Smaller companies within a wider portfolio

bostonerimus

If you include Mid-cap as well then 20% of my equity portfolio is Mid/small/micro cap

george4064

My ISA is about 20% small cap.

pip895

UK Small cap & VCT's make up 10% with another ~10% in other small cap funds. However many of my other active funds are also skewed toward small cap.

Morphoton

Interesting. So you are saying that a lower equity allocation with a high % of smaller cos, provides better returns over time than a higher equity allocation focused on large caps. Do you have any stats/articles (good ones!)/evidence to demonstrate this? Also, have you looked at the minimum period of time to make this a productive strategy taking into account risk associated with the increased volatility?

I do not have any formal proof that my barbell strategy is any better than others and my strategy has evolved over the past 30 years or so. It was really just reading articles that smaller caps tended to outperform large caps, and it seemed to make sense that it would be likely that there would be lower correlations between countries for small caps vs large caps. I was not too concerned about volatility. In fact volatility can sometimes help with regard to re-balancing / pound cost averaging. Having been through the crash of '87, the tech crash and the last one and seen some peak to trough drops of individual investment trusts of ~80% you somehow get inured to them. However despite these drops I have had the best long term returns from these trusts. For example I have held both TR European Growth IT (TRG) & Baillie Gifford Shin Nippon (BGS) for approx 20 years.

There are various articles that show smaller caps have tended to outperform large caps over the long term.
For example, comparing the various UK sized rank returns since 1955. I have also read that this applies to other countries.
UK Data:
https://www.london.edu/news-and-events/news/uk-small-and-mid-caps-outperform-in-2015#.WlFLyNZFDqA

With regard to correlations there is certainly some US research such as from MSCI at:
https://www.msci.com/www/blog-posts/why-are-small-caps-different-/0736742719

If you are very mathematically minded there is a 45 page paper PDF in the link below. This data is fairly old from the 80's and 90's.
If you do not want to wade through it all you can just read p1 or the following quote from p5:
"As a result, small-cap funds have relatively low correlations not only with large-cap funds but also with each other. In contrast, large-cap funds tend to have relatively high correlations with each other, reflecting their common exposure to global factors. During our sample period, for instance, the correlation between the U.S. and Netherlands large-cap funds is 0.61, whereas the correlation between small-cap funds from the two countries is only 0.17. Further, the correlation between the U.S. large- and Netherlands small-cap funds is 0.21."

https://www.google.co.uk/url?sa=t&source=web&rct=j&url=http://www.cicfconf.org/past/cicf2006/cicf2006paper/20060201092639.pdf&ved=0ahUKEwjisvX7psTYAhWJh7QKHWduAe4QFggsMAQ&usg=AOvVaw2w_SjjjJpityPdf6rNt6Bk

With the above information it just seemed to make sense to me that with a high level of smaller caps in my equity portion I would hopefully benefit from the out-performance effect and the correlation effect.

What I cannot say with any certainty is what the optimal level of large caps / small caps / bonds would be with regard to returns and volatility.

The biggest caveat to all the above is of course that past performance is no guarantee of future performance !

TheTracker

aroominyork wrote: »

Interesting. So you are saying that a lower equity allocation with a high % of smaller cos, provides better returns over time than a higher equity allocation focused on large caps. Do you have any stats/articles (good ones!)/evidence to demonstrate this? Also, have you looked at the minimum period of time to make this a productive strategy taking into account risk associated with the increased volatility?

Enormous amounts of material on this, the small premium was an earlyish piece of modern economic research and theory that remains somewhat solid. Banz found there was a size premium - historically smaller stocks achieve higher returns than larger stocks, albeit at higher volatility.

Further, that returns were higher even when ‘risk adjusted’, that is the higher return was not just due to taking more risk. So yes there is more volatility, but not quite as much as would be predicted for the increased return. Fama and French famously then studied it and included in their 3 factor theory, alongside a value premium and general market beta. I don’t think they addressed the risk adjusted part, but further studies found the same as banz.

Since then there has been much research, and as always in semi-scientific economic theory, much is in dispute. The premium disppears when you cut the data in certain ways, such as time or geography. An interesting idea is that such factors disappear when well known / discovered (they get incorporated in general market returns rather than disappear), though evidence points both ways. Another idea is that size/small is a proxy for some other aspect. Eg, perhaps it arises from illiquidity of very small and micro stocks, so liquidity might be a better factor. Various 4 and 5 factor theories exist with good support. They introduce factors like momentum and profitability.

There are all sorts of ‘wars’ around passive/active. But we can dissolve a lot of that by realising most academics suggest that various fundamental factors account for 90-95%+ of performance, and that portfolios should allocate assets based on diversifying across these factors. As such, if the small premium exists and has higher risk adjusted returns, then one can construct portfolios where “a lower equity allocation with a high % of smaller cos, provides better returns over time than a higher equity allocation focused on large caps”.

Any star manager aspect, or alpha, is about the remaining 5-10%, at best, and very probably through some inherent style that exploits factors that theory hasn’t unearthed rather than a timing ability. This is where ‘smart beta’ comes in, the idea that one can algorithmically exploit different factors in different conditions, and why they exist between ‘active’ and ‘passive’.

In terms of time, my understanding is you need to stick with it 20-30 years to have a good chance of the size premium showing above randomness/volatility, and that’s if it exists at all. Because of this, many people quite validly say Pah! and don’t bother with the faff of deliberately targeting/overweighting small stocks.

Linton

The problem with academic research is that most of it seems to be carried out on US data. The 5year data In my post #10 does suggests that the high return from small companies has not been apparent in the US. Whereas it certainly has elsewhere.

Looking at Trustnet charts small company outperformance seems to be only apparent since around 2000. I can think of good reasons why this may be so, for example, prior to then, before internet access became mainstream, funds would have been mainly bought through advisors and brokers and far fewer funds were available. It is possible that small company outperformance will fade. But one surely can’t deny its has been highly significant in the past 20years or so.

A_T

I believe that the US market is so well capitalised that finding outperformers (including small cap) is extremely hard. Hence Warren Buffet winning his bets that the S&P500 will beat stock pickers over long periods.



I would not bother with US small cap but I do look at UK, Japanese and European mid and small cap.

aroominyork

I want to paste a graph in this thread. Is it possible to paste a graph which I have copied into a Word doc?

bostonerimus

TheTracker wrote: »

In terms of time, my understanding is you need to stick with it 20-30 years to have a good chance of the size premium showing above randomness/volatility, and that’s if it exists at all. Because of this, many people quite validly say Pah! and don’t bother with the faff of deliberately targeting/overweighting small stocks.

I couldn't have said it better myself...Pah indeed!

Filo25

A_T wrote: »

I believe that the US market is so well capitalised that finding outperformers (including small cap) is extremely hard. Hence Warren Buffet winning his bets that the S&P500 will beat stock pickers over long periods.



I would not bother with US small cap but I do look at UK, Japanese and European mid and small cap.

If anyone thinks the PE on the S&P is scary, it is a lot higher on the Russell (reflecting the higher growth prospects).

You don't really see too much of a premium in comparison on the UK though when I have looked at it, which is surprising given how strongly prices for UK small companies have performed, which is one of the reasons I am happy to have Smaller Companies make up a reasonable proportion of what I hold in the UK.