Weekend reading: Alas, Smith and moans
![Quality street [Members]](https://compass-signal.live/wp-content/uploads/2024/01/consumer-goods-EV-gross-profit-ten-years-150x150.jpg "Quality street [Members]")
Previous post: Investing for beginners: All about assets
≡ Menu
The Latest Articles
Essential Reading
Better Investing
Property Posts
Get Money Motivated
Monevator is a place for my thoughts on money and investing. Read our disclaimer. Subscribe to get all our free posts via email. Follow us on Twitter, Facebook, or via RSS. You can also send us a message.
Copyright © 2007-2026 Monevator. All rights reserved.
Disclaimer: When investing, your capital is at risk and you may get back less than invested. Past performance doesn’t guarantee future results. All content is for informational purposes only. I make no representations as to the accuracy, completeness, suitability or validity of any information on this site and will not be liable for any errors or omissions or any damages arising from its display or use. Full disclaimer and privacy policy. This site uses cookies and features affiliate links.
Comments on this entry are closed.
Absolutely, I find Terry’s letter muddled. He seems to lament the drivers of what he calls a bubble over several pages before announcing he is at least to some degree now chasing the momentum.
Reads to me like the hubris of a man who cannot simply say we got it completely wrong sitting in unwieldy global consumer goods companies that were struggling against inflation, rising trade barriers and changing consumer preferences and opting out of the companies on the receiving end of trillions of dollars of AI investment.
It’s incredible how sticky assets can be, he has substantially underperformed with remarkable consistency for 5 straight years and yet retains an 11 figure AuM, how is paying 1% and hoping for a change of fortune preferable to the alternatives, I don’t get it.
On a random aside I find myself on holiday at the same hotel as Neil Woodford! perhaps I should ask his perspective…
A very fair summary of the Inelastic Market Hypothesis (conjecture?) Thank you.
It’s not that the (currently positive) net passive inflow explains all of current prices, anymore than information diffusion under the EMH means that all prices always are perfectly optimal and rational.
Rather non price sensitive inflows tend, over time, to increase each of overall market cap, the ratio of investment / disinvestment flow to price changes, the concentration in the largest names (due to index optimisation) and to increased volatility (due to reduced liquidity in the largest companies as market makers don’t increase stock holdings in them at the same rates as their cap weights rise, given the ever increasing compression in spreads for the largest stocks, and therefore reduced profitability per unit of market cap for market makers to the make the market in them).
NB: very recent falls in mega cap hyperscalers could be interpreted as either confirmative or disconfirmative of the IMH. The disconfirmative position is as outlined in the piece (although the IMH does not preclude non passive / active flows from setting day to day pricing, so it’s something of a straw man). The confirmative position for the IMH is that reducing liquidity (per unit of market cap) and increasing volatility in the largest names (due to passive flow effects) means that they will fall (or rise) more quickly, and go to extremes of over sold (or over brought), more frequently and intensely, than otherwise; and that’s exactly what we’ve just seen taken place, as predicted.
“But [Terry Smith’s] sermons were certainly more Puritan than The Church of England.”
That’s not very difficult, given that the Puritans found the Church of England far too close to the “corruption” of Rome, and mystical tradition.
The whole “Mayflower” episode was about escaping to a new territory in the early 1600s, to be able to establish a Puritan colony. The religious tension in England continued until coming to a head in the English Civil War, and the associated Wars of the Three Kingdoms (England, Scotland and Ireland).
Casual, modern interpretations of the English Civil War and the ensuing Republic (Commonwealth) often emphasise the Crown-versus-Parliament aspect, because that makes sense to the secular mind. However, a large portion of the motivation was the ongoing ferment of competing religious ideas, from Roman Catholicism though to ultra-protestant sects which Cromwell himself found intolerable.
Within the Christian domain, the Church of England tends to be pretty broad, from Anglo-Catholics who shock Romans with their preference for the Latin rite, through to fundamentalists who go beyond what, for example, mainstream Methodists consider proper.
Thanks for the Fleetwood Mac subs, maybe they should include “Oh Well” from the Peter Green days.
The Grim Historian piece is an interesting, if long, read.
It isn’t that index funds are intrinsically bad. Index funds are good. Yes, they are cheaper than active funds but, more importantly, they are much more diversified than typical active funds. Given that 80%+ of single stock equities return less than cash, equity portfolio investment is essentially about not missing the big winners. Owning index funds helps you not miss out on that small percentage of stocks that generate nearly all the actual returns from a market.
The issue with index funds is essentially one of their dominance. It should be obvious that if total weight of all index funds asymptotically tends toward 100% of the market cap, then the market would cease to function. It wouldn’t be a market, it wouldn’t be efficient etc. The issue for me is that we are functionally much closer to that that we appreciate. Index funds may not even be 50% of the market, but when we add in active funds that shadow the index, add in those traders trying to front-run index funds etc, the effective impact of index tracking is huge.
The impact on liquidity is the most obvious issue. Essentially, I feel that index-tracking is generating systemic risks. The simple size of the tracker components of Blackrock, Vanguard, State Street, JPMIM is just too big for the rest of the market to clear.
On an different subject, the link “Why nobody understands the AI boom – Groundbrkr” is pretty much where I am on the AI “bubble”. As someone who runs a “long convexity” portfolio, I’m sort of obsessed with the second derivative. I keep seeing things in the current AI boom that keep reminding me of signals before fixed income crises.
I have a couple of explicit momentum ETFs. Jack Bogle really wouldn’t approve of those at all. I sort of follow Terry Smith’s argument and ZXSpectrum48k adds colour “index-tracking is generating systemic risks”
I was tickled by the post that kicked off the end of the professional writer
I use neither, though I have trialled them. In the case of AI writing you lose your voice. And in the case of Google maps you don’t know where the hell you are. I can still use London buses because I learned the high level layout of the city in my student days, well several decades ago. The bus routes are different now, but knowing where they are headed and that old map of the territory is good enough, plus you get to see the city unlike the troglodytes of the Tube. I do appreciate that a retiree is going off-peak and has time, the Tube wins hands down on speed 😉
Asimov’s The Feeling Of Power is salutary. By all means use these tools to explore the problem space, but once you have a solution, make it your own. Waze showed me a great route from Suffolk to Somerset. But I took the time out to identify why it was better than what I had used before. And to get the key switchpoints so that I can drive it unassisted. Because: Asimov knew.
There’s a commonality in the AI writing and the indexing thing. You’re coasting in neutral, and there appears to be a downside of that.
“Coasting in neutral”-what a great phrase however…..
“Investment Coasting” does take a lot of self discipline-the human urge to continually tinker with their investments is all encompassing
However many “amateur “ investors seem to able to manage “Investment Coasting” with outstanding end results outperforming well over 80+% of “active “ investors-that’s a winning investment plan for them-so far
xxd09
I imagine the market being like a polkadot balloon, index funds may inflate the balloon but wont change the relative spacing of the dots. They proportionately scale up the effect of what traders do, rather than resist it. If index funds were sold, it’d take air out the entire balloon, but the relative price differences between different companies wouldn’t change. Likewise an overly large proportion of money in index funds wouldn’t create real trading opportunities because the relative price of everything moves in proportion
Smith, etc, justifying job. There’ll always be active people for various reasons
@Matthew Ainsworth #9: it’s a rather beautiful and very clear analogy which you use (it reminds me somewhat of the ‘smooth tension’ features of the cosmological constant model of ‘dark energy’, or of Roger Penrose’s ideas about Cyclic Conformal scaling Cosmology).
And I agree with you that Terry Smith is just talking his own book.
I also think that he makes a rather poorly argued case in his letter against passive indexation (and is not convincing to me in explaining, less still in justifying, his underperformance, especially given Fundsmiths’ hefty near 1% p.a. fee (compared to the just 0.25% to 0.45% p.a. charged for the UCITS Quality Factor ETFs on offer in the UK).
But, I have to respectfully disagree fundamentally with your logic elsewhere, which reflects a consensus about passive share which, IMHO, is both flawed at the level of first principles, and perhaps already empirically falsified.
Passive flows aren’t neutral scalers.
Cap weighting creates actual differential impact upon relative prices.
Index funds buy in proportion to current market cap.
When inflows arrive, they disproportionately purchase the largest constituents (e.g., Apple, NVIDIA, Microsoft) because those already dominate the weights.
Smaller stocks therefore must and do receive far less absolute buying pressure.
Moreover, most capital in index tracking ETFs and funds uses some form of optimised methodology, rather than buying every stock in the index at cap weight.
Large stocks are the least elastic (they are the hardest to substitute for in the index), which means that the index tracker ETF or fund using optimisation must buy them at whatever price.
And so in/out flows push their prices up/down even more again than for their smaller peers.
Accordingly, a dollar or pound going into passive funds doesn’t just “inflate the balloon uniformly.”
Rather. it creates outsized, persistent upward pressure, especially and specifically upon relatively inelastic mega caps.
Empirical work (Gabaix/Koijen, Michael Green’s analysis with others) uses granular instrumental variables to trace this. Flows drive prices more than fundamentals in inelastic regimes (e.g., $5–$8+ of market cap increase / decrease per $1 inflow /outflow arises, as compared with passive advocates’ assumption of next to no impact, and the observed effects being even higher for the largest stocks, and rising over time as passive share grows).
Your analogy assumes active traders (“what traders do”) can maintain relative spacings.
But, as passive share grows, so the free “float” available for active trading shrinks, especially in mega caps. In this regard, it’s the ratio between the active share and the passive share that matters here.
This leaves active managers facing a thinner, more inelastic market where, in effect, more than half (and rising) of the owners (passive) won’t sell at reasonable prices
@Delta hedge – maybe blowing up a surgical glove is a more accurate analogy! – some parts of the “balloon” are in different ratio to each other – some parts more elastic than others in preknown ways decided by traders, the nominal pound amount is growing more where the market cap is already more (the fingers of the glove can stretch more), but I think it’s maintaining some sort of predictable proportionality – the same overall shape as such. If percentages drift maybe something is squared somewhere, but ultimately that predictability makes it hard for anything it does to not already be priced in, which is why I think increasing indexing doesn’t create much opportunity by itself as it doesn’t increase the unknown – you know one finger of the glove isn’t going to suddenly get 3x longer than another finger when the glove as a whole is inflating
A thinner market for traders should mean that when they collectively buy or sell, price moves more because they are operating from a smaller pool, they have to bid higher to actually buy or lower to actually sell, as the index fund is like an unresponsive obstacle – almost a like large stone inside the balloon, maybe like how all the lost bitcoins that’ll never be sold mean people trading that operate in what’s a smaller real market
@ZX – I’d appreciate your take on this, but the market cap framing for passive funds has always struck me as the wrong metric. What we should be looking at if we’re concerned about that is how much of the daily trading volume is index funds vs. active market participants, as price discovery happens at the margins.
The natural extension of this is the inelastic markets hypothesis, but this should theoretically create arbitrage opportunities for anyone paying attention with the capital to act on them, so either the arbitrage is real and it gets traded away, or there’s a specific reason why it can’t be (not enough risk capital, no way to know the ‘correct’ price etc.).
[As an aside, re Smith – it always frustrates me that people never properly benchmark him. He’s a quality growth manager, so judging his performance fairly means comparing him to something like the MSCI World Quality factor, not the MSCI World – otherwise he just gets credit for a period where quality happened to be doing really well.]
@xxd09, SkinnyJames: returns, volatility, liquidity all come in streaks.
Regimes come and go.
Managers and styles like Woodford’s and Smith’s come in and out of fashion.
There are cycles.
Things rhyme, then reverse.
A benchmark like the MSCI World or World Quality Factor indices can’t fully capture these regime changes.
They only compare like to like (equity v equity, quality v quality). What about like to dislike?
A truly successful strategy is one which can survive across many regimes with a high likelihood of high CAGR, low max drawdown and low drawdown duration, with higher reward per unit of risk etc.
Show me a mainstream fund manager or passive ETF which had done that over generational timescales, and where it was consistently and reliably identified in advance (using an easily repeatable framework) who they were and which funds they were.
Growth and tech are having a long day in the sun.
Since it’s inception on 11th February 2010 the 3x NASDAQ LETF TQQQ has returned 37,360% (374.6x) cumulative nominal, 24,021% (240x) real terms cumulative and 43.5% CAGR net nominal.
Just from dumb and they come buy and hold.
But, had it (TQQQ) existed on the same terms as now (but with the prevailing funding costs then for the leverage) then from April 2000 to March 2009 a (synthetic simulated) TQQQ would have fallen by 99.98% cumulative, turning $100,000 into just 20 bucks.
It’s like that film “Trading Places”. It all depends which side of the tracks you find yourself on.
With B&H ‘passive’ approaches time it right (even once in generation sometimes) you can do very well. Time it wrong you can eviscerate yourself.
@ermine #7: “explicit momentum ETFs”. Building in a simple to operate ten months’ (two hundred trading days’) Simple Moving Average on/off filter rule to a index tracking momentum ETF (SPMO for top decile momentum within the index) does wonders.
It (as such a simple unlevered system) virtually matched the performance of selecting the single overall
best total performing of all the 29,078 US single stocks listed over the whole period from December 1925 to December 2023 (Philip Morris/ Altria, which would have turned $100 into $265 million for a 16.3% CAGR): see my comment as HBCTIA (Howard Beale Can’t Take It Anymore, “Network”, 1976) to the Reddit post here:
https://www.reddit.com/r/LETFs/s/4om8dvzXxj
@Delta Hedge – I’d argue that proper benchmarking does capture regime change (so long as the manger’s style doesn’t change), if quality is out of fasion and Smith is doing as badly as the MSCI Quality Factor benchmark, then that’s justifiable. The point on proper benchmarking is to be able to seperate the factor’s beta with the manager’s alpha (or lack of it).
@DH #13 > (two hundred trading days’) Simple Moving Average on/off filter rule to a index tracking momentum ETF
Thanks for that. I moved into momentum (and value momentum) because of the same reasons all the passivistas on here seem to be leaping off the wagon.
But with generic momentum there is the philosophical problem of when/how do you get out of momentum, it’s not entirely B&H The second fund is exceptionally evil in that it shifts into other asset classes, I will be interested to see how that works in the AI crash yet to come. But for the more generic value momentum, that SMA could be informative.
Actually, I was thinking that maybe there won’t be an ‘AI Crash’, and that perhaps there will be a steady deflation.
Big companies come, big companies go.., bust ( Northern Telecom ), taken over, crippled by lawsuits ( Minolta ), merge ( Konica Minolta ), sold off ( Konica Minolta camera division to Sony ), get out of retail, and go for the easier life of B2B, or government contracts ( IBM, Fujitsu, Blackberry ( RIM as was ). And all the rest.
So, with AI, the littler (!) players give up, scale back into some niche, get absorbed for their customer base/some successful part of their business/take your pick. And the mightier ones, and everyone else, stops talking about AI as it becomes part of the infrastructure of business. Or…, who knows ?
As for index investing, I think I’ll just keep on in that direction. Active investing mostly feels to me like just another worry. And the nearest I’ve got to following ‘star’ investors is my less than a handful of BERKS (B).
Delta Hedge-agree returns,volatility,liquidity (risk?) are what it’s all about for the investor
Unfortunately there is no sure way(investment policy) of handling these tremendous variables by the amateur investor so “buy and hold” ie index investing has turned out to be one reasonably successful way of dealing with with these conundrums
Investors should then concentrate their efforts on matters under their direct control ie saving as much as they can,living frugally and watching costs
Mr Market seems to do best when left alone!
xxd09
#16:@Trufflehunt: The US March/April 2000 to September/October 2002 (UK Dec 1999 – March 2003) tech crash saw remarkably few real economy effects. GDP growth slowed substantially, but it still grew in a muted fashion.
And this was against a backdrop of a 78% fall in the broader index (Nasdaq) and an 82% crash in the associated (unlevered) tracker ETF (QQQ) (if you were levered 2x or 3x then you would have taken a virtual wipeout).
So your thesis is plausible.
But, notwithstanding that slightly reassuring near term historical precedent of sorts, this one does look worryingly big.
Data centre Capex is now such a high proportion of US GDP growth.
Maybe any crash stays in the stock market, but, like Ed Zitron is, I’d be quite worried where this all ends up eventually for the average Joe in the street.
Quite possibly not well, like it didn’t after 2008, even if Mr and Mrs ‘Joe ordinary’ never puts so much as a single penny down direct into either semiconductor or memory stocks, or even into the hyperscalers themselves:
https://youtu.be/3u0KeTC7jso?si=icNRADEUHe8LhMJh
Still, I think Ed’s maybe a tad too harsh on whether there’s actually a product here.
This isn’t, I think, another case of the Crypto Bros’ ’empty boxes’, as SBF revealing confessed to just before FTX imploded, and he went to jail.
There is something here.
On the one hand, it’s clearly not either ASI nor even AGI, and I don’t think it can even properly be called AI yet, because its intelligent sounding, sometimes (remarkably) accurate (sometimes shockingly not) output comes from a non conscious/non self aware process.
Moreover, on that score, even if the deal breaker on the “I” in AI is not the issue of the foundational probabilistic nature of LLMs, versus the received template (pre 2017 or so) that ‘real’ AI would be deterministic (symbolic language based, i.e. programmed); then there’s still the bigger, practical (emergent, functional) issue that (whilst LLMs can be an incredible and useful ML tool, albeit one uneven in so, so many ways) they’ll never ever behave like we expected AI to (* see below on why).
However, on the other hand, they (LLMs) are something.
On a good day at work (a focussed, productive one) using Enterprise level GPT 5.5 Thinking I can in ~8 hours output what would realistically take me at least 24 hours of solid, non stop work in pre LLM world. It’s just a fact.
I see it with my own eyes (and I’m not hallucinating the productivity gain, like the LLM might a confidently misstate a fact, although it does this ever less so in my experience).
Sure, it requires enormous levels of human professional end user intervention and interaction, but it (ML) is capable, in many tasks, and when used correctly, of either massively speeding up the task, doing the task to a level of thoroughness time would not otherwise permit, or of getting through far more tasks in the same time.
So, whilst I do find Ed very persuasive (and easy to listen to, and to read on his blog); I am still thinking that, for all of the collosal issues and problems with the LLM economic ‘ecosystem’ (circular financing, a massively loss making business model, no way to fund or build so many data centres, or to provide the power on time or at all etc etc) that, eventually, there’s still got to be an eventual demand out there from the end user consumer facing businesses equivalent to a multi hundreds of billions to some trillions of dollars of annual revenues, even if we don’t get beyond the current LLM type of ML, and there’s no agentic AI revolution which sees mass replacement of people with ‘thinking machines’ (although, Small Language Models, running with specialised chips on device/on phone might upend all that entirely).
Which is all a very long way of saying that a crash might or might not tip us into an abyss; and, as you say, there’s also either a slow grind sideways or a gradual deflation of expectations set of scenarios (although, I seem to remember people saying much the same things about Dot.com in the later nineteen nineties – certainly everyone back then expected the crash, which didn’t make it any easier for those invested then when it came, and was, at it turned out to be so very steep; to win you have first to survive, after all).
(* see here today for why:
https://youtu.be/yRZC6KfcO88?si=sCPqzDeYybPIMiUQ )
Interesting discussion, as always. Smith makes half-baked arguments, hope he is more rigorous with his stock selection. Others here (@DH, @ZXSpectrum) have made a more convincing case. Though I keep coming back to the point: Passive flows are predictable (fact). If they change prices, then active investors should be able to profit, and arbitrage the distortion away. If this is not happening, why not? Maybe the spoils go to certain hedge funds, not traditional buy-and-hold managers.
@Sparschwein #20: The very short answer is that there are real limits to arbitrage.
Passive inflows can induce prices to drift away from fundamentals if they make informed active investing relatively less effective.
The binding constraint is essentially whether someone can safely and cheaply trade against the over or under valuation.
Mispricing can persist and worsen when arbitrage is costly and/or risky; as is especially the case with, say, shorting a rising (high momentum) stock.
There’s a lot of academic work on this in relation to (Fama-French) factors (value, size, liquidity, profitability/quality, low volatility) and in relation to momentum (which isn’t, IMHO, as such, a factor per se but, instead, I think more like an abiding and underlying feature of market structure itself; i.e. sometimes value stocks have high/low momentum, sometimes growth stocks; and sometimes it’s large caps, and sometimes small).
That work is possibly relevant here because the effects of (and also perhaps, to an extent, the mechanisms of) passive flow effects are somewhat similar to momentum effects.
This paper (“The Limits of the Limits of Arbitrage”) from 2009 is a must (if only for the title):
https://www.bartlit-beck.com/assets/htmldocuments/limits%20of%20the%20limits.pdf
The paper predates the passive flow debate by a while, but is, I think, still relevant to it because it does conclude that the survival of overvaluation depends upon broader frictions like idiosyncratic risk and shorting/implementation costs, and that these are real, substantial and common.
Those types of risk are present (indeed acute) in an active arbitrageur trying to trade against the direction of passive (indeed any) capital flow.
Looking at all the empirical evidence, the authors of the paper find that, for overvaluation anomalies (such as passive flow is being put in the frame for now), the limits to arbitrage model works fairly well (but, oddly perhaps, it doesn’t explain convincingly undervaluation).