Odd asset allocation. 20% UK 17% property 23% bond

kidmugsy

In his writings John Kay repeatedly makes an excellent point. Your concern should not be with the risks of each different investment, it should be with the risks of the whole portfolio. So volatility may not matter anything like as much as correlations - if A tends to rise as B falls, then a portfolio of A & B will be substantially less risky than either component individually.

System

kidmugsy wrote: »

In his writings John Kay repeatedly makes an excellent point. Your concern should not be with the risks of each different investment, it should be with the risks of the whole portfolio. So volatility may not matter anything like as much as correlations - if A tends to rise as B falls, then a portfolio of A & B will be substantially less risky than either component individually.

He is an economist so he should understand the capital asset pricing model (CAPM)!

Audaxer

kidmugsy wrote: »

In his writings John Kay repeatedly makes an excellent point. Your concern should not be with the risks of each different investment, it should be with the risks of the whole portfolio. So volatility may not matter anything like as much as correlations - if A tends to rise as B falls, then a portfolio of A & B will be substantially less risky than either component individually.

That's true and if you get asset classes that are individually volatile but move in different directions at different times, then that is where I think your returns can benefit a lot from rebalancing. The bit I would find difficult if picking a growth portfolio of single sector funds, is being confident in getting the right percentages in Emerging Markets, Small Companies etc., and the right mix of bonds and alternatives for a well structured portfolio.

kidmugsy

Economic wrote: »

He is an economist so he should understand the capital asset pricing model (CAPM)!

Or: perhaps he is enough of an economist to dismiss the CAPM.

kidmugsy

Audaxer wrote: »

That's true and if you get asset classes that are individually volatile but move in different directions at different times, then that is where I think your returns can benefit a lot from rebalancing. The bit I would find difficult ... is being confident in getting ... the right mix ... for a well structured portfolio.

He's pretty good on this sort of thing. For example he points out that buying different companies is a pretty weak diversification if the companies you pick largely sell to the same consumers.

What you want are investments that will react differently to the trends and shocks that might happen.


(P.S. Last time I read him he dismissed fixed interest bonds altogether - far too expensive.)

Audaxer

dunstonh wrote: »

If you think of a something being tugged in different directions by different things. Asset mix, currency risk, political risk, industry risk etc etc.

If you tug a bit harder in one direction then you have to give a bit more in the others. That is effectively what volatility based models do. They may say that by having a bit more in UK equity, they can a bit less in bonds and a bit more in global equity. If they go heavier in higher risk global equity (such as emerging markets, including asia) they would then have to go heavier in lower risk assets to compensate. This could see the middle range lose out.

There are usually constraints placed on maximum allocation to each area to prevent rather strange models for appearing. I have the software that allows this to happen and when you remove constraints, you get some quite unrealistic allocations.

Most treat UK equity as having lower volatility over the long term as historically that was the case. Going forward post globalisation, it is debatable whether that is the case or not. But then in day to day performance, bar any localised events, currency fluctuations do account for a lot of the volatility.

Thanks dunstonh. It would be good if that sort of software was available to DIY investors so we could get a better idea of what percentage and mix of different asset classes is likely to produce the best returns for different risk levels.

System

kidmugsy wrote: »

Or: perhaps he is enough of an economist to dismiss the CAPM.

Why would he dismiss the CAPM when the 'excellent point' he made is an insight from the CAPM?