Current market carnage - anyone selling or buying?

N1AK

Linton wrote: »

A bit of academic research for you - if you must invest in a tracker dont make it a cap-weighted one. See here

Now where do I get a good range of non cap-weighted trackers from?

It's an interesting document, and it's certainly got me considering things. There's a couple of points that mean I'm not going to rush out to sell my trackers just yet:

1. As is shown by their Scrabble index past-performance isn't an indicator of future returns. Scrabble stayed very close to the market cap index for 40 years, then managed to double the value of the other index in the following 15 years. Perhaps Scrabble will lose all that excess value in the next 15 years... I'm going to be doing some further investigation of alternative indexes and how they performed over regular intervals.
2. It provides no information on the methods used in the alternative indexes. Fundamentals for example can be based on anything from Profit to number of employees in theory.
3. It assumed that the market cap tracker would value companies on a yearly basis which has issues both in making timing more influential, and on making divergence from the market during the year a factor. This isn't realistic as market trackers are constantly re-adjusting.
4. They used a merged sample of three American indexes. I'm not sure how reliable that is as guidance for all markets.

I'm certainly open to the concept of using smarter indexes, the issue in practice is finding one that is likely to outperform the market in the future...

TheTracker

bowlhead99 wrote: »

And if your active manager of choice employs strategies which the "evidence based investor" observes that Cass found likely to deliver better outcomes, "simply, their bias toward small and value stocks — factors which have been shown to deliver higher returns, albeit with greater volatility, over the long term"... then it should be quite trivial to be able to pay a manager a fee and have him deliver a result better than the cap-weighted index over a long term.

It is interesting/refreshing/bemusing, some may say confusing/despairing/worrying, that a single article can be read in many ways to draw opposite conclusions. Politics and investing are never far apart.

It is commonly understood that over the period of 10+ years "active beats passive". Usually, that means the average active fund (which riles active investors, since they jump from fund to fund over longer periods), versus big benchmark indices (which riles passive investors, few of which stick to big names like the FTSE100). But mathematically, it is trivial to show the average actively managed pound returns less than the average passively managed pound.

I think the point EBI is making is that since average passive investing, fundamentally built on cap weighted indices, beats average active management which has no cap weighting constraint, then if you measure active against a non-cap weighted benchmark (random, monkey, scrabble) the difference is even more stark. It's a fair comment. Active funds should beat and be measured against random funds. On the other hand, the cap weighting is a by product of what all the active funds are investing in, so perhaps not.

The second point is that non-cap weighted indices seem to work well because of underlying factors or "smart betas" like small, value, momentum, volatility, liquidity. So the challenge is that actively managed funds should be benchmarked against relevant passive vehicles. I don't think that gets enough attention. I like the deconstructions Larry Swedroe performs on etf.com where he shows fund managers who publicise how much they beat the s&p500/ftse100 but simple analysis against factor indices shows underperformance.

So no, I wouldn't give an active manager money to expose my investment to a factor/style unless they could show that after costs they were beating a factor index reliably.

Back to the article. It's one of those clickbait studies that shouldn't be taken too seriously. It's perfect timeless fodder for telegraph journalists to file for publication months before they take their winter holiday in the sun. And I am checking out from further comment on this one, until the next.

Linton

TheTracker wrote: »

........

I think the point EBI is making is that since average passive investing, fundamentally built on cap weighted indices, beats average active management which has no cap weighting constraint, then if you measure active against a non-cap weighted benchmark (random, monkey, scrabble) the difference is even more stark. It's a fair comment. Active funds should beat and be measured against random funds. On the other hand, the cap weighting is a by product of what all the active funds are investing in, so perhaps not.

No No and no again. The investor now has to invest in the funds that exist now. The choice for the UK ISA investor is active of many different types or cap weighted passives. (you can counter this with IUKD if you want but I wouldnt go there if I were you!). In the ideal world perhaps we would have Smart Beta type trackers, running with some human intervention to avoid the obvious howlers that happen mathematically to meet the criteria but have serious flaws. However in the real world it doesnt matter if an active fund would perform badly agains a non cap weighted passive if one cant actually invest in a non cap weighted passive.

Linton

TheTracker wrote: »

.....
. But mathematically, it is trivial to show the average actively managed pound returns less than the average passively managed pound.

.....

Only if you make questionable assumptions. Look at the 10 year data for the UK all companies sector. There isnt much evidence that the trackers are rising up the performance list. Suggest you try the following...

1) Go to trustnet
2) Select the IA UK All Companies Sector
2) Select the "custom" options and tick cumulative 10-yr and then exit
3) Sort by 10 year performance
4) At the top you will see the "IA UK All Companies Index" 10-year value is 58.1%.
5) Search down the fund list to find the highest all-share tracker. The highest I could find was the Scot Mutual All-share tracker at 54%. There are quite a few all share tackers at 40-50% return.

Surely after 10 years there ought to be any inherent advantage of an all share tracker should have become apparent.

bowlhead99

redux wrote: »

It observes an interesting phenomenon, and partly explains it, namely that their Scrabble method tends to turn towards smaller companies which historically outperform the largest,

Yes you're right, Scrabble would produce a more balanced portfolio not so heavily weighted to the biggest companies as the typical cap weighted SP500 or FTSE100, which should give it a long term advantage because on average your money is deployed relatively more heavily into the smaller companies.

Actually, the EBI piece doesn't properly explain this at all. It notes the researchers conclusions that the alternative index constructions benefited from their favouring of value or size when compared with cap weighting, and then the EBI writer says he doesn't understand why Scrabble index would do well because he's not aware of Scrabble being related to a size factor....

The researchers concluded that the principal reason the alternative indices delivered better outcomes was, simply, their bias toward small and value stocks — factors which have been shown to deliver higher returns, albeit with greater volatility, over the long term.

Of course, that doesn’t explain the success of the Scrabble index; at least, I’m not aware of any greater tendency among small or value stocks to have Xs or Zs in their name. That result, perhaps, is one to put down to random chance.

The EBI writer is bemused and puts it down to random luck.

However, it's easy to see that a Scrabble weighted index exploits the so called "size factor" compared to cap weighting.

If you look at cap weighting of S&P500, the $600bn Apple or $300bn Exxon have 50-100x the weighting of the smallest company allowed in the index at $6bn. If you look at FTSE 350 the £90bn HSBC has 150x the weighting of the £600m Lookers group. So, the asset allocation is skewed particularly heavily towards massive behemoths.

If you allocate based on Scrabble scores of two-to-four digit ticket symbols the money is allocated haphazardly based on distribution of the different letter scores in the name. The ZZ of Sealy group or QQ of Quinetic gets 20. AA only gets 2. But you don't get a score that's 50x or 100x or 150x another.

So relatively speaking it is closer to an equally-weighted index which means that even if Apple or Exxon or Shell or BP or Glaxo still had the biggest scores, they wouldn't have the biggest scores by so much of a differential, and your money would be allocated to relatively smaller companies than it would be in a cap weighted all share.

And of course Apple and HSBC and Shell do not have the largest scores. Without looking them up, on average they will get an average score and get an average amount of money. In a situation where your big companies can be expected to get an average amount of money rather than a massive amount of money, and your small companies can be expected to get an average amount of money instead of a tiny amount of money, you are effectively "overweighting" companies with relatively small market capitalizations, compared to the "normal" cheapo cap weighted tracker.

So, Scrabble mixes things up a bit but succeeds in getting rid of the massive differential between highest weight and lowest weight and focusses relatively more money in non-giant companies. Smaller companies are acknowledged to deliver outperformance in the long long term. So as for "why would it out perform", i think QED.

Presumably this "more equal-weight" allocation requires a rebalance every 6-12 months, which a cap weighted index does not need (ignoring new entrants) so there is a small cost implication of additional brokerage fees. Or a large cost implication if the manager is being "active" and researching Scrabble scores manually instead of using a computer lookup, to justify getting paid enough to buy his yacht. But such costs may be offset by the extra returns from following the more lucrative long term strategy as compared to pure S&P or FTSE tracking.

The Scrabble technique is obviously not necessarily any better than what an equal-weight tracker would deliver, but equal weight trackers are not particularly prevalent in the market anyway, so it just joins the list of alternate theoretical indexes which can be dreamt up but are not practical investment options for most

gadgetmind

Linton wrote: »

Look at the 10 year data for the UK all companies sector.

All share trackers will hold only about 15% FTSE 250 and very little smaller companies. We've had a period of out performance for smaller cap versus large cap, and the larger companies in the FTSE 100 have taken a bit of a licking recently.

What's the typical cap split for active UK all companies funds?

masonic

Linton wrote: »

The investor now has to invest in the funds that exist now. The choice for the UK ISA investor is active of many different types or cap weighted passives. (you can counter this with IUKD if you want but I wouldnt go there if I were you!).

Actually, these days there is quite a choice of pretty cheap non-cap weighted passives, e.g. a number of iShares ETFS...

IWFV - World value; IEFV - Europe value; IWFM - World momentum; IEFM - Europe momentum; IWFQ and IEFQ (quality)... All at a TER of 0.25-0.3. None of which are much more than a year old.

I'm in no rush to throw my money at these, but I'll be benchmarking some of my favourite funds against a few of them going forward.

Linton

masonic wrote: »

Actually, these days there is quite a choice of pretty cheap non-cap weighted passives, e.g. a number of iShares ETFS...

IWFV - World value; IEFV - Europe value; IWFM - World momentum; IEFM - Europe momentum; IWFQ and IEFQ (quality)... All at a TER of 0.25-0.3. None of which are much more than a year old.

I'm in no rush to throw my money at these, but I'll be benchmarking some of my favourite funds against a few of them going forward.

Are those US based? Can they be held in ISAs? I cant for example find IEFV on the ETFs list on trustnet, Bestinvest. or iii.

masonic

Linton wrote: »

Are those US based? Can they be held in ISAs? I cant for example find IEFV on the ETFs list on trustnet, Bestinvest. or iii.

They are LSE listed ETFs. See...

https://www.ishares.com/uk/individual/en/products/etf-product-list#!type=emeaIshares&tab=overview&view=grouped&f2=50584
(select smart beta)

Some of these only launched earlier this month.

Edit: And they seem to be on iii:

bowlhead99

As mentioned on other threads (or maybe it was this one, can't be bothered scrolling), I have IEFS and IWSZ, following the MSCI Europe size factor and MSCI World size factor indexes respectively. They haven't been running a year yet and only have a few million of assets. Not a huge amount of daily trading in them. But they're out there and may get bigger if the proof-of-concept takes off.

Vanguard also launched a suite of alternate index ETFs last month: value factor, momentum factor, liquidity factor, minimum volatility factor. All UK reporting, Irish domiciled ETFs with fees under a quarter percent.

Personally if I want to tilt to value or dividend yield I'm not sure I trust the entirely formulaic approach used by passive funds and there is not enough history to really understand how good the methodology works compared to some of the active managers with longstanding funds open for scrutiny. But something simple and cheap like "6 monthly rebalanced equal weight" is a reasonable concept IMHO for broad midcap exposure.

Edit- as mentioned elsewhere the size factor does give quite different weightings to individual countries compared to a cap weighted global or regional tracker. For example IWSZ has under 40% US and about 20% Japan from memory. Japan just has a lot of companies in that size range, relatively, and it only gets rebalanced every so often. But people managing their own allocations will always try to look at the "portfolio as a whole" when deciding if their ending mix works from a regional and sectoral exposure perspective, so one can always find ways of incorporating funds one fancies.