Is there an efficient frontier for including smaller companies alongside an index fund?

aroominyork

The efficient frontier is usually discussed in equity/bond allocations, and says that although equities are higher risk than bonds, an 80/20 portfolio not only has better returns over time than 100% bonds, but carries lower risk. The shape of the frontier (the point of least risk) has varied in each decade over the last century, with the 80/20 curve representing an average picture.

Recently I invested about one-sixth of my US allocation in a US smaller companies fund, not to boost returns but to give me some diversification from the FANG-dominated index. I then wondered whether an 80/20 index/smaller companies equity allocation would provide lower overall risk than an 100% index fund.

For the US, UK, Europe and Japan I used Trustnet’s portfolio tool to combine:

-          80% of Fidelity index fund

-          20% of the smaller company fund which is the FE median in the sector over three years. For example, there were 21 funds in US smaller companies on a 3 year view; I then ordered them from high to low FE and selected number 11 in the list.

This is what I found.

US:

Fidelity US Index, FE87

Brown Advisory US Smaller Companies, FE131

80/20 portfolio, FE82

 

UK:

Fidelity UK Index, FE93

Premier Miton UK Smaller Companies, FE102

80/20 portfolio, FE88

 

Europe:

Fidelity Europe ex-UK Index, FE86

Schroder European Smaller Companies, FE97

80/20 portfolio, FE86

 

Japan:

Fidelity Japan Index, FE83

M&G Japanese Smaller Companies, FE110

80/20 portfolio, FE86

 

So this shows that for the US and UK, an 80/20 portfolio is lower risk than an Index fund.

For Europe, the risk is the same.

For Japan, the risk is a fraction higher.

 

It would take more work to look at performance since the median risk smaller companies would not necessary deliver average/median performance, but I thought I would post this and see what others make of it.

masonic

Well done on doing the analysis. It's something I'd assume would be true. I dare say returns from smaller companies will be higher in most regions, but in the US smaller companies have tended to struggle to outperform the S&P500 (though there was a fantastic buying opportunity at the bottom of the Covid crash).

aroominyork

Just realised it's easy to run three year sector graphs on HL so here they are. Surprisingly little performance benefit in small caps, though I do not know how the sector performance is calculated. Still, the point about small caps reducing volatility is interesting.

masonic

aroominyork said:

I do not know how the sector performance is calculated

I think it is just the mean 3 year performance of funds within the respective sectors (i.e. the sector average). It will include the good, the bad, and the ugly of funds included in each sector. However, the results do seem to bear out the prevailing wisdom - that small caps are considerably better in performance in the UK, quite a bit better in Europe, and less so elsewhere.

Thrugelmir

aroominyork said:

The efficient frontier is usually discussed in equity/bond allocations, and says that although equities are higher risk than bonds, an 80/20 portfolio not only has better returns over time than 100% bonds, but carries lower risk. The shape of the frontier (the point of least risk) has varied in each decade over the last century, with the 80/20 curve representing an average picture.

Have you asked yourself the question why diversified funds, such as multi asset,  have historically held bonds?  No need to reinvent the wheel. 

aroominyork

masonic said:

aroominyork said:

I do not know how the sector performance is calculated

I think it is just the mean 3 year performance of funds within the respective sectors (i.e. the sector average). It will include the good, the bad, and the ugly of funds included in each sector. However, the results do seem to bear out the prevailing wisdom - that small caps are considerably better in performance in the UK, quite a bit better in Europe, and less so elsewhere.

I'm not sure about Europe. I have bought and sold Barings Europe Select twice and, if I could not control my itchy fingers, might easily do so again. For the UK - yes, I have a few % of All Companies but my overweight is all smaller companies.

bostonerimus

DFA might have some analysis, but they won't be sharing it. If you put small caps an an efficient frontier I'd expect the spread of the data to be far greater and your SD axis would have some very big numbers.

jamesd

It's drawdown-specific but it might be of interest to read Small-Cap Withdrawal Magic which found a significant increase in safe withdrawal rate by changing from 50:50 to 35% large cap, 20% small cap and 45% intermediate duration bonds. Don't miss page two.

Small cap has done relatively poorly compared to long term in the US due to the success of some very large companies. Even so, there's some benefit, showing up more over five years than three.



In performance order they are Light blue: ASI global smaller companies, green: ASI UK smaller companies, yellow: UT North America smaller companies, darker/brighter blue: UT North America, Brown: FTSE small cap ex investment co, Red: FTSE All share excluding investment co. The ASI funds are two that I've been using for many more than five years in a small cap heavy portfolio. Notice the considerably greater small cap benefit in the UK comparing yellow and brown with red.

Vs three years:

maxsteam

aroominyork said:

The shape of the frontier (the point of least risk) has varied in each decade over the last century, with the 80/20 curve representing an average picture.

Haha. I bet you are a treckie. 👍

If you are looking for least risk, you should stick to bonds and definitely avoid small companies.

If you are looking for a level of risk that is acceptable to you, then there is no magic number. A risk that is acceptable to you will not be acceptable to another person.

Your post seems to make the false assumption that risk can be avoided by dividing your portfolio into certain ratios.

aroominyork

bostonerimus said:

DFA might have some analysis, but they won't be sharing it. If you put small caps an an efficient frontier I'd expect the spread of the data to be far greater and your SD axis would have some very big numbers.

If you look at FE scores for pan-cap vs. small cap I think you will be surprised. Small cap funds are not necessarily the most volatile.

maxsteam said:

If you are looking for least risk, you should stick to bonds...

No, that's the whole point of the efficient frontier data. Although bonds are less risky than equities, an 80/20 portfolio is less volatile than a 100% bond portfolio.

maxsteam said:

 ... and definitely avoid small companies.

I've just presented data that contradicts that. Look at the data instead of what you assume 'must' be right.

maxsteam said:

Your post seems to make the false assumption that risk can be avoided by dividing your portfolio into certain ratios.

Reduced, not avoided. It's called diversification.

Thrugelmir

Bonds in more recent decades of portfolio management have provided diversification. In that when equities fell , bonds rose in considerably in value. With the auto rebalancing of funds such as VLS60 and 80 ( in the USA).  "Expensive" bonds were exchanged for "cheap" equities. A forumula that worked for some decades. Now that link is irretrievably broken. The trade is over. 

For a portfolio to be diversified. The holdings need to be uncorrelated.    If many of the US small caps you've invested in are Apple suppliers then there's far less diversification than initially appears to be the case. Likewise your portfolio will be impacted by wild swings in the £- $ exchange rate or changes to US corporation tax rate policy.