Are chimps making chumps out of trackers? In a trial of the fittest run by academics at Cass Business School, traditional market cap weighted indices were beaten by indices picked by 10 million monkeys (or rather by their randomly generated computer simulations).
Market cap indices are the spine of the vast majority of index trackers. News that they’re apparently taking a beating from primates therefore caused some consternation here at Monevator’s Passive Investing HQ.
So should we be pounding the earth like Charlton Heston and preparing to worship our new masters in a passive investing version of Planet of the Apes?
Er, no.
The Cass Business School’s two-part paper is called An Evaluation of Alternative Equity Indices (here’s part one and part two).
The research reveals that investment strategies with the power to capture return premiums like small cap, value and low volatility have historically delivered more OOMPH! than you got from the regular old market.
But this is no surprise. Academics have been telling us this for 20 years and we cited similar revelations a month ago on Monevator.
Broad market indices are designed to capture beta – the return of the market. But there are other sources of return out there.
Essentially, this paper shows that other types of index can angle their satellite dishes so they receive some of the beta channel, and also the small cap and value channels, too.
We’ve previously discussed how you can invest in some of those approaches – like the fundamental indices – by putting money into value-tilted passive funds.
Mind the cap
The standard argument for market cap indices is that they are transmitters for the wisdom of the crowd.
The gestalt human investment brain allocates capital efficiently to the firms who will make best use of it, and this is reflected at light-speed in the index.
The case against market cap indices is pretty clear, too. They are passive victims of the misjudgments of the market. For instance, they are mechanically forced to accept overvalued equities that balloon in the index with the smiley face of irrational exuberance painted on.
You’ll often see passive investors challenged in heated comment threads about how their portfolios would have been swamped with Japanese equities in 1989 or tech stocks in 1999, like a council landfill site crammed with trash.
These accusations fail to account for the sheer global diversity of a passive portfolio. Nor do they acknowledge just how many active investors get trapped in expanding bubbles.
Active investors are the ones making the active decision to chase the bubble higher, after all.
The pros of market cap index tracking
There are still good reasons for holding funds that track market cap indices. It’s worth reflecting on those as an antidote to getting too clever by half:
Simplicity – Market cap indices are easy to understand. The bigger a company is relative to the rest, the greater its presence in the index. That’s it, bar common-sense rules to guard against over-concentration in the event we all go bananas and back SnakeOilSystems.com to take over the world.
One of the issues with the alternative, so-called Smart Beta indices is they are often presented in the language of a black box. Too often proprietary strategy jargon and marketing pseudo-science seems designed to lacquer on an extra layer of fees and play to our desire to believe in magic rather than to promote understanding and transparency.
Without transparency we break the first decree of the DIY investor: ‘Never invest in something you don’t understand’. Without transparency we risk investing in complexity rather than efficiency, and paying extra for Advanced Methyl Ether Rejuvenator Balm when we just need creosote.
Low costs – Broad market cap indices contain the most liquid equities and have low turnover which is why they can cost 0.10% to own in fund form.
Academic research rarely takes the actual cost of implementing alternative strategies into account. It deals in a frictionless world where equities can be freely traded and there isn’t a maze of middlemen to dash through.
The Cass Business School paper demonstrates that some alternative indices beat the market cap approach by around 2% per year. In the UK, you’ll be subtracting 0.3 to 0.5% from that in higher Ongoing Charge Figures (OCF) plus bid-offer spreads and any tracking error that creeps in.
Costs are nailed on. Potential returns aren’t.
Availability – We’re now served by a wide range of market cap index trackers that encompass both the weird and the wonderful.
In contrast, many of the indices cooked up in the academic labs just aren’t available as funds you can buy into in the UK.
No guarantees – Debate rages as to why the return premiums exist. Are they rewards for extra risk (in which case they should persist) or the behavioural errors of flawed humans (therefore erodible through exploitation)?
So far the return premiums have stood up to the scrutiny of the markets, but you have to be prepared for the possibility they could evaporate in the future. As Vanguard founder John Bogle says in The Little Book of Common Sense Investing:
I’m skeptical that any kind of superior performance will endure forever. (Nothing does!)
Personal risk factors – If you work for a small cap firm then a good deal of your future returns (i.e. your salary) are dependent on the fortunes of that part of the market. Tilting your portfolio in favour of small cap funds would concentrate your risks rather than diversifying them.
The same principle could apply to some extent to most of us. Consider how your earning power correlates to the economic cycle before tilting your portfolio in favour of, say, value strategies that generally take a beating during recessions.
Performance chasing – Different strategies work best in different time periods, as shown in the Cass paper. Even market cap has beaten all-comers during a raging bull market. Strategies will be declared ‘hot’ based on recent performance – witness the flurries over low volatility and high dividends recently – reducing their likelihood of outperforming in the future.
Johnny-come-latelys will swarm in, then dump the funds when they fail to make ‘em rich. As ever we must resist the temptation to jump on board a bandwagon.
Tracking error regret – How will you feel when your alternative strategy eats the market’s dust for five years straight?
When returns soar we take it as confirmation that we’re as brilliant and blessed as we always thought we were. But when our high-cost strategy is left billowing black smoke in a lay-by as the rest of the market whizzes by, how happy will we be then? That’s not a pain you’ll ever have to feel when you simply invest in the market.
Hard evidence – Twenty years is a reasonable amount of time to judge a strategy’s performance. Ten will do, five is barely acceptable. Anything less is irrelevant.
It’s hard enough to get ten years of performance data for a market cap index fund. We’ll have to wait years for useful numbers on the latest financial products.
Back-testing is used to fill the gap, but be aware that many strategies that glitter in the data mine have lost their lustre in the cold light of day. As ever, investors are wise to be cautious and wary of products with insufficient miles on the clock.
Enough monkey business
I say all the above as an investor who’s convinced by the arguments for factor investing, smart beta, fundamental indexing – whatever you want to call it.
But I know it’s a risk and so I won’t be entirely abandoning market cap investing.
Both approaches will work in tandem in my portfolio because, although I’m prepared to take a chance on the extra reward, sometimes you just can’t beat the tried-and-tested.
Take it steady,
The Accumulator
@The Accumulator,
Re the Cass Study and the shares chosen by monkeys. Given that any selection method is alleged to be better than market cap over the period under consideration, how come that every actively managed fund (especially small cap) in existence for 30 or 40 years has not outperformed the corresponding market cap-based index?
Either I’m missing something or there’s something wrong with the study or the way in which the results are being presented. Given the recent revelations in the 2010 Reinhart/Rogoff work, I’m minded to ask a few questions:
What am I missing ?
Has the Cass study been peer reviewed ?
Why hasn’t any prior study identified the superiority of the monkeys ?
Why hasn’t anyone (such as Vanguard) launched a monkey tracker ?
@Grumpy Paul — A big part of the reason will be fees and expenses. Fund managers aren’t slouches, but besides the theoretical underpinnings of why it’s so hard to beat the market (in short 50% must UNDER perform to make up for those who do better) its very VERY difficult to then overcome an extra 2-5% or what have you in costs.
@ grumpy old Paul I shared your skepticsm when I saw the monkey portfolios. I can see how the monkeys might over perform – just a very significant small cap tilt. What I can’t follow is how they managed to do this with better Sharpe ratios. Can anyone explain this?
The paper on the fundamental indexes just seem to confirm that there is a significant value factor in fundamental index construction.
What is I am sure now wrong is the attempt by some backers of Fundamental Indexes to claim that there is more to their out-performance than the value tilt ie that there is some inbuilt headwind that is intrinsic to cap-weighted indexes. I saw I nice quote from French (of Fama and French) saying that “Fundamental indexes are “just value funds for clients who don’t understand ratios”. Nothing wrong with a value tilt IMHO by the way. It is the stronger claims of the Fundamental Indexers which I think are wrong.
Thank you for posting the reasons why it’s a bad idea to jump into alternative index investing. I was contemplating to switch to fundamental index, and your post clearly laid out the pro of market cap index.
https://www.google.co.uk/url?sa=t&source=web&cd=1&ved=0CDkQFjAA&url=http%3A%2F%2Fdiscussions.ft.com%2Falchemy%2Fforums%2Fedhec-risk-forum%2Fdoes-theory-justify-fundamental-weighted-indices&ei=EKR2Ud_LFuik0AWd9ICIAg&usg=AFQjCNGQAx2KRk17I_yKis8UzqOGaGcuUw
This is a good recent summary of the academic literature on Fundamental vs Cap-weighted indexing debate
To be honest I think the precise asset allocation decision is not of paramount importance as luck will have a big part play in returns turn out to be, even over 30-40 years
I would weight these four factors pretty much in this order
– saving rate each year
– minimise tax
– minimise expenses
– asset allocation
For instance:
– if you spend 95% of your income it hardly matters what your asset allocation is, your savings won’t amount to be a % of what you earned in your working life
– if you are a higher tax payer putting money into a SIPP your contribution is near doubled straight away just because of the tax advantages of the SIPP
@Passive Investor,
Fascinating article which I’ve bookmarked for occasional rereading.
@Neverland,
Agreed. But there’s also maximising one’s income whilst working while maintaining an optimal work/leisure balance.
Not to mention choosing the right parents so as to win the ovarian lottery!
I sometimes think that we are at risk of spending too much time trying to chase the last .5% of return by fine-tuning passive portfolios, to capture value and small cap premiums etc especially when the latter are difficult to find in low-cost funds or ETFs in the UK. However, I can’t deny it makes for interesting reading.
@old guy
I agree with you about fine tuning passive portfolios. I think the best thing is to get the lowest level of cost with a good level of customer service and financially secure administrator
The extra 0.5% per annum from cutting fund expenses is certain, while all this stuff about chasing higher average real returns from different asset classes is very uncertain
0.5% fees saved on a £0.5m portfolio (which is what you will have to fund a comfortable retirement) is £2,500 a year
Assuming you are drawing down say 4% a year off that portfolio to live off, you would be taking out £20,000 a year
So screwing the fund expenses down gets you almost an extra month and a halfs income each year – its like being able to afford another holiday each year
I think I have answered the question about the monkey portfolios performance & Sharpe ratios.
The top 500 companies in the US have a total market cap of around $10 trillion the next 500 only have a market value of $1.5 trillion. The random picking monkey portfolios will therefore very disproportionately (compared with market-cap) weight the smaller companies. For example the 1000th smallest company (market cap about $130 million) will feature in the monkey portfolios as often as Apple or ExonMobile (market cap 3,000 times larger).
So the outperformance of the random monkeys is very likely to be a size effect. For reasons I don’t understand it seems to be well recognised that portfolios of smaller companies have higher Sharpe ratios. For example it seems from appropriate tracker funds that the following Sharpe ratios apply
5 yr Sharpe Ratio
S & P 500 0.3
Russel midcap 0.56
It will be interesting to see if the Cass papers attract professional / academic comment in due course
As far as simplicity and transparency are concerned, I would argue that equal weighting is even easier to understand that market cap weighting. Market cap based indexing was invented as a gauge of the prosperity of the respective stock market; mechanically turning a gauge into a ready-made investment portfolio sounds illogical to me (as long as my intention is to make money over the years rather than to follow the stock market for the sake of it).
Besides, equal weighting provides real diversification, while buying a market cap weighted index fund defeats the purpose of “investing in the whole market” in that respect: in the case of the UK virtually all of the money ends up invested in some 30 companies – so much for the whole market!
I am not aware of any equal-weighted FTSE100/350/all-share funds, so I am currently stuck with a 50-50 combination of VUKE and MIDD (I like ETFs for minor technical reasons) – which is, of course nowhere near equal weighting, but it still has given me both better real diversification and better performance than FTSE100, FTSE350 or FTSE all-share would have done. When it comes to S&P500, my money goes into the equal-weighted XSEW rather than into VUSA, even though the TER is higher by a whopping 0.21% – once again, better diversification (guaranteed) and better performance (so far).
@oldthinker
Well, a cap weighted fund is simpler as it only needs rebalancing when the index changes, which isn’t that often, whereas any other sort of tracker will need to rebalance at periodic intervals.
However, you are quite right to point out that instead of a fundamentally weighted portfolio that biases towards smaller companies, it is perfectly reasonable to get a large-cap and a small-cap tracker which together give something similar, even if the actual holdings are quite different.
Personally I avoid the FTSE100 as I don’t like its components. I have a FSTE250 tracker and see no need to have a piece of every pie. (Having said that, I do have a significant amount of ‘incidental’ large-cap exposure from my IT holdings.)
I haven’t dipped my toe into non-cap weighted funds, but I plan to the next time I add to my US exposure: I’ll probably go with the iShares minimum volatility ETFs. One fund I’d like to see is a volatility-optimised ETF that has a _high_ beta filter. It would be interesting to see if one could get a mix of volatile companies that still average out to give a low overall volatility. Of course, this would be swimming against the tide if the previously covered anomalies persist…
@Greg
I meant simplicity as defined by Monevator in this article, while you are referring to the ease of operating the fund, which is a different thing. The ease, or otherwise, of rebalancing is not of investor’s concern as long as TER remains reasonable. Simplicity is a different matter: try explaining equal weighting and market cap weighting to a child, and you will see :-).
Moe importantly, I like to understand why my money is going where it is going. Unlike you, I am broadly OK with the composition of FTSE100, so I am happy to be invested in, say, both Shell and ARM amongst others. However, I do not understand why for each £1 invested in ARM I need to invest £10 in Shell, which is what a cap weighed index fund would do on my behalf. The relative capitalisation argument does not convince me at all; as far as I am concerned, capitalisation was already taken into account when both companies got included in FTSE100. Investing into them £ for £ makes more logical sense to me, as doing this would give me better diversification, while I would still be investing in large British companies. It annoys me that I cannot do this simple thing with FTSE100; thankfully, an equally weighted ETF for S&P500 is available.
@oldthinker
With equal weighted ETF, I’m assuming that you want to tilt your portfolio towards small cap and value, correct?
The way that I built my portfolio is similar to yours, I’m using US listed ETF’s instead, with VTI (total market) and VXF (extended market, only mid cap and small cap). I have VTI at 17% while my VXF at 8% of my total portfolio, but I might gradually increase my VXF holding.
@oldthinker. The reason for investing in £10 in shell for every £1 in ARM is because that is how the market is allocating capital. Even if the market isn’t perfect (with bubbles, over reaction to unexpected information etc it clearly isn’t) it is a big step to say that you or anyone else or any fundamental valuation system can do better.
The reason many people justify fundamental or equally weighted indexes is to avoid the valuation errors in the market price. The trouble is that imperfect though it is there is a lot of information in the market price (eg about growth and risk). By deliberately ignoring this information ‘fundamentalists’ are saying that the fundamental valuations are a better judge than the market of fair value. I have seen no convincing empirical evidence that this is the case and the two main theoretical papers (arnott and hsu) have been heavily (and effectively in my view) criticised.
In all this it seems to me (to paraphrase Churchill on Democracy) that “the market price is the worst form of estimating fair value except for all the others that have been tried from time to time”
@dmdave. I am interested in your using US listed trackers as I was considering doing that to get a Value tilt. I was put off by the approx 2% currency spread at Hargreaves Lansdown which would be payable on all dividends too. I would be interested to know if you have found a better option?
@DMDave
I would say that I primarily want to diversify for real. Tilting towards smaller cap and value is a potentially useful side effect, but my primary intention is diversification across the companies included in my chosen indexes. The split does not have to be absolutely even, but I do not want to invest 50 times more into company A than into company B – this is not diversification; B can as well be ignored in this case.
@Passive Investor,
Yes, following the way the market is allocating capital is a strong argument for a passive investor (which I am). However, I am not at all happy with the degree of diversification achieved this way – the result looks too much like all eggs in just a few baskets. I sleep better at night knowing that I am properly diversified, and if I may take a performance penalty for this, so be it. So my strategy is not based on the assumption that equal weighting is likely to beat market cap weighting. In fact, I reckon that the results will be reasonably similar in medium to long term – but I will sleep better throughout.
I would quite like to use US listed ETFs but:
– Inside an ISA I would be hit with an extortionate currency conversion each time I bought or sold.
– Outside an ISA I would be hit with swinging taxes as they don’t have reporting status.
🙁
A friend claims to use a misogynist portfolio: if a woman joins the board, he sells.
@oldthinker. I am concerned about the over-concentration of FTSE 100 and just looking at current weighting HSBC is nearly 8% which is more than I am comfortable with (or realised). Still by the time this is diluted by FTSE 250 / foreign / cash and bonds depending on exact figures the weighting if HSBC in a whole portfolio may easily be only 2%.
The trouble with an equally weighted FTSE 100 portfolios is that the smaller constituents 0.04 % will with at 1% have a weighting of 25x the market weighting. This seems to be swapping one risk for another to me. …
@ Passive Investor
Sorry but I should’ve mentioned that I live in Canada. In a sense yes, by tilting your portfolio to have a more exposure to small cap, it’s a higher risk, but historically small caps do have higher rewards.
For example, small caps have the potential to become large caps and give out big rewards. Just look at Apple >10 years ago when it was on the verge of bankruptcy. At least that why I’m doing what I’m doing.
@ DMDave @ oldthinker I have a small value tilt for the reasons you say. The equally weighted fund seems an extreme (and possibly expensive) way of doing this. Also it may not achieve exactly what is intended in the sense that the S&P 500 or FTSE 100 don’t contain small caps. An equally weighted S&P fund is more of a tilt to ‘larger mid caps’ I guess.
That should say ‘small company’ tilt.
The alternative index approach sounds appealing but I can’t help feeling that the attention it’s receiving is at least partly due to its success in the past 10 years.
I suspect I’ll end up using both approaches ..
hmm. I am coming to the conclusion that the new fancy alternative index products are, bottom line, a strategy to entice us to pay higher costs to the providers. I know myself that I’m a sucker for ‘stamp collecting’ and can’t help but think ‘oooh yes, that looks shiny and new and interesting – I want some of that’. The hardest thing about passive investing (or indeed any investing) is maintaining the discipline to keep the strategy simple and above all to control costs. I totally agree with a previous comment about the marginal returns of all this tinkering with tilts. Best thing to do is to keep socking away money into the lowest cost tracker you can find and then ignore it…. (now, how can I make myself take my own advice??)
@vanguardfan. I am completely with you. The urge to ‘tinker’ is very very strong and needs to be resisted. Investment is all about knowing and dealing with your own psychology, I more and more realise. Wherever possible I use Vanguard too. The only thing I don’t like about them is that they are even cheaper in the US where to be fair they have huge economies of scale
Almost completely. I take a small cap tilt and would have a value tilt too if it was easily / cheaply available.
@Passive Investor
Swapping one risk for another is the name of the game – sadly, there is little else that can be achieved by our investment decisions. Being the kind of person I am, I would sooner put my trust in the law of large numbers than in the collective wisdom of fellow humans, so I feel that it is more important to be diversified than to be aligned with what the market as a whole is doing. Therefore, the concentration of FTSE100 in a small number of companies and sectors is of more concern to me personally than the remote chance of many smaller companies underperforming all at once (which is the only scenario where my relative overexposure to them would seriously hurt me).
You are, of course, right that diluting by FTSE250 alleviates the problem to some extent, but I see the resulting bimodal distribution as a poor substitute for roughly even allocation across the constituents of FTSE350, which I have no way of achieving at a realistic cost. Diluting by other asset classes also helps, but this is a separate dimension of diversification, and many other factors start coming into play here.
Although I don’t follow your strategy I can see the rationale for it. As I mentioned I hadn’t realised that HSBC is nearly 8% of FTSE 100 – I had it in mind that it was nearer 5%. For what it’s worth I keep FTSE 100 at 60 % of my UK shares but even so that leaved more than I would like in a single company. All the best
Apart from women’s handbags what other products or services do we buy more of as they get more expensive?
Intuitively, everyone knows that mkt cap tracking is sub-optimal. The fact that it works better than most active funds is not, in my view, a good justification for allocating capital by price.
The problem is that if everyone else is doing it the fundamental approach is drowned out by the competion. If everyone else thinks South Sea Shares or Enron is the place to be any fund that is underweight those gets crucified. Until the market corrects. But that could be five years or longer.
Surely the market cap of a company is, in part, simply a reflection of its size? HSBC has a bigger market capitalisation than other banks at least in part because it has much bigger assets than those banks and big operations in many countries. But does that make it any more appealing as an investment than a smaller bank?
Surely, the factors that influence whether a particular company is one that you want to be invested in (directly or indirectly) is the dividend yield that you expect to receive from it, the capital growth that you expect to receive from it and the level of risk that you might not receive what you expect. If the market changes its view on those three factors, it will certainly affect the share price of the company and therefore its market capitalisation, but it is only one factor.
If one bank has a market capitalisation 10 times bigger than another, then it seems to me this is not sufficient reason to put a 10 times bigger share of your portfolio into the first bank than the second (or to invest in a market weighted index fund that has this effect). I had never really thought about it before, but I’m starting to agree that it might be better to get the greater diversification that would come with equal weighting.
If there was any evidence that investment returns are higher for the companies with the biggest capitalisation, then market weighting would make more sense to me, but I understand that the evidence is that investment returns are, in the long run, higher for small capitalisation companies. I think it is true, however, that large capitalisation companies tend to be less volatile, so if you prefer to have lower volatility and accept a likelihood of slightly lower returns, market weighted index funds would make sense. Personally, that is not my preference, so the above posts have made me wonder about adjusting my index fund investments to improve my diversification.
@ivanopinion,
You have very succinctly summarised my conclusions on reading the above posts. But there are very few UK passive funds that are suitable for increasing diversification. As others have stated, FTSE 250 trackers help by skewing towards mid-cap companies but are presumably themselves market cap weighted. The HSBC FTSE 250 tracker has a pretty impressive 16 year record but did seem to drop more steeply than the All-share index in 2008/9 and August 2011. But if you have the temperament to accommodate that volatility, then that’s not a problem.
What I’d like to see are more funds like the Vanguard Equity Income fund but with lower caps on individual share and sector holdings and maybe a further set of variants tracking the FTSE 250. I realise that what I’m really talking about are further customised indexes and funds which track them. Such funds would have the following merits:
– low cost
– passive
– greater diversification
– tilt towards smaller companies
– tilt towards dividends
I’m a simple old geezer and am slightly wary of ETFs because I’m not confident of my understanding of them and am also wary of counter party risk.
Finally, common sense says that if it were that easy to beat the All-Share index, then far more active managers would do so consistently.
@ivanopinion,
You have very succinctly summarised my conclusions on reading the above posts. But there are very few UK passive funds that are suitable for increasing diversification. As others have stated, FTSE 250 trackers help by skewing towards mid-cap companies but are presumably themselves market cap weighted. The HSBC FTSE 250 tracker has a pretty impressive 16 year record but did seem to drop more steeply than the All-share index in 2008/9 and August 2011. But if you have the temperament to accommodate that volatility, then that’s not a problem.
What I’d like to see are more funds like the Vanguard Equity Income fund but with lower caps on individual share and sector holdings and maybe a further set of variants tracking the FTSE 250. I realise that what I’m really talking about are further customised indexes and funds which track them. Such funds would have the following merits:
– low cost
– passive
– greater diversification
– tilt towards smaller companies
– tilt towards dividends
I’m a simple old geezer and am slightly wary of ETFs because I’m not confident of my understanding of them and am also wary of counter party risk.
Finally, common sense says that if it were that easy to beat the All-Share index, then far more active managers would do so consistently.
I share the concern about the over-concentration of FTSE 100 in a few huge companies and for me this makes it sensible to over-weight the FTSE 250 in a market-cap-weighted index tracker.
The major disadvantage to equal weighting is that costs are likely to be significantly higher (equal weight charges seems to be o.2 % – 0.3 % over market cap weighted funds in US).
There will be periods when equal weighting out-performs and periods where it under-performs. But (I don’t think) there is any evidence or theory to prove convincingly that once size & value factors are accounted for the equally weighted will out-perform in the long run.
The cost differential is likely to be higher in the UK so why pay more in costs than necessary?
@ Grumpy – like the idea of a monkey tracker.
If you want to diversify your large caps then consider a global fund that will include the UK alongside US, Europe, Japan etc, such as:
Vanguard FTSE All-World ETF (VWRL) or a LifeStrategy fund. It’s about as simple a way to diversify as you can get.
If you want to tilt away from market cap then look at the PowerShares RAFI funds or a high dividend fund or the DFA funds.
http://monevator.com/uk-value-premium-funds/
Or split some of your FTSE All-Share allocation with a FTSE 250 tracker or something like the Credit Suisse UK Small Cap ETF (part 250, part AIM, part FTSE Small cap) as many here have suggested.
If you subscribe to the 3-Factor or 4-factor model then all but a few percent of equity returns are explained by a fund’s exposure to beta, small cap, value, momentum and perhaps another couple of return premiums. Hence why the monkey tracker (small cap) has a good chance of beating market cap and a fundamental fund (mostly value) has a good chance of beating market cap.
No magic, no trickery, if a fund is tapping into more sources of return then it should beat a market cap fund which only earns beta.
@Greg> I would quite like to use US listed ETFs but … outside an ISA I would be hit with swing[e]ing taxes as they don’t have reporting status.
Check out
http://www.hmrc.gov.uk/collective/rep-funds.xls
A lot of Vanguard US ETFs look like they are listed. VOO S&P500 tracker, for example, 0.05% TER.
https://institutional.vanguard.com/iippdf/pdfs/FS968.pdf
Matching the cusip number to the ETF ticker is a bit of a bear, though.
Another way to make the same point regarding market weighting would be to imagine if, instead of a FTSE 100 tracker, you had the choice of two trackers, the FTSE 10, based on the biggest 10 companies in the FTSE, and the FTSE 90, based on the next biggest 90 companies. For every £100 you have available to invest in the FTSE 100, would you really put £43 in the FTSE 10 and £57 in the FTSE 90? Because that’s effectively what is happening if you invest £100 in a FTSE 100 market weighted tracker.
I take the point about it being difficult in the UK to come up with a way of adopting a different split whilst maintaining a passive investment approach. I’m going to check out some of the alternatives highlighted above.
@ivanopinion
“For every £100 you have available to invest in the FTSE 100, would you really put £43 in the FTSE 10 and £57 in the FTSE 90?”
There are three or four reasons not to, but in my opinion only two of them are correct.
1) The FTSE 100 is as we have agreed is over-concentrated in the mega-caps (HSBC, Vodafone, BP etc etc). This can be easily overcome by taking a ‘total market approach’ and investing in FTSE All Share (which is roughly 80% FTSE 100, 16% FTSE 250, 4% smaller). The over-concentration can be further diluted by investing in foreign markets (of which the US is obviously much the biggest). For what it is worth I think this makes sense.
2)The second reason to deviate from FTSE 100 is to take a small cap tilt on the grounds that this will probably lead to greater returns (at the price of greater volatility). I do this to0.
3)The third reason which I think is misguided is to suggest that there is some intrinsic reason why market cap-weighted indexes are inherently ‘inefficient’. This is the argument made by Research Affiliates with their Fundamental Indexing approach. To my mind their arguments have been convincingly and mathematically been demonstrated to be wrong. (RAFI make a make a strong claim that their indexes offer much more than just a value tilt. They claim to be able to avoid an inherent arithmetically driven weakness in market-cap weighting)
4) The fourth reason (that I can think of) would be to take a value tilt either using RAFI or not. I would think about doing this but the danger is that you can end up actually being even more concentrated in very large stocks.
I think a lot of the hype over alternative indexes (fundamental, minimum volatility) is pushed on investors for the wrong reasons. The investment industry is extremely worried about market-cap index investing among private investors. It is the cheapest possible form of investing and with Vanguard charging 0.15% in UK and less in the US their profit margins are evaporating.
This is a very large driver indeed for the industry to come up with alternative indexes.
It seems very unlikely to me that the extra (?0.3-0.5%) cost of alternative indexes is worth the POSSIBLE extra returns. All the alternative indexes have sometimes under-performed market-cap indexes for very long periods on back testing and their returns going forward are likely to be less anyway. (There is always a back-testing bias, and the market is likely to erode future returns over time.)
I plan to stick with cheapest possible market-cap weighted indexes but to dilute out the FTSE 100 mega-caps by diversifying as noted.
Hi PI,
If we’re agreed that the returns from fundamental investing are value returns then there’s no need to rely on back-testing (though I appreciate that new fund ideas use this to make their case).
But you can go back as far as you like, using real data, to see that the value premium exists. A fundamental index may not be the best way to capture it, but we’re kinda low on options here. So far the market hasn’t snuffed out the value premium and there’s no way of knowing whether that premium will wax or wane in the future.
The evidence for the size premium is weaker than the value, so the diversification argument applies more strongly to value companies than small companies.
I completely agree with you that the motivation of the industry to dream up new ideas is based on profit margin, but not every new idea is bunk. Just 99% of them are 😉
I think we are basically agreed and (although I might can see I implied differently) I do agree with you that the value premium is likely to be persistent.
If Vanguard or iShares offered a World or Small Cap value physical index tracker, at a good price, I would put perhaps 10% of my stock portfolio in it. (Even so I would be a bit worried that I might not stay the course if there was a repeat of the late 90s bubble.)
My concern with Research Affiliates is that they claim that their recent (and back-tested) out-performance is about more than just capturing Value. I think this is to put it politely highly controversial and it has the effect of weakening investor confidence in market-cap weighted indexes which despite the noted disadvantages have some major advantages (eg very low turnover, representing the view of the capital markets etc).
This article by Kaplan from 2008 “Let’s not all become fundamental indexers just yet” explains it quite well
http://www.nxtbook.com/nxtbooks/morningstar/advisor_2008spring/index.php?startid=23
All best
Good article. Thanks for sharing. I’ve seen references to Arnott climbing down on his claims for fundamental indexing and admitting that in reality he’s just capturing the value premium.
I say ‘just’ – capturing the value premium is good enough reason for me to use fundamental indexing, though I can see why the earlier hubris puts you off the brand. I’d be happier if Vanguard released some of their broad market value funds over here. One day.
Using the FTSE All Share helps a little bit with overconcentration, but I don’t think it is a full solution. If 80% of the All Share is the FTSE 100, and 43% of the FTSE100 is the FTSE 10, then putting £100 into a FTSE All Share tracker is still equivalent to putting £34 into the FTSE 10.
I don’t really see investing in foreign markets as being a solution to this, because I tend to adopt a top-down strategy to allocating my portfolio. I decide first of all how much of my portfolio to allocate to the UK, versus North America, Asia, Europe, and emerging markets. But I still don’t want to put 43% (or even 34%) of my UK portion into just ten UK companies.
I have the same problem with using global trackers. The Vanguard FTSE Developed World ex-UK Equity Index fund is heavily dominated by US equities, because that’s where most of the world’s biggest companies are. 58% of this fund is invested in North America, but I don’t want to put more than half my portfolio in one region of the world. Fortunately, this is easy to correct by buying different proportions of regional trackers.
I’m not sure that the RAFI funds are a solution to overconcentration, either. I just had a quick look at the FTSE RAFI UK 100 fund and more than 50% of it is invested in just 10 companies, so it is even more over concentrated than a market weighted tracker. To be fair, I don’t think they claim to be an antidote to over concentration.
On the other hand, another recommendation made above was Dimensional. Their UK Core Equity Tracker explicitly sets out to underweight the large cap equities. This has a 35% concentration in its top 10 holdings, which is better than 43%, but still pretty high. I would really like to see someone like Dimensional take this a step further and produce a tracker with equal weighting in each of the FTSE 100 companies.
Then again, if you are just going to put 1% of your portfolio (or the portion of your portfolio that you’ve allocated to the UK market) into an equal amount of each of the 100 companies, then that’s a strategy that is much simpler and cheaper to operate than a market weighted tracker, because there would be no need to rebalance in reaction to swings in value of each stock. You might want to rebalance periodically, if some stocks fluctuate massively and become worth significantly more or less than 1% of your portfolio. But there is nothing magic about the 1% figure, so I think this could be fairly infrequent. This would mean that trading costs would be pretty low. I wonder how big your investment would have to be before it would be cheaper just to buy the underlying stocks yourself.
If you shop around, you can get trading cost down to about £6 per transaction, so buying 100 stocks would cost £600. Let’s assume you end up rebalancing 10 of them per year, so that would be 10 sales and 10 purchases, so a further £120 per year. Average cost over 10 years, £168. Even if you had £100,000 to invest in this self managed FTSE 100 equal weighted tracker, that works out at annual costs of 17 basis points, which is more than even some market weighted FTSE 100 trackers. You would think that a fund with the same strategy could get the annual costs down far below this.
By and large mkt caps refelct the underlying fundamentals of the companies. Even Fundamental Tracker funds have a large concentration in the big caps.
The reason is that the main driver of equities is dividends and the largest dividend payers are the big caps. This demonstrates that the business logic behind consolidation actually works.
It is just that in the short term they can be overshadowed by whatever is currently flavour of the month.
Since QE started 4 years ago grwoth stocks have been favoured at the expense of value stocks, as Buffet has so elonquently pointed out.
@ Ivan – The problem is how do you decide when an index is too concentrated? If you believe the market is putting too much store in America, what do you know that everyone else doesn’t?
Equal weighting just concentrates your risk in different ways.
Equal weighting was tried in the early days of indexing and was found to be problematic. High turnover and high costs.
Still, no point hankering after a fund that doesn’t exist. There are plenty of other ways to diversify as outlined above.
@ivan @accumulator
http://www.rickferri.com/blog/investments/no-free-lunch-from-equal-weight-sp-500/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+RickFerri+%28Rick+Ferri+Blog%
Here is Rick Ferri making a similar point about equal weight today. May be of interest
@Accumulator – I accept there is no objective measure of when an index is too concentrated. But the FTSE 100 does seem to be quite extreme. The top 10 holdings in the Vanguard US Equity Index fund represent only 15% of the fund, compared with 43% for the FTSE100. So I’m ready to make the subjective judgement that the FTSE 100 index is too concentrated. So is the FTSE All Share, with 34% in the top 10 holdings. I’m not so worried about the S&P Total Market Index, which is what is tracked by the Vanguard US Equity Index fund.
And in relation to the market cap of the US market versus the rest of the world, I think this reflects many factors in addition to investors making top-down decisions about how to allocate their portfolio across the different regions in the world. For a start, many investors have a huge inbuilt bias to their home market. The UK market is approximately 10% of the capitalisation of world stock markets, but most UK investors have considerably more than 10% of their portfolio in the UK market. Even in New Zealand, most investors primarily purchase NZ stocks, although this is a fraction of 1% of world markets. Similarly, most US investors will have a huge bias towards the US market, regardless of its share of world market cap. Therefore, to some extent, the market capitalisation of the US market is simply a measure of the capital available for investment by US investors.
I accept that judgements about likely returns and risks of the US market versus other markets are part of the explanation of the market cap of the US market, and I would not presume to second guess the wisdom of the crowd on that. But I don’t think it is the only factor explaining why the US market is more than 50% of world indices and I’m not going to slavishly follow market cap as my only guide to geographic asset allocation. (I suspect very few of us do.)
In many ways, the US market is hugely diversified, but it is still heavily dependent on the fortunes of the US economy and if things start to go badly for the US, most US companies will be adversely affected and so the US market as a whole would perform worse. I don’t want to bet 58% of my portfolio on the long-term performance of any one country. (However, I will probably put more of my portfolio into this region than any other region, because market capitalisation should not be ignored altogether.)
I do agree that equal weighting (even if it were available for a UK tracker) has its own problems. I’m not sure I can see the need to rebalance quarterly, which is what Ferri says is done by equal weight trackers. Why not rebalance only when the weight of a stock has fluctuated by a given percentage over or under the starting allocation? This would cut costs considerably.
I take your point about other ways to diversify. Certainly, the fact that more than 7% of the FTSE 100 is HSBC should be more of a concern for the investor whose only investment is in the FTSE 100, in comparison with, say, someone who has most of their portfolio invested in other regions/countries.
Isn’t it reasonable to have a home bias, provided you expect to stay in your home country, because of the foreign exchange risk?
It may well be, but what I’m questioning is whether it is reasonable to let the market capitalisation that results from the home bias of US investors influence me (as a UK investor) to put 58% of my portfolio into the US market.
More ammunition for the debate
http://www.etfstrategy.co.uk/david-stevenson-on-smart-beta-etfs-including-high-dividend-etfs-and-low-volatility-etfs-45269/
Being influenced by the verdict of the market is at least as reasonable as trying to guess what effect ‘this factor’ or ‘that factor’ may be having – which is nigh on impossible. It may be less fun but it is simple and it is reasonable.
So are you saying you have more than 50% of your portfolio in US equities?
And only 10% of it in UK equities?
@acc @ivan.
I follow and would argue for a half way position. Foreign equities make sense because they offer diversification, some degree of non-correlation (though not enough when you need), access to more dynamic economies than the UK and I agree that ‘the wisdom of the market’ is not to be dismissed lightly.
The disadvantage is exchange rate risk. Over the long- term this is actually fairly low but the trouble for the near-retiree is that you could hit a major appreciation of sterling in say your last 5 years of accumulation / first five years of draw-down. If that happens it won’t be much consolation to know that ‘in the long run everything will be all right’.
The obviois compromise is to put 40-60% in the UK and the rest in non-UK markets. Hopefully this gives the best of both worlds.
All,
I didn’t realise that this debate was still continuing but interesting it sure is!
Having read all of the above comments, I’m still inclined to stick with simple market cap. trackers for the bulk of the equity assets within my portfolio but will, when I next rebalance, tweak it with a nod in the direction of value and small caps. However, I am aware that if this “nod” is, say 20%, of my equity assets, and has an out-performance of 2.5% (after additional charges), then the extra performance of my equity assets as a whole will have been enhanced by, you guessed it, .5%!
Regarding the weighting towards the US, I think I’d have to look at factors like the turnover, dividend, profits of US companies compared to the rest of the world to help decide how much exposure you feel is appropriate. Also we need to remember that a big chunk of US companies’ business is generated overseas so you’re not just betting on the health of the US economy. But, yes, currency exposure can be a risk.
Regarding the over-exposure to FTSE top 10 companies resulting from a market cap. tracker, yes it does mean that you get caught up in market bubbles but also, presumably, you also benefit from momentum. If there is a market bubble (anyone expecting that in the next 5 years?), then disciplined rebalancing between equities and other asset classes and between the UK and overseas equities will help reduce the impact of the bubble bursting.
I would argue that is possible to recognise when equity markets are in “bubble territory”, even at the time. The difficulty, is jumping off the bandwagon early, watching it carry on moving, and enduring the scorn of any acquaintances who know what you’ve done! Bubbles can last a long time with the most extreme case being the UK residential property market.
“Jumping of the bandwagon early” is of course not passive unless it happens as part of your passive rebalancing strategy. Which leads me to wonder whether limiting market sector allocation should also be part of a passive strategy and, if so, how it could be implemented for UK investors.
It’s a pity that also there aren’t FTSE 100 decile trackers (top 10, 11-20, 21-30 etc), even if cap. weighted, which would allow a passive investor to very simply restrict exposure to a hand full of elephant cap. companies as they see fit.
I’m deliberately not attempting to quantify anything (e.g. maximum sector exposure, maximum exposure to one company, maximum exposure to top 10 companies) because different investors would want different targets but I do feel there is a place for a variety of simple cheap cap. weighted trackers for passive investors wishing a greater level of diversification.
@ Grumpy – good thoughts. One of the things that’s becoming apparent from the debate on this thread is that we’re a bunch of very committed ‘passive’ investors who like to finesse. So I think we implicitly accept a higher than average level of complexity in our portfolios.
But think one of the great selling in of passive investing is its simplicity. Anyone worried about the concentration of the FTSE 100 or All-Share would be well served diversifying across the standard asset classes – international, emerging markets, fixed income, real estate, maybe commodities, maybe small cap and value.
After that you could overweight into a FTSE 250 fund. There is no real need to go any further except because you love it. As we clearly do.
@ Ivan – it’s tough to commit 50% of your portfolio to the US when there are all the asset classes above to fit in plus a home bias because like PI and PC I am concerned about currency risk.
I am 25% UK – diversified across market cap, small cap and value.
I am 50% international – same split.
But my market cap and small cap international funds are both 57% US and I don’t worry about it at all.
I’m well diversified across the asset classes and sources of return and what metric would I use to change the verdict of the market into a more rational decision?
As Grumpy says, if I could see a massive bubble inflating then I’d be prepared to take action, but it’s easier said than done. I wonder how many investors got out of Japan in 1989 or tech stocks in 1999?
http://www.thereformedbroker.com/2013/05/06/fundamentally-weighted-vs-cap-weighted-indices/ yet another article on cap weighted. v fundamental indexes
Dimensional has popped up on TD Direct.
http://www.ossiam.com/index.php/solutions/ossiam-etf-stoxx-europe-600-equal-weight-nr#
Thoughts?
TER of 0.35%. The best physical market cap index is 0.3%(HSBC) and the artificial dbx at 0.2%.
Its artificial.
Only started in May 2012 so no indication of how good the tracking is. Total of about £60 million so not promising for tracking economics of scale.
Nearly non existent volume but authorised participant should keep the spread low.