This article can be read by selected Monevator members. Please see our membership plans and consider joining! Already a member? Sign in here.
What to do if you’re queasy about the US stock market [Members]
Next post: Trend following: is the trend your friend?
Previous post: Weekend reading: Spring Budget 2024 roundup
Comments on this entry are closed.
> into an ex-US global tracker
That’s all very well, but I have yet to find something aimed at a non-US-market. The index FTSE All-World ex-US and the ETF exists – VEU. No, not that EU, this is marketed at US good old boys, not effete yurpeans (or non-yurpeans). I haven’t found anyone selling that index to Brits, presumably on the principle nobody would be mad enough etc etc.
I want this, to split some VWRL I have to sell to harvest this year’s CGT allowance, and perhaps boot the VUSA element into next year’s ISA and leave the deadbeat FTSE All-World ex-US ETF in the GIA to deprive the taxman of revenue by concentrating the profitable part in the ISA.
So it’s not quite going along with the thesis of this article, but abusing the principle for different aims.
UK ISA to the rescue? Some UK gov bonds and shares to dilute US? For Cap gains harvest I’m currently cashing out into 5% cash rather than switching to a different global tracker.
My plan, is to keep my equity asset allocation as is (65% in my ISA), but have this to from 65% VWRP to 55% VWRP and 10% VHYG.
This will do a little bit to reduce my overall US equity exposure slightly, and bring a slight value tilt to my ISA.
Lots of interesting information here but does it alter the underlying premise of just investing in the world stock market index in stocks and bonds and leaving your investments alone-well not for me
The delight of a total stockmarket tracker is that it automatically adjusts to the index -so if you are worried about a US preponderance then the tracker will adjust as the US goes off the boil (when ever that is) without any worries from the investor getting his hands dirty performing market timing/adjustments etc-all those great excuses for tinkering !
Regular market drops in equities and bonds and their subsequent recoveries are a constant market lesson that reinforces the fact that nobody knows anything !
Having just sat through many of these ups and downs( with my selected Asset Allocution in place ) without doing anything at all -over my 78 years -it does seem to be a successful if very boring investment policy
Betting against the US of course has been a losing investment plan for many decades-so far!
xxd09
@ermine — Eek, really? I guess this is what happens when the stock picking junky makes assumptions about funds.
Oh well, pending a solution from anyone else I suppose you could just buy a European and an Asian tracker and get much the same. I’m more cavalier about this stuff though than purists I realise.
Ridiculous there’s no ex-US tracker when it’s 65% of a standard world equity fund though!
I tried to reduce my exposure to the US, but as Ermine mentioned – there’s no fund that does that in the UK. So I ended up getting two different funds, one for european stocks and another for emerging markets.
I think it’s a valid decision for passive investors, because – entirely consistent with acknowledging that we can’t reliably pick winners and losers, there are many risks which are not directly related to market-capitalisation.
To explain – consider that every $ invested carries with it a “performance risk”. This is distributed in a similar way to market capitalisation – there are far more $$$ in the US, therefore the risk is larger in the US because it actually has more companies and more value.
On the other hand, some risks aren’t distributed like that. For example, the risk of societal/geopolitical crisis isn’t per $, it’s per country – the US political system isn’t less prone to collapse because it’s economy is larger. This means, for example, that the risk of the US facing an internal political crisis is (to simplify) not materially different than, say, the UK. If the risk is identical, it makes no sense to bet 65% of our future on a country that could fail just as easily as other countries (again, this is hard to measure, but our error could go both ways). Therefore, a somewhat less efficient distribution from a returns perspective can increase diversification when it comes to other risks.
I think what is important from a passive investment perspective is to be consistent and clear with your strategy. As long as you define the distribution you’re happy with, don’t fiddle with it as markets go up and down – that’s where things start to go wrong.
(As an aside – what we do have a lot of is ex-UK funds, which I find a lot less useful. The UK is a small market so we’re not taking too much of an all-eggs-in-one-basket risk (like when US investors invest heavily in the US) by simply using a weighted global tracker. It only really serves to simplify things if you’re going to tilt in favour of the UK, which course feels like a bad idea for other reasons)
Excellent article. Thanks @TI.
The arithmetic of to loss to gain favours smaller adjustments.
Take a toy example:
Starting value= 1, all in shares.
In scenario A shares double.
In scenario B shares halve.
If you go to cash, and A happens, then you miss gains of 1 (2-1).
But, if A happens, then you only avoid losing 0.5.
So, the ratio is 2:1 against going from 100% shares to cash if both scenarios are equiprobable.
If, however, you decide to move 20% from shares into long duration bonds (i.e. from 100/0 to a 80/20 shares/bonds mix) and, say (for the sake of illustration only), such bonds show a -0.5 (i.e an inverse/negative) correlation to equites then:
– In Scenario A you go from 0.8 to 1.6 value in equities and bonds fall 0.2 to 0.15. You end up with 1.75, not far off of the 2 which you’d have ended up with if you’d kept at a 100% equities.
– In Scenario B you go 0.8 to 0.4 value in equities, but bonds go 0.2 to 0.25, leaving you with 0.65 overall.
So, overall, at the risk of losing 0.35 in Scenario B (1-0.65), you can gain an extra 0.75 in Scenario A (1.75-1); which is a 2.14:1 ratio (i.e. 0.75÷0.35) in favour of moving from 100/0 to 80/20 in equities/bonds (and only 0.25 less than if you’d stuck with a 100% equities).
Basically, in adjusting asset allocations, less can be more.
IMO, return stacking (i.e. WTEF UCITs ETF in the UK, NTSX ETF in the US) can be a way of achieving this without sacrifing too much expectation return. WTEF is 1.5x a 60/40 SP500 / intermediate duration US Treasuries (comprised of 90% S&P 500 capitalisation weighted physical shares and 60% exposure to intermediate duration US Treasuries with 10% in 6x Treasury bond futures using a bond ladder of different durations).
60% in WTEF, 20% in non-US regions’ shares (a capitalisation weighted mix of European, Asia Pacific ex Japan, Japan and Emerging Markets equity trackers) and 20% in intermediate duration Gilts (conventional and/or index linked) would give you an exposure of:
– 54% to the S&P 500 (compared to 65% in Vanguard’s FTSE All Cap Global Index Fund);
– 36% in intermediate US Treasuries
– 20% in ex-US equities
– 20% in gilts
For a total exposure of 130%, and giving a rebased to 100% allocation of:
– about 42% S&P 500
– about 28% US intermediate treasuries
– about 15% ex-US equities
– about 15% gilts
@TI. When I looked I didn’t see much available in world equal weight funds, though there is an Invesco RAFI fund which charges around 0.4%.
I am trying to derisk and move towards a sustainable yield drawdown type approach and unfortunately things like Life Strategy don’t offer much yield (even an ‘income’ class isn’t much use if subfunds are acc.).
Because of my particular circumstances I am trying to derisk into more bonds which is difficult as it means giving up on growth ( the compensation of less exposure to drawdown somehow doesn’t have enough emotional recompense), bond duration not to be more than about 5 years and lots of short duration index linked, and adding a few side hustles to the equity side (Latin America and Pacific index funds…..surprisingly good yield) and some commodity exposure ( cheers TA) in case inflation gets away from us.
US exceptionalism
Land & resources – continental advantage (friendly neighbours, separated from nearest enemies by oceans), significant natural resources base & energy independence, manufacturer of most things
Human capital – 60% of worlds best universities, global centre of tech innovation, attracts and retains best and brightest
Geopolitics – $ the principle global medium of exchange / reserve currency & store of value. A unipolar financial superpower. Military superiority. Strong influence over world order & supranational organisations
Rules based framework – capitalism and free trade beneficiary, independent institutions, constitutional checks and balances
Culture – capitalist, pro investment and wealth creation, shareholder friendly
The structural advantages support dominant capital markets, lowering the cost of capital, attracting more flows. Ever higher share of index tracking funds feeds the momentum. The loop feeds on itself until something breaks it
What shifts the momentum and who takes the market leadership
Europe with our ageing population, low productivity, less capitalist culture, lack of resources, old world industry weights and an ongoing war on the periphery
China under the CCP
India maybe but a long way to go to create the rules based framework for growth that sustainably correlates with equity returns. Not exactly cheap after the rally either, neither Japan
Wider EM not getting anymore stable
The UK…
US equities highly rated for most of the past 15 years. Why mean reversion now. How to weight the risk of being long and wrong vs the risk of being underweight and wrong
Grantham / GMO’s flagship benchmark free allocation fund annualised 3.07% in the last 10 years so a real return of zero over a decade
With a sub 15 VIX and call put skew at a record I wonder about hedging with puts as a means to mitigate the downside but struggle with it philosophically vs reinvesting all cashflows
Erratum (and breaking my rule to only comment once per article): it should be 1.7 v 0.65 and 0.35 v 0.7; not 1.75 v 0.65 and 0.35 v 0.75 (with perfect symmetrical -0.5 correlation).
So, the difference in outturns between scenarios A and B of a small change (100 shares going to 80 shares and 20 bonds) v a large one (100 shares to all in cash) is actually 2:1 in favour of the small change v 2:1 against the large (and not 2.14:1 as stated).
In reality, it would be surprising for bonds to halve if equities doubled, but plausible for bonds to go up 25% if equities fell 50%.
So, by chance, the 2.14:1 ratio in favour of a small asset allocation change, v 2:1 against making a large one, probably is closer to IRL.
Also, small changes make more sense simply because no one knows anything for sure about future prices. Only the other day I came across a 2017 Monevator weekend reading link to a fund manager confidently saying that the equity market was in a bubble. The fund manager in question? Neil Woodford.
We miners are simply minded passive types, we’ll stick to the plan and heed what my grandma used to say “make hay whilst the sun shines”
We are interpreting that as keep the eq:bonds/cash ratios in check with
rebalances and don’t get greedy.
For me, I see the primary risk to be that the US market returns are just really poor for the next 10 years, whereas if the US market crashes then everywhere crashes. So while I don’t feel a need to reduce the overall weighting to the US market (it is after all the most liquid and dynamic stock market globally with earnings sourced globally), i do think there is merit in moving from a market cap weighting for the S&P to an equal weighting index. Only in the last few years with the FAANG then Mag 7 domination have the market cap and equal cap weighted indexes diverged, so it seems reasonable to expect them to reconverge at some point. You’d think. Quite how to implement that strategy when you have a single global market cap weighted equity fund is a different matter though….
Timely article, thanks TI! As many pointed out above, in the absence of an ex-US fund I’ve compensated the Developed World tracker with smaller allocations to Europe, Japan, Pacific ex-Japan, and Emerging markets to get the US allocation of 45%. Could be simpler, but is doable.
Great article. Unlike most others here, I have moved some funds (30%) of my ISA to a Money market fund. I had been considering this anyway as I anticipate I may need a good dollop of cash in a year or so and my ISA was heavily tech and US focused and the index is high. Sometimes it is necessary to take some money off the table. I shall try not to look at the S&P for the next year (ha!). My Sipp remains unaltered as I hope it is a long term investment (I am retired).
What I do with my money is assess the risk. These articles are nicely balanced to pick out the opportunities and threats with enough caveats to only shout “that’s market timing” for giggles.
I have parked these reflections in my subconscious for the next time I can be arsed to tinker with my strategy.
FWIW I think AI is still in its infancy so a portfolio positioned to take advantage of that might be a good bet. The key word there is “bet” and I try not to speculate too hard because of the nosebleeds. Feeling like I am informed of risk is enough, especially the risk that no one knows for sure what’s going to happen.
Unfortunately, I’ve been taking most of these approaches for the last eight years or so, and watched the US tech giants climb higher and higher. Thank you for at least reassuring me that I’m not completely deluded to think the global tracker is a bit unbalanced. However, I have upped my US weighting over time to keep somewhat in line with the rest of the world! I’ve come close to switching to a global tracker a few times, not least because the lack of an ex-US fund means my portfolio is more complex than I’d like.
@Ian — Thanks for sharing, and yes it can be lonely going against the crowd. (And expensive, at least in terms of opportunity cost). I’ve been in-part a growth/tech investor for most of the past 15 years, so dialling down this exposure in my own portfolio more recently hasn’t been easy I concede, though perhaps not too expensive so far.
However I would stress ‘dial down’ rather than ‘avoid’. I still have plenty of tech exposure and a fair bit of US exposure, though far below benchmark now.
As always I think going all-in/all-out something is a very risky call when it comes to broad markets (versus individual stocks or similar). My knowledge is far from perfect, and my judgement is even more fallible. No worries, everyone’s is, but it’s a reminder not to be absolutist IMHO. It’s also hard setting one’s face against momentum from a proven return premiums point of view!
Hence my suggestion in the post of targeting US exposure at 50% — down from 65% but not to 25% let alone 0%. 🙂
Just came across an article from last week by passive investing guru Larry Swedroe on high US valuations over at Kitces.
FWIW Swedroe writes in the exec summary:
But there’s a *lot* more in the full article:
https://www.kitces.com/blog/us-equity-valuations-investment-cape-10-diversification-stocks-sp-500/
A new ETF …
https://etf.dws.com/en-gb/IE0006WW1TQ4-msci-world-ex-usa-ucits-etf-1c/
@Kwaker — Nice spot and how timely!
I could only access the fact page by saying I was an intermediary. Perhaps platforms are going to restrict it to somehow to professional investors/advisers? Hopefully not (that really would be barmy) and it’s just because it’s new.
EXUS (MSCI world ex-USA) is now available to retail investors on AJB. It’s not currently available on ii or Fidelity. I haven’t checked other platforms.
A reminder, if any were needed, that the only constancy is change.
From this week’s IC: 1949-59 Japanese equities return an astounding 1,500% (albeit off of a very low, war devastated base). Nonetheless, bar the oil shock in the 1970s, they contine rising for another 30 years.
By 1989 Japan appears unassailable, with a 45% share of global equity market cap, as compared to just 29% for the US.
Fast forward to now, and despite Japanese equities finally regaining their prior nominal all time high from 35 years ago, the 60% US share of global equities cap is now 10 times the Japanese figure (of just over 5.5% of the global equity MSCI benchmark in 2022).
Re: #8 and #12 above, and observation in @TI’s piece that, “You could track an equally-weighted index, for instance, instead of a market cap-weighted one”:
this 2018 article on portfolio construction/diversification somewhat pours cold water on equal weight being superior than market cap weight:
https://qoppac.blogspot.com/2018/11/is-maths-in-portfolio-construction-bad.html
@Delta Hedge — Well it depends what you want your equal-weighted index tracking to achieve.
It wasn’t mentioned here as a superior form of index tracking, in particular. It was mentioned as a viable way of not being so exposed to 5-10 extremely large companies that seem to be on a very extended run (amazing businesses though they are) if one is minded to do that.
Equal weight definitely avoids risk of over concentration within indices (at the risk of over dispersion). It also is a, if not the, logical choice if/when uncertainty about the path of future returns, volatility & skew is of paramount concern, & if confidence in the past as a guide is weak.
I’m finding it v. difficult to guess whether (or not) to actually take any action over the Mag 7 hitting 30% of the S&P 500 cap, and the S&P 500 hitting (depending on the Dev. World tracker methodology in question) a ~60% to ~70% share.
I suspect that Lars would say do nothing. Philosophically that would be the pure passive approach.
On the one hand, you have this: US Stock Market Cap as % of GDP:
1984: 42%
1994: 63%
2004: 93%
2014: 114%
2024: 187%
On the other hand, the combined revenue of just Amazon, Apple, Google/Alphabet & Microsoft hit $1.53 trillion over last 12 months – larger than the GDP of all but 15 countries.
It’s v. unclear whether & to what extent to be concerned. As always, time will tell.
More context from JPM just in on mega cap US tech outperform: “The good: We’ve reached the 30th anniversary of tech sector profitability crushing other sectors. As things stand now, tech and interactive media free cash flow margins are 3x higher than the rest of the market. Since 1990, large cap US tech stocks outperformed the MSCI World Equity Index by 8,000% to 1,000%. Also good: unlike in 2020/2021, there’s less of a spike in the market cap of unprofitable tech companies (the YUCs).”
The bad is that, even on a forward earnings estimate basis, their valuation is back to 2021 levels.
Interestingly perhaps, on data that I think goes back to 1990, Equal Weight S&P 500 shows an outperformance in 53% of periods and market cap weighted S&P 500 in 47%, so very close, with average P/Es of 15.5x v 16.2x respectively. However, currently the P/E difference is 17x v 21x.
Moreover, since 1996 the top 10 stocks in the S&P 500 have had average forward P/E ratios of 20.3x compared to 15.7x for the remaining 490, but currently the difference is 28.4x v 18.3x.
Always worth remembering that no one knows anything for sure about what valuations truly mean for the future path of returns, and which metrics are relevant (and which are not) to that future.
Inverting here and asking instead what evidence points away from excessive richness in US stock prices:
Paulsen Perspectives: on Market-cap vs. equal-weight: “The SP 500 MC index has outpaced the EW SP Index by quite a margin in recent years, but this chart suggest it could surprise most by continuing to outpace for some time and by much more.” Chart@:
https://x.com/jimwpaulsen/status/1804120013810417861
Thoughts on CAPE etc @:
https://open.substack.com/pub/paulsenperspectives/p/a-different-view-of-valuation
Sam Roe via UBS: “P/Es today are a full standard deviation lower than in the late 90s. On a FCF basis, the S&P 500 is 2 standard deviations cheaper.” Chart@:
https://x.com/SamRo/status/1804170533241512343
The latest on US stocks v the rest and US stocks v other assets’ valuation, from Top Down Charts / Callum Thomas:
https://open.substack.com/pub/topdowncharts/p/12-charts-to-watch-in-2024-q2-update
4 things stand out:
1). The risk of deflation, notwithstanding a residual risk of reflation / ‘higher for longer’. A long duration US Treasuries’ (UST) ETF could be a way to deal with the former, and a Broad Commodities ETF with the latter.
2). UST v SPY ‘Valuation v History’ chart (1983 to date): UST looks a ‘bargain’ and SPY not, relatively speaking. Buying a bargain and selling expensive often doesn’t work though; so perhaps one way to possibly play this one might be to – in effect – take both sides of the bet with a returned stacked product like WTEF ETF, which is 1.5 x a 60/40 (i.e. a 90/60) SPY/UST, rather than just selling SPY to buy UST.
3). US Tech Stock Valuations chart (1984 to date) with equal weighted composite Z scores for CAPE (‘PE10’), Price/Book, P/Div, EV/EBITDA and P/Cash Flow: the slope is not yet as parabolic as it was in 1998-99, but it is steepening. The Z scores are shown, as Sam Ro noted in the link in my last post, as being a good StdDv below the crazy 1999/2000 peak, but this doesn’t mean a blowout will or won’t occur – less still tell us when and how, if it does.
4). Global ex-US Small Value v US Large Growth chart (1995 to date): what a stonking difference over time. Former is now at around a fifth of the relative price (ratio) of latter as compared to 1995, 2007 and 2009, and is equally relatively cheap as it was in the crazy times of 1999. Extraordinary. I’m overweight both ex US Developed and Emerging Market Small Cap Value (‘SCV’) in consequence (as EM is now relatively very cheap indeed compared to history and to DM generally and the SPY specifically). I wonder, therefore, if I should be still more overweight ex US SCV. I’ve kept direct SPY to 50% and full look through US stock exposure to 56% so far. A tilt away from the US rather than something more radical.
@Delta Hedge — Thanks for the updates. I continue to soldier on with I’d estimate about 35-45% US exposure in my equities. (I’m being imprecise because a bunch of it is on a ‘look through’ basis via global / tech trusts and similar.) The sledding is not easy in the face of such a headwind, but I do rightly or wrongly feel good about being less exposed to nVidia when it was at $3trillion for three seconds.
Macron hasn’t helped things though, with some European quality stocks I hold selling off again with his snap election…
@TI – regret minimisation’s a plan, but with losses 2-3x as painful as gains are satisfying, my best guess is that an ability to take on some pain of being ‘wrong’ in the here and now (i e. overweight ex-US stocks and commodities) could in the future be asymmetrical rewarded.
The one thing which I’m not doing though is taking any actions based on predictions for index levels in the near term (i.e. SPX Dec. 2024 targets now at from 4,200, from JPM, to at 6,300, from GS) .
That game scared me out of equities for a while, at just the time that I should have been doubling down following your prescient March 2009 shares on sale piece.
Case in point: Dec. 2008 year end targets for 2009 for the FTSE 100 ranged from IFA Alan Steel’s 5,600 call to Sandy Chen of Panmure Gordon’s 3,500 – in other words, literally useless.
Better, IMO, to think in broad, long holding period asset allocation terms.
What’s the likelihood for the primary asset allocation trend over the rest of my life?
Is it more or less likely to be a rising rates, multiple compression, higher inflation scenario, like the mid-sixties through to 1981, than it is a falling rates/inflation and rising multiples (and earnings) scenario, like 1982 to 2021 (albeit via a valuation bubble in 1995-99 and global economic stroke in 2008-09)?
The effect of the last two primary trend regime changes are seen strikingly in the different ‘behaviours’ of the median CAPE for SPX over different timescales, namely:
– The median CAPE ratio was 15.04 times earnings from 1871-1980.
– The long-term median CAPE is 16.52 times earnings from 1871-Present (all years).
– The median CAPE is now at 23.70 times earnings from 1980 to the present.
Arguably, the shift higher in median SPX CAPE was likely a function (to a significant but debatable extent) of falling economic growth; and a lower discount rate being applied to projected future earnings as a consequence of the lower for longer rates regime of deflationary pressure (with China etc joining the world economy).
I asked Perplexity AI (which is closer to a – IMO – much improved Google search engine, but with useful cross referencing provided, than it is to say Chat GPT) whether there were any buffer ETFs available in the UK for the S&P 500.
This is what it said:
“Global X offers two buffer ETFs that are listed on the London Stock Exchange: Global X S&P 500 Quarterly Buffer UCITS ETF (ticker: SPQB for USD shares, SQBP for GBP shares). This ETF aims to provide a 5% downside buffer on a quarterly basis while tracking the S&P 500 index. Global X S&P 500 Annual Buffer UCITS ETF (ticker: SABP for GBP shares). This ETF seeks to provide a 15% downside buffer on an annual basis while tracking the S&P 500 index. Both of these ETFs use options strategies to create the downside buffer, while capping upside potential. They are UCITS compliant, domiciled in Ireland, and available to UK investors. The Quarterly Buffer ETF has a total expense ratio of 0.50% and was launched in February 2023. It uses a synthetic replication method with an unfunded swap structure”.
I then checked them each out, and they each have trailed the S&P 500 bigly over the past year (they each have short trading histories, so a longer comparison is not possible).
It’s a really huge amount of relative underperformance. Basically we’re into a 5-10% returns versus a nearly 30% return territory here.
So it looks like neither of the buffer ETFs so far available in the UK for the US market are the answer if you’re concerned about US crash risk.
A 20%+ p.a. performance lag (i.e. over two thirds) is not IMO ever an acceptable lag for a downside hedge product. There has to be a better way.
Relatedly, I see the range of outcomes covered by the forecasts for the S&P 500 by end 2025 has expanded from 3,750 for the pessimists to 7,000 for the optimists.
Chat GPT suggested the following to reduce US exposure to 50%
Allocation:
77% in Vanguard FTSE All-World UCITS ETF (VWRL)
23% in Vanguard FTSE All-World ex-US UCITS ETF (VFEG)
US Exposure: 50.05%
Non-US Exposure: 49.95%
Overall TER: 0.23%
This seems easy enough and also manages the risk somewhat. Any arguments against this?
@Saba: I think ChatGPT may have let you down here: VFEG is an emerging markets ETF. It reduces US exposure but it’s not All World ex US. It would be useful if there were such a thing available here, although I understand one is available in the US with the ticker VEU.
@Saba worth checking out @Finumus’ ‘Investment costs how low can we go?’ piece on Monevator on 4 May 2023.
For example (not a recommendation, only an example for hypothetical illustration purposes) one could use the swaps based SPXP.L ETF (the Invesco S&P 500 UCITS ETF – Accumulation), at a Total Expense Ratio / Ongoing Charges Figure of 0.05% p.a. (with no FX fees, nor any US withholding tax on the dividends), for the US weighting (assuming, of course, that one was happy with having only US large capitalisation stocks and a market weight exposure).
If someone did this, they could then use region/area specific ETFs, per the @Finumus’ piece, for the non-US countries (i.e. those ETFs covering Europe, Asia-Pacific excluding Japan, Japan and the Emerging Markets).
This enables some tailoring of regional exposure, if desired.
If someone wanted to basically stick to a global market capitalisation allocation, but just to dial down the USA to 50% of the total, then one would look to see what percentage of the non-US share in (say) Vanguard Global All Cap Fund was both allocated to each of the non-US developed market regions and to the Emerging Markets, and then apply those shares/percentages to the example 50% non-US allocation.
A few years ago, Moneyweek had a short article on a similar theme to this one and proposed a GDP weighting rather than market weighting of investments. At the time this was:
North America (25%)
Developed Europe (25%)
Japan (7.5%),
Emerging Markets (37.5%)
Developed Asia Pacific ex Japan (5%)
(The writer also showed how to do this with 5 Vanguard funds.)
The percentages are, no doubt, slightly different now but still not concentrated as 65+ % in US.
Excellent idea @MogulMarkR.
Investors Chronicle today looking at global valuation and the US risk and suggesting:
– No more than 60% in equities.
– Of which reduce the US market share from 65% or so by capitalisation weight down to 40%.
– Put the 25% difference into discounted UK investment trusts.
Going down to 25% for the US would then enable 25% each into the US, UK ITs, Europe and Emerging Markets; which, if nothing else, has the ‘benefit’ of simplicity.
US Stocks vs. US GDP. For the first time in history, the market capitalization of all active stocks in the USA is more than twice the nation’s Gross Domestic Product.
How much is down to the zeitgeist over AI?
The turn up in the US market did roughly coincide with the release of the first useful version of ChatGPT.
Coincidence is not causation but…if this goes south it’s hard to imagine US mega cap tech not playing an outsized role.
Ed Zitron thinks LLMs are heading for their Wile E. Coyote moment (or should that be jumping the shark?):
https://www.wheresyoured.at/godot-isnt-making-it
Gary Marcus agrees:
https://open.substack.com/pub/garymarcus/p/chatgpt-at-age-two
US cap weight share in large cap developed world trackers now an unprecedented 74%.
So if we’re not into a new paradigm, and they’re right, then brace for impact because it could be a long way down.
Referencing back to @AoI #33 on JGPI ETF in the comments here:
https://monevator.com/how-gold-is-taxed/comment-page-1/#comment-1849265
The LSE GBP equivalent JEPG ETF (listed as JEGP ETF at HL) could be a way to ‘hedge’ the risk of equity overvaluation – summary of the main pros and cons here:
https://europeandgi.com/etf/jepg-etf/
TL:DR Sideways market and the stocks should get called less meaning premium kept and stocks kept. Down market similar but with covered call premium to buffer drawdown in capital value of stocks. Up market underperform even with premium from covered call writing. FWIW I’ve dipped my toe in this ETF today.
A hypothetical ideal all in one product here might offer covered call writing on defensive global equities (60%), gold (as with GLDE) (20%) and broad commodities (20%), as well as combining the characteristics of JEPG/JEGP with the capital efficient leverage of WTEF by holding 90% of the 60% in the equity sleeve as physical stock replication (with covered call writing) and using the balance to lever up 6:1 exposure via rolling 3 month futures to intermediate Treasury bonds. This would give a 30% overall leverage (130% total exposure) with 94 of that 130 generating cover call writing premiums and with an economically equivalent stock/bond / gold/commodity ratio of exposures of 54:36:20:20. One day perhaps.
Then again, perhaps the FCA will ultimately amend the PRIIPS regulations and allow ISAs and SIPPs to hold the Direxion HCM Tactical Enhanced US Equity Strategy ETF, giving the option of addressing US valuation concerns through a fully automated Tactical Asset Allocation strategy to the S&P 500.
Although valuations are stretched by orthodox metrics (P/E, P/S, P/B, FCF yield, ERP, Buffett’s Market Cap/GDP); since 1957, when the S&P 500 replaced the S&P 90, its total (dividend reinvested) nominal return has been 10.2% p.a., whereas the return from 2004 (at which point in time most observers would not say now, or have said then, that it was stretched) to date comes to ‘just’ 10.1% p.a. (albeit that the return since 2014 has been a more creditable 13.2% p.a.).
The ‘meaning’ of valuation is very much in the eye of the beholder. It purports to be a wholly objective yardstick, but nothing is, or can ever be.
From GMO at minus 2.9% p.a. to Columbia Threadneedle at plus 5.4% p.a. , what stands out in this piece today from Joachim Klement on the US ERP is both the range in the forecasts of what to expect from the US coming out of different firms of institutional investors and the underlying optimism on the US from pension funds relative to wealth managers:
https://open.substack.com/pub/klementoninvesting/p/how-big-is-the-equity-risk-premium
The first point seems to underscore that when it comes to the future no one knows anything, whilst the second point resonates with the economist Thomas Sowell’s observation that if you show me the incentive then I’ll show you the outcome.
Re my #38 – a word of caution. Popularity in the US of short-dated options (zero day options typically) has structurally impeded implied correlations as hedging increasingly moves to single-day events from the 30-day horizon, leaving the monthly options with little demand and large aggregate supply due to increased overwriting of index calls for income. This suggests to me call writing ETF strategies – whilst potentially less risky (as based on global defensive stocks, which are less likely to get called, rather than US growth stocks) – are perhaps not quite such a great deal.
Regarding availability of Xtrackers MSCI World ex USA (Acc).
I asked ii if they are going to offer it.
Their (long) answer was:
Unfortunately, you cannot purchase EXUS through ourselves as there is no or failed VFM (Value for Money) information on the stock.
If we do not have a completed VFMA (Value for Money Assessment) for a stock or if the completed assessment confirms that a stock is not value for money for investors, we cannot set it up for buys.
This is due to the new Consumer Duty Regulations.
Consumer Duty is a new standard introduced by the Financial Conduct Authority (FCA), which came into force on 31st July 2023, and sets higher and clearer standards of consumer protection across financial services.
This requires a “Value for Money Assessment” (VFMA) to be performed on products to determine if they should be offered to retail clients.
Please note, fund managers and providers of ETFs have had to confirm if their ETF/Fund is good value for money and justify their charges by completing a VFMA.
It is the responsibility of the fund provider/ETF provider to do this.
Some fund managers/ETF providers have said their fund/ETF are not good value for money, and some have just not provided the information yet.
It appears that trading212 do, so I plan to open a (free) account with them for this year’s ISA.
https://www.trading212.com/trading-instruments/invest/EXUS.GB
Perhaps consider using Merton share to calculate a reasonable dynamic % exposure to the S&P 500 (Merton of 1997 Nobel Prize in Economics fame, and LTCM infamy). Merton share calculates optimal allocation to equities based upon expected equity return (for which could use S&P 500 inverse CAPE / 10 yrs’ average Earnings’ Yield), risk-free rate (could use the 30 yr US Treasury bond yield), volatility (i.e. square of S&P 500 annual rate, expressed as a fraction, e.g. 0.2 squared), and risk aversion (1 as average).
Details here:
https://elmwealth.com/how-much-happiness-does-money-buy/
Calculator here:
https://evergreensmallbusiness.com/merton-share-estimator/
The orange agent (emissary?) of chaos has now pulled the tariff trigger.
Is it all a big bluff, like Nixon’s ‘madman’ strategy which reputedly brought both the North Vietnamese to the negotiation table and the Soviets to Detente?
Or does he really and truly believe all this stuff (God help us all if he does)?
Anyhow, uncertainty again abounds:
– Does DeepSeek change the game? Whatever becomes of LLMs, and whatever comes after them, will the US hyperscalars and semis be the winners?
– Can any of this stuff actually be effectively commercialised? Whose going to pay, who will get paid, for what, and how much? Can’t see capex alone infinitely sustaining US mega cap prices. At some point there’s surely got to be some compelling pay for use cases with chunky revenue streams.
– OTOH, noone knows anything. Whose’s to say that we don’t get glimmerings of AGI and then the market looses its mind in euphoria and we melt up in short order to 10,000+ on the S&P 500. Sounds crazy, but think back on what we’ve been through just since Covid. None of that was ‘in the script’ in 2019.
I’m an equity guy. I like the uncapped upside asymmetry. But bonds look in some sense ‘cheap’ compared to at least some of the more traditional valuation metrics for the US large cap indices.
5% in bonds to sit on the sidelines beats sub-1% from the 2010s, that much is for sure.
But is that the winning (right) mentality here? Does avoiding the big loss mean loosing the big gain?
Are the possibilities of AI/AGI/ASI (and, for that matter, of quantum computing, of mass produced autonomous & humanoid robotics – including FSD & fully automated manufacturing – of nanotech, and of next generation genomics) actually already ‘priced in’ for the S&P 500?
Think about it. Does 6,000 on the S&P now feel ‘right’ if Wall Street truly believes that those things are really round the corner (5-20 years’ horizon)?
I doubt it.
If WS was truly treating those possibilities as being more than remote (1%-5%) then the index would surely be over 10,000 right now.
The sorts of impacts which full realisation of those technologies likely would bring would be at the scale of multiples of the present Gross Global Product.
They also probably would be highly deflationary, thereby likely leading to low, or even negative, rates which would, by itself, significantly push up valuations.
Moreover, the winner takes all nature of such a scenario would seem more likely to raise the share of profits within economic output.
That all implies that WS thinks this very likely is not going to happen – i.e. assuming only a low single percentage figure that it does.
The Street may very well be proved right here. It does all seem rather far fetched, albeit a scenario which is clearly possible in principle.
But *if* the Street has underestimated either the likelihood of the technology realising, or the extent of its impacts if it does, then it will have substantially undervalued the present value of big tech and mega caps generally.
This to me suggests that a deep out of the money long dated index call option strategy (like LEAPs, but over 5 years, not 2) is perhaps, in some sense, more attractive than might now be thought to be the case under the current concensus view.
Losing say 5% on such a call option on SPY, SPX or QQQ ETF and then putting the remaining 95% into risk off assets like bonds, gold, commodities and/or non-US deep value or SCV equities might be a better bet than staying 100% in the S&P 500 or the Nasdaq.
If America ends up disappointing, then the alternatives should provide a return.
And if, alternatively, there really is going to be a technological singularity in the offing, then the deep OTM long dated call option could explode in value, making up (and more) for what would then be the drag of the risk off assets and value shares in the portfolio.
@Delta Hedge — I think you’re being rather confident that achieving the singularity would be a net gain for S&P 500 companies (/the economy) — leaving aside ‘for humanity’ 😉
Having ‘data centres full of geniuses’ on demand, perhaps being tapped for 97% less energy per DeepSeek (I know…) sounds potentially rather challenging for a load of incumbents…
Of course you might still argue that people will shop at Walmart and drink Coca-Cola and that even these consumer goods companies will be run more efficiently thanks to AGI, but equally I might retort not if half the population is out of work…
I am very fearful of AGI on multiple fronts, as you know. If I was going to bet on it via very long options I’d want deeply out of the money calls AND puts… 😉
p.s. This work is interesting on the prospects of an AGI revolution / impact on valuations, via the state of long-term real interest rates, albeit over a year old, prior to DeepSeek, etc:
https://basilhalperin.com/essays/agi-vs-emh.html
That’s a brilliant article @TI. Halperin is using second order thinking. I suspect the EMH is discounting near term AGI/ASI, whether it’s aligned or not, and suspect also that Mr Market is right to do so.
But….
I’m trying to think in bets and prepare for possibilities, rather than predicting any outcomes per se.
So while I don’t think that there will ever be a technological singularity as the optimists imagine myself, that’s just one guess by one person, and not any sort of knowledge.
Likewise, I think it more probable than not that the US is, in reality, looking quite stretched on valuations now (less than 1999 I’d conjecture, but more than in 1987 and 2007) and will, therefore, probably deliver rather disappointing returns over the next decade or so, perhaps especially compared to say fixed income, with bonds quite possibly being good value now.
But again, that’s just one guess, not knowledge. I have no privileged insight whatsoever. No edge at all.
However, I still think that it might be possible to hedge out some of the extreme regret that I know that I would feel at missing out on the possibility of not having some full throttled exposure to the far right of the totally outside of the Bell curve distribution of returns that a tech singularity might perhaps bring (if it doesn’t wipe out life as we know it first).
At the same time, the proposed upside hedge (whilst a binary 100% loss or a likely multiples of capital employed gain) would not involve remaining, as I am now, still substantially committed to the US.
Indeed US exposure as a share of original capital employed (my preferred metric) would be just 5% (the option premium), as opposed to nearly 50% now through being directly invested in US large caps (via XSPX ETF and, to a lesser extent, WTEF ETF – I also have a much smaller US exposure through the Russell 2000 US Small Cap Quality ETF, RTWP – which pushes current US allocation share just over 50% overall).
Whilst America (it pains me to say this as someone with left wing sympathies) has the best companies in the world and one of the best environments for business in the world; if these were just ‘ordinary times’, and not possibly *the* hinge to history, with what might be the most impactful series of technologies that we’ve ever invented round the corner, then I’d just go with what the current valuations seem to be saying.
That’d mean, in turn, fairly agnostically allocating 35% to cap weight global equities, 35% to global AAA gov bonds hedged to sterling (or maybe 35% to long duration US Treasuries not hedged), 10% to trend following, 5% to REITS, 5% to commodities and gold (and/or gold miners), 5% to BHMG and 5% to various other ideas (e.g. special situations, especially SCV and deep value plays).
But since Chat GPT came out in 2022 the prospect that the next 5 to 20 years could be *the* hinge to history (a pivot between extinction and finally getting going rapidly up the Kardashev scale) has gone from no more than vanishingly small to actually something to seriously worry about.
I hate missing out here. That’s what you have to bear in mind.
Example: I first understood BTC in 2013, when Cyprus went belly up and people were using it to get money out. Whilst I loathed the idea of crypto instinctively (and am still strongly ideologically opposed), the potential was obvious. But it then went parabolic from $150 to $1,200. So I avoided. In any event there was no time or opportunity to act. Then Mt Gox failed. I felt smug for staying out of it all. And so I watched it fall over 2014 and the start of 2015, eventually 85%.
But when it got to about $200 to $300 in 2015 I thought to myself ‘look, it’s just completely nonsense, but why not get 10 or 20 coins anyway? You’ll probably end up getting scammed or your wallet hacked or it goes to zero, but if it should turn out to be digital gold then it will be $100,000s per coin’.
There were far fewer and much less trustworthy ‘on ramps’ back then, and so I dithered and as I procrastinated the price rose to $700 and to $1,000.
A BTC ATM appeared near the London office for work in 2016. I almost used it to try to buy some coins, but now it was $1,500, then $2,000 and then, by time I’d read the first BTC piece here on Monevator in Dec 2017, I’d given up on the idea, as it had surpassed $17,000 or so.
If I’d acted back in 2015 then I’d now have had £600k min extra after the CGT for 10 coins, roughly 400x what I would have been risking (fully expecting to loose it all) back in 2015.
It’s just risk to reward.
If the 5% allocated to paying the premium on the OTM index call option ends up getting incinerated then I won’t be too bothered providing of course that the remaining 95% that would then be in the non US ‘safer’ (or at least cheaper) assets pulls through and give a creditable (say min RPI + 3% p.a.) return.
Looks like IBkr or opening a US brokerage account – both taxable – are only ways to get hold of 5 year SPY and SPX index ETF (call or put) options in the UK.
I’ve considered LETFs like 3LUS and QQQ3 here, but the constant leverage trap could screw it all up if the index soars as hoped for but then the VIX and VIXN rise sharply as well. If it’s a tech singularity then it’s likely also I’d guess to be a choppy ride up even if the US market then reacts to it as hoped i.e. up on the chart parabolically to the right.
Sorry this is so long.
BTW love that Halperin attaches various confidence levels to his essays.
In his Annexe 1 essay (“Against using stock prices to forecast AI timelines” – NB: Open Philanthropy has the whole work as an 80 page PDF to download) I’d guess he’s being maybe just a tad too pessimistic about the utility (or futility) of trying to discern what the market collectively ‘thinks’ using stock prices.
Take Palantir as an example (disclosure: I own, and am perhaps ‘bet happy blind’ here as up 3x since the autumn).
Palantir guides for $3.75 bn in sales for 2025 and its market cap is $250 bn. That’s a Price-to-Sales of 66x.
Reverse engineering this, the market is perhaps “saying” something like that Palantir now has maybe a 20% chance of growing sales by an average 60% p.a. for a decade (for context, it’s increasing its sales growth quarter on quarter, most recently from 30% to 36% annualised), which would then be 100x growth of sales volume over 10 years, to $375 bn annually.
Its run margin now is only 20%, but its free cash flow margin is up from 40% to 45% in the most recent results (so now scoring 81 on the Rule of 40 test).
So assuming the market thinks it will eventually align GAAP profitability with FCF that would then give about $150 bn of accounting profits in 2034/5.
If it’s then selling for, say, 16x earnings at that profit level then that gives a market cap that year of nearly $2.5 tn, or about $1,000 per share.
Applying 20% odds of this happening and then using a discount rate of 6% p.a. for the risk free 10Y T-Bill rate plus, say, 1.5% on top of that (in order to allow for some further shareholder dilution) and you get back to today’s $110 price.
Now had the opportunity to go through the whole 80 pages of Halperin and Co. and it is illuminating. I don’t think the suggestion there of using inverse Treasury Bill ETFs is going to work well. The 163% gain on a 3% rise in the real rate from 2021 levels (based on an 85% chance of aligned and 15% chance of misaligned AGI/ASI) is doubtful given that the real rate has risen a good percent or so with the rate hike/Ukraine war energy disruption/Trump reflation/US debt anxiety trade and I don’t think those ETFs have shined. In any event, it’s not a huge payoff for taking a punchy directional active view on real rates. I don’t know what the margin and collateral requirements would be (maybe @ZX could advise?) but hedge funds often use the repo market for short-term borrowing, leveraging their positions significantly. For example, as of late 2023, hedge funds were exposed to over $550 bn in Treasury trades with only $10 bn of their own capital (the ‘basis trade’). If a big HF therefore used $1 bn of their capital to borrow and short sell $50 bn of long duration TIPS that would seem to me to provide a much bigger and surer payout (if the AGI/ASI realised in short term scenario plays out) than an inverse Treasury/inflation linked Treasury Bond ETF. Just a thought.