepted this may not be normal in the more high-flying careers of your readership which take a Neil Young better burn out than fade away approach.
@TA #9 > The long-term returns show the Permanent Portfolio growing at the rate of the 60/40 portfolio
JP Greaney (who does show his working) just plain doesn’t agree and the shortfall is over 2:1 w.r.t 60/40 over 30 years. I’m sorry, but if J.P.Greaney is right the PP is definitely an old man’s portfolio.
I have much love for the PP, well, two of the components anyway, the third I hold for free and cash I always detest other than a full Premium bonds allocation and perhaps the same elsewhere, but I keep on trying to get that down, but JPG informs the prognosis for me, and equivalent 60/40 it ain’t
He could be wrong, but he comes across as a reasonably careful worker. I do hope you are right and that the prognosis is better, but I’ll be happy with JPG’s result
@TA – great series, thank you.
I de-risked practically overnight several year ago and have given up significant gains as a result. Don’t regret it for a minute.
I’ve got roughly 10% each of 10 years Gilts (VGVA), 3 individual Gilts maturing to 2028, MM and gold. I’ve tried Commodities, but I know won’t be able to stick with the under-performance. As your analysis doesn’t show much difference which maturity you hold, I may simplify a little. By default this will happen anyway as I consume each gilt as it matures followed by whatever else seems appropriate in my glide path to SP and DB, but it’s good to know it doesn’t really matter. At that point I’ll hopefully have 3 or 4 years of top up/fun money left in defensive assets that’ll be consumed and not replaced.
(I’ve also drank the TI’s Cool Aid and have BHMG and Pershing as they I hope they’ll produce a premium over FI and act as wild card insurance. These I treat as defensive assets and expect to consume)
Are you planning to show the effects of e.g. 60:40 with different combinations of defensive assets?
@ermine #8
‘I personally don’t do bonds…’
Me neither. So my version of the PP replaces those with ‘Managed Futures’.
And I wouldn’t exactly say that ‘Managed Futures’ are risk-off; but are very lowly correlated to stocks, so that’s good.
(My Equities / MF / Gold / Cash aren’t in 25% equal lumps, but more like 42.5% / 25% / 20% / 12.5%)
@ermine (14)
>….. nobody comes to Monevator because they want to work a traditional retirement to 65 …..
I did. I’ve been reading Monevator since perhaps 2013, at which point I fully expected to keep working until 65. It was only an unexpected voluntary redundancy opportunity at age 63 in 2017 which caused me to retire slightly early.
I came to Monevator because I have always been interested in finance and investing, and I liked its style and content, as well as the large number of well-informed comments.
@Algernond I didn’t know what Managed Futures were until I looked this up, thanks for the heads up!
@DavidV Charge of hyperbole accepted. I was possibly stuck in the past but I have looked at About and it doesn’t mention FI/RE
More generally JP Greaney’s analysis is in terms of drawdown @4%, but I see no reason why the relative performance of the PP and the 60/40 should differ that much in drawdown relative to accumulation, though you’ll end up with 4%p.a more. So you could in theory accumulate in the Permanent Portfolio, but it is still tortoise to the 60/40 hare.
Optimised Portfolio did the accumulation comparison and it’s difficult to intuit the % difference due to the log scale. The PP is a smoother ride, but still slower.
I was tickled by the US author’s view
Them days are gone, my friend, the spirit of Volcker stalks the world no more, he is the ghost of a different America.
@Ermine – Re: J.P.Greaney portfolios: 30-year time period, US returns, different inflation experience. It’s not apples to apples. Though at least it shows that outcomes can be quite different over an individual’s time horizon.
Regardless, I don’t think it’s about being right or wrong or that long-term GBP results show that someone should plump for the Permanent Portfolio necessarily. I think the results show the potential advantage of broader diversification. I’d advocate adding index-linked gilts to the mix, too.
@Brod – Cheers! Yes, my own derisking path was pretty steep and sudden too.
“Are you planning to show the effects of e.g. 60:40 with different combinations of defensive assets?” I’ve started noodling behind the scenes. I worry a little though about over-optimising on past asset allocations. Do you think that’s an issue or is it just interesting to see what difference playing with the allocations makes?
@TA – I wasn’t so much focussed on optimising a portfolio for a past that will not repeat, though may rhyme of course, but demonstrating what would have happened if you had mixed a few different asset classes while drawing down 4% + inflation in a bear market from the asset class that did best.
E.g. in the Dot.com crash this is what would have happened if you held 60% equities, 40% Intermediate Gilts. However, if you allocated 10% of Gilts to Gold, this would have happened. And if you allocated another 10% of Gilts to MM this would have happened… However, in 2008 these are the results… And in 2021, the would have happened…
Although that’s maybe a lot of work… 😉
Me, I don’t know what’s going to happen so I’ve been holding MM, Gold, 10 year Gilts and short, individual Gilts to maturity. And BHMG and PSH just in case. And only 40% equities.
@david V
and
@ermine (14)
>….. nobody comes to Monevator because they want to work a traditional retirement to 65 …..
same here – 67 and working (yes by choice) .. and here because I’ve learned so much about personal finance from Monevator
@Brod – yes, I can do that these days. I’m working on making it less work than it was, too. Plus, if I spread it across enough different posts…
Thank you for the thoughts, it all helps. If anything else comes to mind then please chuck it in the thread. That goes for all @Not-Brod too 🙂 It’s really useful to know what people are interested in and this series was prompted by a reader email – though I’ve lost track of who it was. Perhaps @TI can remember?