The Investor is unwell, I mean on holiday. Definitely not too drunk to write his column this week. Nuh-uh. No way, Jose. Nope.
Hi! The Accumulator here. Just covering while my good friend The Investor is having a nice rest.
OK, links is it? I’ve got loads. Because we’ve been planning this for weeks. Sure have.
Anyway, one article that sobered me up this week is a penetrating critique of defined contribution (DC) pensions written by the esteemed William Bernstein and Edward McQuarrie.
They elegantly show that most people relying on DC pensions to provide for a successful retirement need:
- Much higher savings rates than is commonly admitted
- 100% stock portfolios throughout their entire investing lives (accumulation and decumulation combined)
- A dose of luck: in the form of a benign sequence of returns and average historical return rates (Woe to thee if you’re below average.)
The savings rates required to retire on a portfolio of low-risk assets (e.g. index-linked government bonds) are just not doable for most people. From the article:
Grim indeed: using historical data, our analysis shows that not until the savings rate approaches 25% does the saver have more than a 50/50 chance of success, and to approach certainty requires savings rates in the 40% range. Lower savings rates require a market return that has seldom been on offer.
To bring savings rates down to something half manageable, it’s 100% equities all the way:
It turns out, counterintuitively, that only one maneuver improves the success rate, and that’s a 100% stock portfolio both during accumulation and retirement.
Even then you need a 20% savings rate to push down your chance of retirement ruin to 4%.
How likely is it that the majority can achieve that? We’ve known for a long time that the median UK pension pot is ridiculously underfunded. And those who struggle to save likely face bleak retirements, or a working life that stretches far into old age.
Bernstein and McQuarrie’s prescription:
The current system doesn’t need more nudges; it needs dynamite and rebuilding from the ground up on the DB [defined benefit] model.
That isn’t going to happen here. Nor in the States. Indeed, the authors’ aim seems to be to push back against libertarian forces who seek to dismantle all forms of social insurance, and leave individuals at the mercy of the market.
Whatever you think of the politics, the underlying research paper by Bernstein and McQuarrie is a clear-eyed education in investing risk. Most of all, it relentlessly strips away the many myths that comfort us when we look at a global equities returns chart and notice that it’s done pretty well for fifty years.
Have a great weekend.
From Monevator
FIRE-side chat: after the rollercoaster – Monevator
SIPPs vs ISAs: battle of the tax shelters – Monevator
From the archive-ator: Bear market recovery times – Monevator
News
Crypto ETN ban could be lifted for UK retail investors – Which
Revenge tax menaces foreign holders of US assets – FT
Pension reforms ahoy. Just what we need! – This Is Money
UK Tesla car sales down by a third: analysts stumped – Guardian
Another fintech snubs the London Stock Exchange – FT
Source: This Is Money
Tesla price plunge: a textbook case of idiosyncratic stock-risk – This Is Money
Products and services
Banking switch offers are hot right now – Be Clever With Your Cash
Best mortgage rates for first-time buyers – This Is Money
Hack: How to ‘spend’ on your debit card without spending – Be Clever With Your Cash
Avoid these travel insurance nightmares – Which
UK property hotspots – This Is Money
WASPI women: watch out for scam websites – Guardian
Get up to £1,500 cashback when you transfer your cash and/or investments to Charles Stanley Direct through this link. Terms apply – Charles Stanley
Care-home fee black spots – This Is Money
Get up to £100 as a welcome bonus when you open a new account with InvestEngine via our link. (Minimum deposit of £100, T&Cs apply. Capital at risk) – InvestEngine
Nintendo Switch 2 review – IGN
Homes for sale in cultural hotspots, in pictures – Guardian
Comment and opinion
US safe-haven status in peril – Paul Krugman
FIRE sceptic rethinks their biases – Morningstar
How to avoid the big investing mistakes – Behavioural Investment
How to rationalise dreadful investment losses [Satirical] – Acadian
The UK doesn’t have a productivity puzzle – FT
Investors do better in target-date funds – Morningstar
Does small cap value improve your safe withdrawal rate? [Plus ERE vs The Golden Butterfly portfolio] – Early Retirement Now
Value is working quite nicely outside of the US – Morningstar
Sage investment wisdom from Benjamin Graham x Jason Zweig – TEBI
Common FIRE traps not to fall into – The Purpose Code
The dangers of home bias versus the UK growth agenda – Archie Hall
Choosing where to live after financial independence [Slides – US but translates] – Harry Sit [Video version – Harry Sit via Bogleheads]
Naughty corner: Active antics
To earn the big bucks you’ve got to take the big losses [Research paper] – Morgan Stanley
Using ChatGPT to optimise your trading strategy – Quantpedia
Don’t bet against AI stocks say Wall Street analysts – Sherwood
Hedging AI risk – AWOCS
Life is harsh (and short) for underperforming funds – Jeffrey Ptak
Bitcoin ETFs are up! – Humble Dollar
Pudgy Penguin NFT ETF = End Times – FT
If you like risk, you’ll love Bitcoin treasuries – This Is Money
Kindle book bargains
How to Own the World by Andrew Craig – £0.99 on Kindle
The Algebra of Wealth by Scott Galloway – £0.99 on Kindle
The Big Short by Michael Lewis – £0.99 on Kindle
Skunk Works: A Memoir of My Years at Lockheed by Ben Rich – £0.99 on Kindle
Or pick up one of the all-time great investing classics – Monevator shop
Environmental factors
Green-hushing: ESG survival strategy in the Age of Trump – Semafor
Why batteries make a renewables-powered energy grid affordable [US but translates] – Construction Physics
Hybrid electric vehicle sales rocket in the US – Sherwood
Robot overlord roundup
How AI is infiltrating the movies – Vulture
Why AI isn’t leading to mass sackings (yet) – Dwarkesh
Not at the dinner table
Trump vs Musk: Gasbags at dawn – CNN
Apparently we’re at war with Russia – Guardian
Reaction to the UK Strategic Defence Review [Podcast] – Chatham House
Trump family get into bed with crypto bros [Voms] – WSJ
Off our beat
How Ukraine’s audacious drone attack stuck it up Putin’s bombers – CSIS
The genius myth [Paywall] – Atlantic
Contrarian views on the big five mass extinctions [Paywall] – New Scientist
And finally…
“Owning the stock market over the long term is a winner’s game, but attempting to beat the market is a loser’s game.”
– Jack Bogle, The Little Book of Common Sense Investing
Like these links? Subscribe to get them every Saturday. Note this article includes affiliate links, such as from Amazon and Interactive Investor.
McQuarrie 05/2025 paper (pp. 10-11): “use
two benchmarks: the 6.6% average real rate of return on stock investments in the US over the past two centuries, and the 4.4% average real rate on World ex-USA equities since 1900, per Siegel (1994 / 2022) and Dimson, Marsh and Staunton (2025).”
By those metrics it looks grim.
But there may be a third way, beyond increasing your savings rate and / or accumulation period and / or decreasing your consumption and / or your time in decumulation: namely using part of your asset allocation to play ‘Bessembinder roulette’.
The only way to achieve a higher return than the market is to assume some concentration risk (or to use leverage, but God only help you doing that); but not so much that it ruins you if/when it fails.
As Bessembinder shows in his studies: 50% of excess returns (equities over T-bills) comes from just 0.25%/0.3% of all listed firms.
Finding one of those is like finding hen’s teeth.
But if you do…
So say put 80% of your portfolio and your future investing into VDEV/VWRL/Vanguard Global All Cap, and the last 20% into the best, most thoroughly researched, long term buy and hold idea that you’ve got for a life changing single stock name.
If you lose on the single name then you’ve still got 80% of what you would have had you been 100% passive.
Obviously not great, but no disaster either.
And if you win….
Since IPO in 1999, Nvidia has had a total return of around 3,700x; and since 2006 it’s around 329x.
While predicting Nvidia’s ascent to an AI behemoth from the vantage point of its IPO would be an act of sheer prophecy, by 2006, the picture had changed dramatically with the launch of CUDA (Compute Unified Device Architecture), such that a thoughtful and technically-astute observer could have reasonably foreseen that Nvidia was positioning itself for a dominant role in the future of high-performance computing and, by extension, the nascent field of machine learning.
Analysing DC schemes and contribution rates and concluding they need to be well funded to be successful is no great surprise. The industry has been banging that tub for years, including legislatively here in the UK via schemes like NEST. But then leaping to conclude DB is the way forward isn’t logical to me.
The effect of tax breaks and policy were entirely ignored also. Of course people want the assurance of DB but the whole point of why DB was abandoned over DC is the cost – money doesn’t magically appear out of nowhere. There is no free lunch. Someone has to take risk somewhere. Choose how to pool that, sure. But the benefit level needs to fundamentally linked to personal contribution level over the person’s entire lifetime to avoid imbalance. Trying to escape that invariably leads to other problems.
I found your FIRE interview on the Australian superannuation system most illuminating and sensible in that regard. It tries to strike a better balance between personal contribution and wealth, while still providing a tapered minimal backstop.
@ The Accumulator – congrats on the promotion for the weekend and your pick of the links.Strangely in among the heavier articles i found the Targeted dated story the most interesting,the last time i looked a couple of years back it seemed only Vanguard offered them via diy platforms but may have changed.
While pension funds offer their own glide type products etc it seems that target dates have not really caught on over here.Even Fidelity who are One of the biggest providers of them in the States seem to prefer adding the word retirement to their Multi Asset Allocator Growth fund and having Two identical funds for some reason.
@TA:
Interesting paper from William Bernstein and Edward McQuarrie. However, it is not really all that surprising for the reasons given below.
B & McQ make two critical assumptions, namely that: a) DC is the only source of retiree income; and b) retired lifestyle costs are same as before retirement minus the cost of DC plan contributions.
Back in 2019, Pfau, Tomlinson, et al in the Stanford Spend Safely in Retirement Study (SSiRS) concluded that if middle income workers optimise US social security (ie delay taking it) it could provide them with between two thirds and 80% of their retired income needs. So removal of SS (as assumed by B & MCQ) will have a huge impact; and indeed it does as B & McQ demonstrate over some 100 plus pages.
For me, the most eye catching figure in all of this has always been the relative generosity of US social security!
Loved the ‘Common FIRE traps not to fall into’ – The Purpose Code article – couldn’t agree more.
I’ve always been more interested in the FI than the RE.
To call for DB schemes is to call for somebody to pay a subsidy. Who?
Since “the State” is broke the subsidy could come only from cutting other expenditure. What?
Do Americans do things differently to us?
Say they leave university at 21, this presumes retiring at 51. I had a 30 year working life, but I consider that I retired early. A conventional working life is much closer to 40 years than 30. If you want to retire after working a quarter less than most people, it stands to reason you have to do something different to the norm, and yes, they have summarised that reasonably well. I’m not under the impression that Americans generally retire earlier than us.
@dearieme > To call for DB schemes is to call for somebody to pay a subsidy. Who?
It doesn’t have to be the State.
In your case I believe, and in mine, DB is deferred pay from while working.
I knew we were going to have a good discussion about this 🙂
Re: cost and subsidies – When every individual must bear the cost of being unlucky then the bar is set very high to avoid falling into that trap. But a collective approach smoothes away the outliers.
The extremely lucky subsidise the extremely unlucky. Everyone can enjoy a reasonable outcome. We do this all the time. Home insurance, NHS, State Pension. Japan does this for social care.
In the case of an annuity, the subsidy works in reverse. The extremely unlucky (those who die early) subsidise the extremely lucky (those who live to a ripe old age).
@John Charity Spring – all good points but despite tax breaks and auto-enrolment, individual risk remains and can be better solved by a collective approach. I agree the Australian system seems to strike a better balance.
DB pensions allow for the impact of personal contribution. The State Pension does too. You could also still include a DC component for the well funded.
@Al Cam – at some point your retirement expenses are going to be your working age income minus savings. The more you must save – or the less you earn – the more likely that is. I don’t think the authors’ concern is for the well paid or those receiving higher rate tax reliefs. It’s also worth bearing in mind that recommendations to delay US social security max out at age 70. In other words, it’s a manoeuvre more easily afforded by the well-resourced and healthy. People with good options.
I’m not sure how our State Pension compares with US social security? I keep stumbling on articles suggesting the UK is about as miserly as it gets among our peers (or wished for peers).
@Howard – that’s essentially what happens I think. People take big bets. Often they don’t pay off. Situation is worse. It looks like there’s a better way.
@klj – ha, ha – thanks. I’m hoping to get demoted again in time for next week 😉 Yes, you’re right, Target date products do seems to be largely confined to pension schemes in the UK. A few Target Date ETFs have emerged, too – in Europe and the States – but not here as of yet. I do think they’re an excellent idea for anyone who just doesn’t care about investing, or doesn’t trust themselves.
@Ermine – perhaps 30-years is a good average for the accumulation years?
Lots of people likely don’t start saving for retirement until nearer 30 (or beyond!). Many are probably in insecure employment for a long time. Perhaps don’t have access to pension schemes.
Others are in grad school etc so perhaps don’t start earning until age 27 or 28. Many women will take time out to have kids. Some will suffer bouts of unemployment. Finally, a substantial slice are probably forced out of work early due to ill-health – their own or a family member’s.
Every study shows that 100% diversified stock portfolios always do better than anything else over 25-30 years, bonds can just reduce volatility. Consequently it always has made perfect sense to have a 100% stocks allocation throughout.
And the “attitude towards risk” question always asked is somewhat stupid. We all want the same from long term investments, a fair return with minimal chance of failure.
A diversified 100% stock portfolio has always been the way to go, why I haven’t done that is another matter.
I remember being told by an HR bloke long ago that I needn’t worry about pensions until I was 38 !
Of course that was in the times of dB pensions and pre-Gordon Brown.
I always wondered how you could work and earn money for say 45 years and somehow save enough to fund a further 20 odd years of leisure. Now we know you can’t really.
Of course if you are employed by the state and have a dB pension you might think not your problem.
However if you have a clear disparity between comfortable retirees living off the unfunded civil service pensions, a group of retirees struggling on dc schemes, plus workers paying taxes to sustain both, you might wonder if there is a fight coming.
@ Mr O
Perhaps it was also in the times you could put in over £200k a year, so could just stuff your pension in your final few years of work when hopefully children had left home and had fewer overheads. They used to think about these things and try to help people.
@TA (#8):
Try: https://researchbriefings.files.parliament.uk/documents/SN00290/SN00290.pdf
AFAICT, the paper’s primary purpose (although IIRC not explicitly stated) is do not mess with US SS – IMO the rest is just noise around the edges.
The current UK SP system is effectively flat rate and means tested – ie you get the same out no matter how much you pay in, which IMO makes it pretty mean. Prior to 2016 there was at least an element that was earnings related. Anyway, the current UK SP system has been around for some ten years now and finally people are slowly waking up to what it is – must therefore be about time for another change that will no doubt also promise far more then it ever can deliver!
@Mr O (#11):
Re: “However if you have a clear disparity between ….”
Maybe, but it has been a heck of a long time coming.
My (largely, but not entirely, tongue in cheek) proposal is that the next set of pension reforms start with MP’s pensions ONLY and that the changes be put to a national referendum that can only proceed if a clear majority (say 60%) of the population eligible to vote (ie not just those who can be bothered to vote) agree; otherwise MPs pensions stop entirely! That might help focus some minds?
As an optimiser, I often find myself spending far to long worrying about the intricacies of my own asset allocation. Currently I find myself sitting on an 80/20 allocation for accumulation, and 60/40 for deaccumulation in my investment plan.
In the past year or so, that whole debate seems to have revolved around the Scott Cederburg paper “Beyond the status quo” which the Bernstein link refers back to. Would be very interesting to know Monevator’s thoughts on that one, as correct me if I’m wrong, but I don’t think I’ve ever seen a standalone article on this site discussing that paper, and it’s recent revisions.
It seems like some very big online financial personalities are now backing its findings (think the likes of Ben Felix and Pensioncraft who seem to now favour 100% equity allocations) and the results are also backed up by the Wafe Pfau paper referenced in the recent UK SWR series you’ve been running.
Will there come a point soon where we feel forced to ditch the bonds? I suppose the reasons for holding you could say are to reduce volatility and the risk of selling in a downturn (ie addressing investor behaviour risk), but I’m more interested the benefits bonds can have through reduction in sequence of returns risk. I wonder if UK history can’t tell us the whole story here, but we have to in fact rely on block bootstrap and monte carlo simulations to assess whether there are any benefits to holding bonds from a purely mathematical point of view. It’s interesting that the US data shows a 75/25 to be more optimal compared with 100% for the UK, but I guess that is down to the higher UK inflation in the past, favouring a 100% strategy.
I suppose ultimately thoughts of this kind are meaningless if you have a large enough pot. A 2-3% SWR is likely going to be safe no matter what asset allocation you choose. Maybe with longer retirement horizons though (think 40-50 years for the FIRE community) it does make sense going 100% if inflation then becomes the biggest threat after sequence of returns risk is dealt with.
@Mr O #11. “Unfunded civil service pensions”. Why would the government pay an amount every month into a fund which has to be managed when it can simply not pay anything at all for the entire working life of the civil servant until their pension begins payment? Taking this enormous cash flow advantage is an obvious measure for government which can simply pay its long deferred liability out of regular taxation, not an avenue readily open to other employers. It would also have had to pay the nominal 8% deduction made in arriving at pay scales that I was told about when I signed up as a civil servant in 1977. By the time I left, I was having an increasingly large slice taken from my salary towards the pension which, given there was no fund for it to go into, was simply an extra tax just for civil servants. I’m not moaning though, just giving an alternative view of the matter.
@JCS (#2):
I asked @LALTA for some further details – which she has kindly provided at her fireside chat post. Under the hood, the Aus system looks suitably complex to me. So without a lot more work I conclude it is different to the UK system. Also, I suspect that the B & McQ paper may be more appropriate to Aus than either UK or US.
I’m sorry to be that guy TA but….
“ How to How the World by Andrew Craig”
DB is still widely available in the public sector. I’ll wait for the stampede of applicants…
For what it’s worth I did a stint in the public sector to build up a baseline of 10k per year DB and now enjoy twice the salary but a less generous pension in the private sector. I feel that kind of combination might work for many people as the ‘pseudo bond’ part of their portfolio, allowing them (us) to go 100% equities in DC part whilst losing less sleep.
The fly in the ointment as always is investor behaviour especially during periods of stockmarket downturns
Even though much more is known in this area now it doesn’t seem to change stockmarket volatility or investors behaviour very much
A key if not the key for investors on the path to successful investing is to ascertain as soon as possible their personal stomach acid levels/sleepless nights abilities ie % of equities in the portfolio -often/usually acquired after experiencing their first stockmarket drop with a reasonably large amount of money saved/at risk ?
(This comes before the next keystone of investment ie Asset Allocation which arguably is almost but not quite as important )
Bonds have classically filled the “safe” investment role in portfolios-other assets don’t seem to be able to compete for long in this area
Very few amateur investors can live with 100% equity portfolios for any length of time especially when the amount saved/at risk has become a reasonably large figure
xxd09
The Bernstein and McQuarrie paper is interesting, but (from a UK perspective) not surprising since this has been under discussion for quite some time (e.g., see https://actuaries.org.uk/thought-leadership/thought-leadership-campaigns/savings-goals-for-retirement and https://www.gov.uk/government/statistics/analysis-of-future-pension-incomes/analysis-of-future-pension-incomes).
FWIW, I’ve done calculations (as yet unpublished) on what savings are needed to provide a retirement income from a DC pension equivalent to the SP. Using 100% stocks with a fast glide to 60% in the last 4 years of accumulation, a 40 year accumulation period, and an expected inflation adjusted payout rate of 3% at the start of retirement (would be a bit better now), required savings per year ranged from about £9k in the worst historical case to £3k per year in the median case, and £1.2k in the best case – and yes, that is really a factor of 7.5 between the best and worst cases. Longer accumulation periods reduced the required savings amounts and vice versa. Lower equity allocations increased the required savings amount.
In the UK, collective defined contribution schemes (e.g., see https://commonslibrary.parliament.uk/research-briefings/cbp-8674/) are one future route since they combine aspects DC and DB pensions. If my understanding is correct, they benefit from collective mortality along cohorts (in the same way annuities and DB pensions do) but also spread investment risks between cohorts – in other words, each individual effectively ‘sees’ an investment return closer to the median, so pension income is less dependent on luck and will (probably) fall in a tighter range from ‘best’ to ‘worst’ than the factor of 7.5 mentioned above. A few already exist.
@Flying Scotsman (#15)
Looking at SWR as a function of equity allocation for individual countries (see Figure 2 in the paper at
https://www.financialplanningassociation.org/sites/default/files/2021-10/DEC10%20JFP%20Pfau%20PDF.pdf) shows that for most countries the highest SWR was not at 100% domestic equities. Using international equities as Cederburg does changes this outcome because of diversification, e.g., the US crash of 1929 was not reflected so severely in the UK stock market. However, it is not clear to me in an era of globalisation how well international diversification will hold up in the future.
I also note that all of the historical studies have, of necessity, used nominal bonds not inflation linked ones and, furthermore, usually bonds with longish maturities. There is no doubt that periods of high inflation have a severe effect on the real returns of nominal bonds even when held to maturity.
@xxd09 (#20)
It is the time scales that make it difficult psychologically for DIY investors. Bonds and cash might (or might not) usually lose purchasing power so slowly that it is not noticeable, but when equities lose purchasing power it is usually very noticeable indeed!
Alan S-indeed
Bonds reduce (a) volatility in a portfolio (except for 2022!)-enabling a DIY amateur investor to be less likely to panic and sell at the worst possible moment ie at the bottom of the market
Half of my income at the start of my retirement came from an investment portfolio-30/70 let me cope-currently 22 years into retirement now at 35/70
(b) appear to preserve wealth in the face of stockmarket shenanigans although preservation might be more apparent than real especially over the long term
(c) Hopefully actually grow in value a little
Did the job for me-so far ……maybe I had saved a lot,live quietly and have very low investment costs-at the end of the day it’s all very personal
xxd09
@TA #8: “people take big bets”: there’s other outperform attempt alternatives to the Bessembinder lottery (i.e. trying to select the one in every 300 or 400 super winners, so 250 to 300 out of 100,00 listed companies globally). One system with no skill, no discretion and no market timing involves concentrating capital solely in the single largest stock by market capitalisation worldwide. Historically, this position has shifted between General Electric, Microsoft, ExxonMobil prior to 2012, and Apple and Nvidia more recently. An annual check of the largest cap leader would have resulted in 9 round trip trades since 1997. As this is the largest company worldwide spreads are miniscule. For UK investors allow 0.5% FX costs for each round trip (0.25% per limb). Over 30 years has delivered 22% CAGR and since 1997 20% CAGR compared in both cases to 10% CAGR for the S&P 500 (pre FX costs). Since 1997 a 60% largest cap to 40% S&P 500 tracker annually rebalanced would have given 16% CAGR at no effort or skill.
@Flying Scotsman – We’ve always known that the biggest spoils go to 100% equity portfolios but that few can live with the psychological pain that it takes to achieve them.
Think of Japanese equity investors or those articles about how much you would have made on Amazon if you’d invested in 1997 (and, oops, we forgot to mention, if you’d been able to withstand multiple 90% drawdowns along the way).
Given the disturbance in the Force whenever stocks go down 20% or so, I suspect that many holdl types could not handle their 100% stock portfolio being crushed 80%.
There’s always the risk of regime change, too. There’s a great paper somewhere on how bonds have outperformed equities for long periods in the past (pre-20th Century).
It is the case that you get higher US SWRs using US bonds – but those studies only count the decumulation phase. Cederburg et al get 100% portfolios by analysing the entire lifecycle (accumulation + decumulation).
To your excellent point about how the calculus changes if you are well funded, it was Bernstein who said, “If you’ve won the game, stop playing.” The paper implies that ideally you’d be sitting pretty in inflation-linked bonds in retirement. I think they’re drawing attention to the impossibility of that happy outcome for most.
@Alan S – great post! Will check out those links.
@No Longer Civil – It makes me wonder where we’d stand if we’d had slightly higher taxes and set some money aside in a sovereign wealth fund.
@Howard – do you follow this system?
It’s a big claim so I think that requires me to ask questions like:
Are the numbers reliable? Have you checked them yourself?
Do you know why institutional investors haven’t arbitraged the effect away?
Does it hold true over 100 years? [28 years really isn’t long enough. Do you know why 1997 was chosen as the start date for the comparison? Whoever came up with this could easily go much further back than that.]
Does it hold true in multiple markets? [Perhaps it’s a data artefact that’s unique to the US market over this specific period.]
Are there periods when the effect disappears? How long do those periods last?
Do the numbers rely on market timing? i.e. the ability to dump the leader at the point of transition?
If this is true then why don’t most active investors beat the market over time? I mean all they have to do is…
Removal of DB pensions happened with little or no compensation which was effectively a salary cut. (Obv it was done for the sole purpose of saving the companies money.) Owners of capital pocketed the difference and dumped the problem of funding retirement on the individuals and society.
This is part of a larger trend in recent decades, global capital has won and labour in the West has lost out. Companies and capital make rampant use of tax havens, labour remains the only goose left to pluck.
Unless we somehow reverse this trend, I think we’ll get ever nastier politics (Brexit and Trump 1 should have been enough of a warning).
Meanwhile, people need to acknowledge that they were made poorer, and either save more or spend less in retirement. The Bernstein paper uses 100% stocks as a scenario but they don’t recommend this – they call it “insanely risky”.
@26 “ Removal of DB pensions happened with little or no compensation which was effectively a salary cut. (Obv it was done for the sole purpose of saving the companies money.) Owners of capital pocketed the difference and dumped the problem of funding retirement on the individuals and society.”
This was sold to the company I worked for (circa 1984) as “granting you freedom to manage your own pension.” Then around 1989 came “personal pensions” as a thing. No one really explained, least of all the “pension consultants “ management brought in, that it was a neat transfer of risk.
@TA/#25: They’re all fair & good questions.
Numbers are reliable.
The effect is stronger in recent times (so cherry picking, but recent is arguably more relevant, even if 30 years is not as good as 200), and, yes, because the largest company has been US during that time, it’s market specific.
It can’t be arbitraged IMHO any more than the SPIVA outperformance can be because it relies on the same cap weight principle of trackers just taken to a logical conclusion – the market thinks X is the best company by giving it the highest valuation, therefore buy X as long as that holds true.
Like indexing, it’s self reinforcing (feedback) rather than self correcting (reversion).
It would work better doing it at the switch in leadership, but annual suffices and keeps turnover / costs to not much more than with the annual rebalance between the tracker and largest company.
Versions work with more than 1 company so that aspect’s not overfitted – but why complicate?
Obviously noone will let or pay for a fund manager / institutional fiduciary to hold just 1 company. In a sense that’s another advantage (for retail).
But how much risk is there over and above indexing by holding the largest component at overweight? If it stumbles then it gets replaced by the next largest, which then becomes the largest, just like in indexing (and then that company becomes the one to own).
And if the big companies in the S&P are falling in value then the index will likely be taking a bath too given current concentration in the largest caps.
Sometimes the best ideas are the simplest. I want a zero skill, low and infrequent effort system which an idiot could follow.
Playing devils advocate- if life expectancy is increasing and legislation and scheme rules become more generous then the “same” DB benefits are increasing in cost and value. Changing them would be necessary to costs stable.
Reducing or removing benefits would become essential to remain a viable and healthy company v new start up who would be unencumbered by the “old world” traditional benefits.
Sort of makes you wonder if public sector DB schemes are killing the UK’s future. We live is globalised world, something has to give to remain competitive.
In a coincidence if you have a reading app such as Readly this weeks issue of the NZ Listener has a comparison between the Australian & NZ superannuation plans
@Howard – The no.1 company in an index by market cap doesn’t necessarily equal best performer. Think of a company that grows until it’s on the verge of usurping the market leader. You miss out on all that return because you’re holding the existing large cap king of the jungle. Meanwhile, that’s growing more slowly (it doesn’t even have to be losing money) and we know that for a fact because its lead is being erased.
We could imagine this cycle repeating. The strategy requires switching to the market cap leader after it’s risen to the top, just as it begins its decline. So there must be periods when it doesn’t work.
If it’s true that the index leader outperforms the rest of the index then this shouldn’t be market specific. It should work across all markets.
But the effect could easily be arbitraged away. If it’s a slam dunk that the index leader outperforms the index then everyone piles into the leader. At some point, its price is forced so high that late entrants cannot obtain a worthwhile return on investment. The stock’s price flattens out then declines – the holders lose, undoing the effect.
Why wouldn’t institutions sit in a single stock if there was no risk? Or even if that single stock was less risky than the rest of the index? It would be the right thing to do. If you were bound by fiduciary duty then you would be doing the right thing.
Aside from that there is plenty of private money, hedge funds, single-stock ETFs and other vehicles that aren’t bound by the restrictions of a fund marketed at retail. They would just eat up this trade. They would be idiots to do anything else, if it automatically worked.
The market is incredibly well informed on aggregate. That’s why it is so hard to beat. The market doesn’t believe in this trade. Listen to the market. You can’t beat it with zero skill.
@Boltt – some (I don’t know about all) public sector DB schemes have been made less generous over time e.g. larger contributions required, later payouts, less generous payouts.
Life expectancy also declining in the UK… yay, us.
But I do agree that mass DB pensions are risky for individual companies. That said, insurance firms manage it with annuities. And the State is another beast entirely.
@klj – which is best?
@TA #31: Framing the issue matters.
If you believe in strong EMH (the market price and weight is always right) as a matter of faith, then every outperforming strategy is either a data artifact mirage or a soon to be arbitraged anomaly. $100 notes seen on sidewalks can’t be real because they have been picked up already.
But strong EMH is a model designed to shoehorn evidence and logic into its assumptions (of perfect markets, without regulatory or institutional barriers, with exclusively wholly rational actors with information symmetry).
Loosen the constraints and things get more interesting.
Markets aren’t perfect. Institutions can’t invest as they please. A HF PM going all in a single stock would get fired – assuming compliance even let the trade execute (Ackman has 25% of PSH in UMG IIRC, but that ain’t 100%).
And given their mandate they’d be right to sack him/her.
Any single stock portfolio is going to show terrible risk adjusted metrics. Volatility will always be way higher (typically double, but it varies) the index. There will also be idiosyncratic risk.
That’s why concentrated portfolios show higher returns. They have more risk, at least measured by volatility.
But we don’t eat risk adjusted returns. We eat CAGR over the investing horizon.
Plenty of zero skill portfolios outperform on the CAGR basis.
Low volatility anomaly persists. No skill required, and the anomaly (unlike other risk factors) does not appear to be compensation for extra risk.
Momentum carries more risk than cap weight, but also outperforms over the whole Fama-French dataset period. Skill required nil. What could be dumber than buying something just because it’s gone up. But it works (granted with periods of underperforming).
Cap weight indexing itself is also an excellent zero skill approach which outperforms active fund managers as a group (even more so allowing for the closet indexers). Maintain market weights regardless of price is another dumb sounding simple strategy that also just works.
And they’re plenty of others.
Take this weekend’s reading.
Michael Mauboussin of Morgan Stanley May 2025 report, at p. 5 and exhibits 4 and 5.
Of the 1842 stocks that dropped more than 95% grom 1985 to 2024 only 16% made it back to ATHs.
But the median 10 year total shareholder return on them once they had first drawn down 95% was a staggering 32.6% annualised (at which p.a. rate a stock doubles in less than 2.5 years, and does over 16x in 10).
Simple and dumb (especially given that so many stocks falling 95% went on to fall another 99 to 100 percent). But again it ‘works’.
And as with the single stock it works because it gives a reward for the risk which exceeds the risk.
Maybe not a $100 bill on the sidewalk, but possibly a $20 one.
And, in fairness, the largest stock by capitalisation in the world is not going to be the riskiest, or even likely to be of average risk, at least in terms of sudden, unexpected and permanent loss of capital.
Every single large cap analyst on Earth will have run their side rule over it.
The market will have decided in its dollar weighted wisdom to give it the largest allocation.
This idea of mixing on overweight to the market weight for the market leader alongside a core index holding just leans into that, albeit heavily.
And the effect is seen for the US largest company only simply because for the past 30 plus years the largest company globally has been a US one.
@The Accumulator – May be partly based on the writers political views (but i feel not due to different governments) that due to mismanagement and changes to the rules etc over more then a few decades that there is not enough money to pay out the promised amount at the age agreed and a Black hole is building etc. that it’s a clear win for the Aussies.
@Howard – Thanks for engaging. I’m going to continue because this is such a big claim and I’m intrigued. In fact, if it works, then I want a piece of it.
The claim is the index leader strategy works and can’t be arbitraged. The lack of arbitrage is huge if true.
The factor strategies you mention e.g. low vol and momentum are different because:
It’s commonly accepted they can be arbitraged. Typically the premiums have decayed in the aftermath of research publication. Credible analysts don’t claim they will always persist, simply that there are reasons to believe that they may do so.
Factors or pricing anomalies that persist require explanation if we’re to have any confidence they’re more than a data mining artefact.
Typically that comes in the form of a risk-based or behavioural thesis. If the explanation is risk-based then we accept that the risk may not pay off, or may not pay off for decades. I’m not sure, but you may accept this is true of the index leader thesis?
If the explanation is behavioural then we have to accept that the anomaly can be corrected once known, and arbitrage is more likely.
We can have more confidence in the thesis if the strategy persists across multiple markets i.e. out-of-sample data increases the probability we’re not seeing a pattern that’s occurred by chance.
Naturally it helps to have a long-term track record. 60-years is the shortest I’ve seen for any factor that survives the charge of data-mining.
The thesis here is that the market leader outperforms the rest of the market during the period in which it occupies the no.1 spot. (This is the opposite of “buy low, sell high”. It’s a “buy high, sell higher” strategy.)
The idea being that because the stock has been good enough to get to number one, its performance thereafter will smoke the rest of the index. In fact, its over performance will be good enough to beat the rest of its index plus its inevitable performance drag when its lead is being eroded by the no.2 stock.
However, the effect has only been identified in one market? For a short period of time? I’m personally worried that without more evidence then that’s a data mining artefact.
If this works then it should be well researched. Academics compete to identify and publish new factors and price anomalies.
The only research I can find (they describe it as a Top Dog strategy) shows the cap-weighted World portfolio outperforms the Global Top Dog (1980 to 2017). See Figure 4: https://www.researchaffiliates.com/publications/articles/674-buy-high-and-sell-low-with-index-funds
The whole paper is worth reading, though.
I can also find a Reddit thread in which the OP calculates a win for the Top Dog vs the S&P 500 1990-2024.
https://www.reddit.com/r/investing/comments/1e5k8n6/would_an_all_in_fund_that_only_ever_held_the/
Unfortunately, the OP calculates price return not total return.
Someone links to a piece citing CRSP and Morgan Stanley Research:
“First, the top stock has historically been a bad investment. Specifically, the arithmetic average of the series of annual returns of the top stock relative to the S&P 500 from 1950 to 2023 was -1.9 percent. The geometric return of the series was -4.3 percent, reflecting the fact that the series was volatile. — Morgan Stanley Insights
Ned Davis Research also had similar findings that showed that from 1972 to 2013, the S&P 500 was up close to 5,000%, but if you had owned just the biggest stock in the index every year, you would have only gained around 400%.”
https://www.marketsentiment.co/p/should-you-invest-in-the-biggest
Again, if this were a no-brainer, then there are plenty of vehicles that would hoover that up. Compliance departments prevent over-concentration because it’s not a no-brainer.
You obviously know your stuff so I’m wondering what’s giving you so much confidence in this? Especially when other factors are available. I’m more than happy to read some research that supports the strategy.
@klj – Cheers! Must look into this sometime.
@TA (#34)
I think this is this Morgan Stanley paper (Exhibit 10) you found referred to
https://www.morganstanley.com/im/publication/insights/articles/article_stockmarketconcentration.pdf
The paragraph under that graph is worth quoting, “Exhibit 11 reveals the second noteworthy finding: the top 3 stocks produced markedly better relative returns from the end of 2013 through 2023 than they did in the past. […] The arithmetic average annual excess return was 15.9 percentage points for owning
the largest stock, 9.8 percentage points for the second largest stock, and 8.4 percentage points for the third
largest stock over this time. The corresponding excess geometric returns were 14.2, 7.5, and 5.3 percentage
points.”
In other words, there may be some (recency?) bias if calculations only go back 20 or 30 years.
While not restricted to the Number One stock, the ‘Nifty 50’ that was popular in the US in the 1960s (not to be confused with the Indian index of the same name) may be a salutary lesson in limiting stock choice to ‘popular’ stocks. While not strictly not a top-30 index, the Dow Jones index can also form a useful comparison with the SP500 index (which of course didn’t always contain 500 stocks) – FWIW, from 1958 to 2025, the annualised returns narrowly favour the SP500 (7.75% to 7.06% according to https://www.measuringworth.com/calculators/DJIA_SP_NASDAQ/).
Thinking on all this, I am sure I have read many articles exploring the maximum swr over time and that a stock allocation between 30% and 80% would do just fine. Am I wrong ?
Getting political, am I the only one sensing this government will go after ‘unearned income’ , to use a phrase from the 1970’s ? A ‘special’ tax on pensions above a certain limit might broadly achieve the same as a lifetime allowance scheme, which I feel they would be loath to reintroduce.
Apols for the delay in replying. Work et al.
The Market Sentiment piece seems to be premised upon the investor hanging on to whichever stock was #1 by cap weight even after it has ceased to be #1. That’s a different strategy.
Ned Davies Research is informative for the period from the crash of 1972-4, and the decline of the Nifty Fifty, through to the beginning of sustained market cap leader outperformance in 2013.
If one thinks the future is going to be like 1972-2013 then sure that suggests avoiding the biggest stocks.
But, as the 2024 Morgan Stanley paper puts it, since 2013 the position has been markedly different: “Owning the stock of the largest company has historically been a poor investment relative to the market overall until about a decade ago. Owning the second and third largest stocks produced excess returns from 1950 to
2023. However, all three of the top stocks have provided stellar relative returns from 2014 to 2023. Where we
go from here is anyone’s guess, but assessments of sustainable competitive advantage and growth will be central to determining that path”.
As a side note, it’s worth contemplating here the recent effect of tech sector concentration on excess returns to cap overweight portfolios, and whether this phenomenon is cyclic or structural.
I personally think indexing is starting to have an effect on market dynamics including liquidity and volatility – so this *could* be structural given that indexation is largely a post 2013 phenomenon in terms of market share.
Maybe coincidence. But possibly causative or at least contributory.
Yes, the market leader approach (for want of a better description) definitely shows recency effects.
Whether that is a bias though depends upon your epistemic preference: should one equal weight all past periods or overweight recent periods (because recent information about performance may be more relevant than older).
I prefer the latter approach personally, but YMMV.
It is fair to point out that 1997-2025 includes both the Dot.com crash and 2008/9, so there’s perhaps some robustness across cycles.
Could it be a data artifact. Sure could. Might it be worth exploring more. Perhaps. Would I dismiss it a priori on the basis only of the EMH (likewise also for any other systemic idea): personally, no I wouldn’t.
Absolutely agree every idea, Fama-French risk factors included, deteriorates post publication. No shock there.
But sometimes ideas go out of fashion, and when matters go to an extreme they outperform.
Value (HML), ex US DM, EM and Small cap (LMS) have all underperformed for ages relative to US Large Growth and Momentum (WML).
But there’s still a T Stat significant anomaly over the whole (1928 to date) period in the S&P 90 (1928-57) and 500 (1957 onward) data.
For Low Vol it’s uniquely difficult to identify a risk which the performance compensates for (therefore possibly a second free lunch, alongside diversification).
I love cap weight indexing but I don’t allocate it to a special category distinct from any other rules based approach.
It’s not for me philosophically either cap weight or the highway.
It can be cap weight alongside multiple other ideas.
Find patterns and capitalise on them until that pond dries up then find and adopt new frameworks.
Cap weight has worked historically, but not necessarily better than other rules based approaches. It depends on the approach and the period.
What it does do is ensure you get the total market return less fees.
But, in doing so, it definitionally ensures that you won’t get a higher return than the index.
That won’t always and everywhere be enough to meet everybody’s needs.
For some, taking more risk than the market or a different approach is the only realistic prospect to hitting their requirements.
@ Mr O
It does feel like “all bets are off” territory. I’m struggling why they can’t understand incentives. Non-workers need to be materially worse off than full time min-wage workers, otherwise what’s the point.
I’m personally aware of a millionaire who bought a 4 year old 7 seat BMW with over 100k miles on the clock – the first owner was a motability customer. It just feels wrong.
Having clearly declared which side of the fence I’m on, I see much logic in equalising tax on income, dividends, gains, etc. And the time has come to merge tax and NI at around 25%. That will the pensioners, which seems to be their objective – it will cost me but it seems sensible. As does no new ISA subscriptions if your balance is over £200k
I hoped Labour had spent their opposition years coming up with some strategy and innovative thinking, but it seems complaining was their main pastime.
Grumpy
@Mr Optimistic – The paper quoted is focused on median (and below) investors across the entire lifecycle – accumulation and decumulation combined.
If you’ve already hit your number then SWR strategies come into play. In that case, bonds can help but it depends on the length of the retirement. Alan S linked to this article earlier in the thread: https://www.financialplanningassociation.org/sites/default/files/2021-10/DEC10%20JFP%20Pfau%20PDF.pdf
See figure 2, page 8. 30-80% equities is the US experience.
@Howard – in the absence of robust evidence for me to get my teeth into I’m going to wish you the best of luck with it. I genuinely hope we meet on a thread somewhere in 20-years time and you can say, “Y’see! I told you I was right. I’ve smashed the market.” And in reply I’ll be able to say: “Me too, check out my small value returns.” 🙂
More generally, it’s a sad truism that those who most need to take risk can least afford to bear the cost if things go wrong.
Hence if you *need* to beat the market then wiser heads than me advise some combination of:
Work a little longer,
Save a little more,
Or live on a little less,
If you possibly can.
But if you have no edge, and take the risk of trying to beat the market anyway, then look for:
– Credible independent research that supports the strategy – including a clear underlying rationale for the outperformance.
– Evidence it works in multiple markets and across multiple timeframes.
– Understand the risks.
– Beware wishful thinking.
Just to weigh in on the pension discussion, I always find it a big red flag when someone pulls out graphs and individual investment advice to explain (multi-)national macroeconomics.
On a fundamental level, all a national pension strategy does is to decide how much of their economic output the working population redirects to pensioners, and how that output is spread around. A national system like a 20% tax that gets spent evenly across all pensioners does that very obviously. More complicated schemes all the way up to DC pensions do the same thing, just murkier and murkier. What none of them do is magic up a baker who makes twice as much bread just because someone found the crazy returns button. While the best case is that linking pensions to stock market gains, bonds or beanie babies will simply mean those returns get damped down or inflated away, the sad reality is that most likely the lofty more-for-less promise will be kept for whoever proposed the scheme while those lucky few of us with the time and means will put the tedious work in to keep up with yet another investment pot and ordinary people who just want to work and retire someday bear the brunt of it.
So whenever someone claims something like “The effective return on this DB scheme is only three percent while that ETF-backed DC will yield 7%”, my mind files them into the same category as people who claim no one will ever have to work again if we only mint a $1 quadrillion crypto coin and take that to the supermarket.
@TA @Alan S thanks for that link. Nicely presented paper.
Looks like UK experience is more sensitive to stock allocation and, historically, less than 70% was below par.
Cheers.
TI must have a hell of a hangover if he is still in bed on Wednesday.
This early (2011) paper from Pfau may also be of some interest: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1876224
The UK SWR was also fairly sensitive to gilt maturity – my paper at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4742456 might be of interest
@TA: … “But the effect could easily be arbitraged away. If it’s a slam dunk that the index leader outperforms the index then everyone piles into the leader. At some point, its price is forced so high that late entrants cannot obtain a worthwhile return on investment. The stock’s price flattens out then declines – the holders lose, undoing the effect.”
But does the arbitrage relationship still hold? Does EMH hold at all? What if the dominant investor group is totally price insensitive and thus, by the axioms of EMH, not rational? If investors make no judgement on future returns and simply invest every day/month/year into a product that they are told will replicate the index return relatively closely. And that product only buys a limited selection of index stocks – those with the largest market caps – since this will replicate returns to that tolerance whilst reducing transaction costs and churn.
Is it totally coincidental that in the 25 years or so since index tracking really started, that the equity market where index tracking is most dominant has ended up with large caps outperforming small caps to the most obvious degree?
@ZXSpectrum48k (#44)
Ahn and Patatoukas “Identifying the Effect of Stock Indexing: Impetus or Impediment to Arbitrage and Price Discovery?” (2022) might be of interest (I don’t pretend to understand the technical details of this paper). One relevant part of the conclusions is quoted below
“Overall, our article provides new evidence on the causal effect of stock indexing on arbitrage conditions and price discovery. The critics of stock indexing often argue that index investing leads to excess comovement and reduces price informativeness. In contrast, our causal evidence shows that index investing facilitates informed trading and increases the speed of price adjustment to news for more arbitrage-constrained micro-cap stocks. To be clear, we do not argue that there is only a bright side to stock indexing. Moving forward, a growing concern with respect to stock indexing is the concentration of ownership and voting power among the “Big 3” index fund managers: Vanguard, BlackRock, and State Street.”
@ZX – I’m open to the possibility that Howard’s idea could be right. But so far no convincing evidence has been offered other than a pattern in the data. Is that a market-beating strategy? Is it noise? Is this time different? Does correlation equal causation? Let’s find out.
#46: Check out “Inelastic Markets Hypothesis” on SSRN (Bouchard 2021, Gabaix 2020/2022/2023, and Ruf 2024 – plus many others) and the long Bogleheads IMH discussion thread linking to NBER and other research. There actually seems to be quite a lot of evidence here.
Maybe IMH is a factor here, or maybe not, but I struggle to believe from first principles of institutional risk management that any concentration effect could be arbitraged by overextension – i.e. by institutions piling into a single (or the top two or three) stock portfolios.
@ZX will be better placed to comment, but I would have thought that a fund manager or PM would be taken out and shot if they put all the clients’ or their employing firms’ own money into just 1 (or even just 2 or 3) stocks. Ark Invest Cathy Wood’s bets would look tame in comparison. I think Ackman runs 8 to 12 stocks at a time, but only having 1 to 3 and then only selected on size and not fundamentals would be something else again.
@TA @Howard – At its core this seems to be entirely down to whether the largest companies in the index/world continue to grow as a proportion of total market cap. They obviously have for an extended period, and if that remains true then it is very likely any strat based on over-indexing on them will outperform the market; I don’t believe there is anything structural about the stock markets and indexes themselves that makes this true and thus their is no inherent inefficiency for arbitrage.
The only question is whether you think the trend will remain true and we’ll relatively quickly reach a point where 5-10 companies will represent half of the total market cap of all traded stocks globally, and within 10-15 years that the same number of companies or less will be worth 2/3rds or more of total market cap or not?
@ZX this seems like a mischaracterisation on how investments in passive funds work. If someone suddenly came along and put £10 trillion into a passive index tracker that money is invested equally as a share of market cap against all companies in the index. If a company is worth 1% of the index before it will be worth 1% after, and the same for one worth 50%. The exact point of index trackers is that they are entirely and only sensitive to price (in the form of market cap).
The only area where the considerable increase in passive trackers is likely to have some impact relates to stocks that fall outside the scope of popular trackers, and stocks that may be perceived as likely to fall out of the scope of popular trackers, but that really is relevant to differentiating 5-10 companies at the top of indexes of 100+ companies from the wider index.
@JohnG. I think you mischaracterize how trackers actually work. The majority of trackers do not buy the index constituents in a market cap weighted manner. To lower costs, they replicate the index using a small optimized selection of stocks. Those stocks are chosen to minimize the tracking error to the index.
The stocks chosen are those that offer the highest beta and lowest alpha to the index. Unsurprisingly, those stocks with the highest market cap weight will tend to always be chosen. You get 35% of the market cap of the S&P with just 2% of the stocks, 60% with the top 10%. The smallest 50% of stocks only provide 9% of the market cap and far far less in return volatility terms. You simply don’t need them to replicate the index.
So, while inflows are occuring (typically in a rising market) trackers will tend to overweight the bigger stocks and underweight the smaller stocks through this replication process. This creates a feedback loop favouring big stocks. Only when you get outflows does that effect unwind. Basically, the market is not perfectly elastic as EMH requires; different investors have different objectives. Flows do matter.
ZXSpectrum48k (#50)
There are plenty of index trackers that hold the entire index – e.g., SP500 tracker VUAG has 503 equity holdings (i.e., all of them) and some that do not, e.g., HSBC FTSE All-World Index Fund has 3479 equity holdings which is quite a bit less than the 4225 in the index. While you are right to suggest the missing holdings are probably those with the smallest cap, needless to say, I have not compared the lists to see which ones have actually been omitted!
It is also useful to note that the indices themselves are only updated periodically (quarterly for the SP500 and semi-annually(?) for the FTSE All world – much less frequently than the monthly refresh of the bond indices). The fund managers have to decide what to do with cash inflows/outflows including dividends and whether to keep small cash buffers for liquidity purposes. I suspect that each fund house does this in a slightly different way and will attempt to reduce management and trading costs while maintaining a small tracking error. I note that the HSBC All world fund has a -0.07% allocation to ‘cash’ and a 0.07% allocation to ‘others’ (digging down, this possibly includes futures) that may be part of this process.
Trackers are increasingly dominated by lowest cost optimisers rather than full replication.
And indices are very concentrated as (cap weight) AAPL, NVDA and MSFT each account for 4% of the MSCI World, with a quarter of the index in tech, 17% in financials and (a more than slightly concerning) whopping 71% in the US.
As @ermine has said before, it’s a WINO (World In Name Only) index now, but the situation may be worse than widely realised as it’s liquidity weight and not cap weight which moves price.
As @ZX says, the feedback loop here is led by optimisation.
But there’s a further issue that the largest caps still move more on flows than you’d expect based on cap weight even allowing for optimisation effects (which are the main, but not exclusive culprit).
The problem seems to be that whilst. in absolute terms, there’s way more liquidity in the mega cap stocks than the rest, this is not so true on a cap weighted basis.
Market makers make such a pitiful spread on the biggest firms in the index – low bps worth – that they don’t tend to run the risk of upping their stock of mega cap shares to keep up with and fully reflect those firms’ increasing market cap.
And so in and out flows affect price more than they otherwise should.
Say imaginary Mag 7 stock A has a 2 bps spread and costs $1,000 per share with 2 billion shares outstanding and with market makers holding 1 million shares.
Now, in contrast, imaginary nano cap AIM stock B costs £1,000 per share with 1,000 shares outstanding, a 250 bps spread, and with 50 shares held by market makers.
Despite market makers holding $1 billion of stock A and only £50,000 of stock B, in practice – and counter intuitively – it’s stock B, and not stock A, which is likely to have better liquidity, given the million and a half fold difference (adjusting for FX) of market caps between stock A and stock B.
@ZX – Thank you for the detail you went into, and you’re correct that my summary ignores the detail of how some trackers create representative holdings. I’d be very interested in seeing any evidence based analysis that this is affecting market prices.
I just checked holdings in two trackers including a FTSE Global All Cap. Neither trackers seemed to be meaningfully overweighted in the largest firms. The FTSE All Cap ‘only’ has 7k of the 10k firms in it but the top 10 firms aren’t over-represented. If a company isn’t buying excess amounts of the top 10 firms in an index of 10k firms then I am dubious, until presented with evidence to the contrary, that index trackers in general are over-buying the largest firms in a way that impacts on their value relative to the wider index.
@JohnG — The argument being made here is broadly the same that Mike Green in the US has been making for a while.
For instance: https://www.etfstream.com/articles/mike-green-passive-ownership-is-approaching-dangerous-levels
If you Google around you’ll see long and sometimes persuasive engagements with it from supporters as well as cogent counterarguments.
Not to stop the discussion here! 🙂 Just because you’re asked for evidence, which to be honest I’m not convinced Green has yet provided. It’s more a coherent hypothesis as I understand it.
I can quite believe a relentless shift to passive will cause distortions at *some* point. Financial markets always eventually have too much of a good thing, and even Bogle believed there was some limit to the passive share versus active re: price discovery.
However I’m no expert on market structure etc (to put it mildly!) nor where these limits would lie — let alone if we have reached them or are decades away –- and am mainly chomping my popcorn and watching!
#53: The point is that not all trackers do optimisation, but all investors in companies in an index are affected by optimisation.
If you look at a tracker doing full replication then definitionally you’ll see no optimisation. They will not overweight.
But that doesn’t mean they’re not affected by (the larger flows) going to (cheaper OCF) optimisers *if* optimisation does change prices (as it will then increase the cap weight of the biggest firms, and full replication trackers will have to follow those increased weights).
#54: On evidence, I would recommend, before making up ones mind on the issue (one way or another), first going through all of the papers on the Inelastic Market Hypothesis on SSRN. There’s quite a lot there.
I also see Klement on investing has also covered this issue a couple of times. He seems to think that, in recent years, about half the price impact of passive flows has persisted adding 1.5% p.a. to returns at the index level with (in another piece) highlighting an estimated current uplift likely about 4% p.a. – N.B. as passive share increases any effect should become more extreme.
Whether this is a problem, an opportunity, neither or both is a matter of opinion, preference and values. Whether or not it exists is a question of fact.
If it does exist then personally I want to take advantage of it if possible.
I have a very tentative idea here that if the effect is real and significant (and persistent and increasing) it might then mean fewer and smaller flows going into the smallest stocks than would otherwise be the case- i.e. the ones not in any tracker. If those firms don’t screen well and given that they’re likely too small to be investible institutionally (wouldn’t move their performance dial but would move the securities’ prices) exploitable pricing anomalies could arise. Less a case of buying when there’s blood in the street and more of buying when there’s noone in the street.
There’s 1205 companies on AIM of which 438 are in the AIM All Share. 150 did less than £150,000 of share trades per month. 120 of those have revenue. Overwhelmingly those 120 are utter bilge. Uninvestable / don’t touch with a bargepole attached to a bargepole territory. But there’s a few in the 120 whose valuation makes no sense in a truly efficient market.
Take Fiske, an investment manager and stockbroker with management fees of £3.9m and stockbroking commissions of another £3.9m in 2024. Expenses totalled £7.2m, leaving operating profit of £660k. Revenue has been growing for the last 8 years, and is up roughly 3x since 2016. Has £6.4 mn cash in the bank. No debt. Market cap is £6.3 mn. A true net net. Effectively you’re getting paid to buy the shares on a sub 10 PE. More interesting Fiske owns 0.071% of Euroclear, which delivered net income of ~£970m in 2024 on a 16% CAGR over the last decade. In December 2024, LSEG sold a 4.92% stake in Euroclear to TCorp for €455m, valuing the business at ~£7.75b. Accordingly, Fiske value their stake at £5.3m, 80% of Fiske’s market cap.
*If* the efficient market hypothesis is watertight then situations like this should not happen.
The story about Green leads to this 2024 paper: How Competitive is the Stock Market? Theory, Evidence from Portfolios and Implications for the Rise of Passive Investing
That paper posits a 33% decrease in stock level elasticity over 20 years. Based on that they say:
“In light of these estimates, the rise in passive investing over the last 20 years has made the demand for individual stocks 11% more inelastic.”
There’s another interesting 2024 paper: Passive Investing and the Rise of Mega-Firms
“Consistent with our theory, the largest firms in the
S&P500 experience the highest returns and increases in volatility following flows into that index.”
An interesting counter here claims that the run-up in mega caps has been driven by Big Tech’s outsized earnings.
https://www.ft.com/content/6f3d9258-b2e6-44da-a1a7-1eda2f5ad796?utm_source=chatgpt.com
Pushback from Morningstar:
https://www.morningstar.co.uk/uk/news/246766/passive-investing-%3D-broken-markets-discuss.aspx?utm_source=chatgpt.com
Pushback from Goldman Sachs:
https://www.marketwatch.com/story/the-3-biggest-passive-investing-myths-goldman-sachs-says-investors-need-to-get-over-334bf3a9?utm_source=chatgpt.com
This 2025 study claims ETFs assist with price discovery
https://www.ft.com/content/e23897c4-0d9d-4b6a-9c43-590870e4eb5e?utm_source=chatgpt.com
There’s plenty more on both sides of a debate that’s been going on for years.
But I still can’t find anyone concluding from all of that: just hold the index leader and you’ll outperform.
@Howard – I don’t believe in the strong version of the EMH i.e. that markets are perfectly efficient. I don’t doubt there are opportunities to be exploited. I do think it takes skill to exploit them consistently though, barring the possible exception of systemised factors – which require great fortitude. I can more readily believe your AIM market thesis is an exploitable inefficiency – at least for a time.
It’s no surprise to read this, I often wonder if the real issue is education. I think the decades gone by of defined benefit pensions have conditioned people to believe putting money in a pension simply might be enough. But most have no understanding of what their investing in, or the potential consequences of this on their future.
@Eddie,
I agree. DC and SIPP completely mis-named. They are IMO tax advantaged saving schemes and NOT pensions!