In the course of my work I have often come across things which have made me think. It could be the way you have to deal with certain companies or it could be a change in legislation.This week I have done quite a bit of work on asset allocation within portfolios and this has made me think quite a bit.
We are often quoted the Brinson, Singer and Beebower research which says that over 90% of the performance of a portfolio comes from asset allocation and the remaining just under 10% comes from stock selection and market timing.
Accepting this as being the case, it is therefore vitally important that the asset allocation of a portfolio is the best it possibly can be.
Fortunately, there are a lot of people who are far more intelligent than me, who have access to much greater research facilities than me and are paid an awful lot more than me, who should be able to tell me what the ideal asset allocation of a portfolio should be.
However, unfortunately once I started to do my research I found that there are wide differences between the ideal portfolios of one research institution compared to another.
This difference can also be quite large – as was highlighted in the discussions on the forum. Standard Life through their wrap which has model portfolios provided by OBSR, recommends holding approximately 24% in property in a balanced portfolio whereas the APCIMS Private Client Indices only recommend holding say 5% in property and 7IM is about 2.5%
Similarly, with cash, Standard Life recommend holding 0% in cash for any Risk profile above 3/10, Selestia, for their Investment Account is 0% for Risk Profiles above 5/10 whereas 7IM recommend about 10%; APCIMS is about 5%
Surely if asset allocation is so important you would think that most optimum asset allocations would be fairly similar and not have such wide variations?
One of the problems we have as paraplanners is when we look at the asset allocations and do not agree with the allocations suggested. What should we do? The asset allocations have been developed by companies who spend their time doing these things so it shouldn’t really be for us to question their recommendations. However when we see such a wide variation in recommendations between companies it is difficult not to want to make some changes.
I don’t have the answer and I wouldn’t even start to suggest that I could do better than any of the people who are paid to do this but it does make our job difficult when making recommendations to our advisers/planners particularly if they then go to the client and quote Brinson, Singer and Beebower and then say that the asset allocation they are suggesting is the optimum.
{ 1 trackback }
{ 6 comments… read them below or add one }
I think we all agree asset allocation is key when investing for our clients. The allocation is adjusted to suit the client’s risk profile. As long as you have a documented process of how you arrived at this allocation and your compliance department have approved it then you should be able to sleep at night.
The fact that there is such a wide variety of different allocations available from supposedly reliable sources confuses matters. I feel there is no true optimum asset allocation out there, we should just recommend one we are comfortable with which will achieve your clients objectives.
Whoops Martin,
The BSB paper actually says that asset allocation accounts for 90% of the VARIABILITY of portfolio returns, not returns themselves. There was an article in New Model Adviser recently to this effect (in their Mythbusters series). Hence there is no ‘optimum’ asset allocation. We use the Selestia model, and all they do is calculate the most efficient combination of asset classes, assuming you want to include as many as possible in each portfolio, within certain % parameters. I think this is a better approach than OBSR, APCIMS and the like because 1) Selestia’s models take account of the tax treatment of the asset within the tax wrapper and 2) by using a mathematical model, it removes the opinion element of asset allocation. APCIMS and fund managers generally may be great at tactical asset allocation, but how do you as the adviser or client control that? (Implications for TCF under designing portfolios that meet the client’s risk profile). I feel it is safer to have an asset allocation model based on forward-looking assumptions for asset class returns that you can then populate with single sector funds rather than stick a metaphorical finger in the air and think ‘what property etc exposure do I want today?’ What do other people think?
Good spot James. Hadn’t spotted that myself and always normally picking people up on the subtle difference. Being outsourced paraplanners we have to use a whole range of systems but we like the Selestia one as you do. However, is it still the case that the model excludes commercial property altogether from certain product classes (offshore bond I think)?
I do think that risk profiling and asset allocation tools are a huge leap forward from where we were 5 or 10 years ago but I have some concerns:
1. Do the people using them really know what is going on behind the scenes? Could anyone explain why one system produces a different allocation than another for the same client? Are advisers calling the market and the systems by changing the model and can they justify it?
2. I think a different approach has a strong case. Rather than producing a model asset allocation based on the client’s identified risk profile, should the asset allocation be based on the risk required? I prefer to work out what the client needs his money to do by producing a cash flow forecast which will then show me the rate of return required and it is from this that the asset allocation is built. If a client is identified as being ‘balanced’ he may well get a ‘balanced’ asset allocation. But if the cashflow shows that he needs a return of 1% in real terms you don’t need to go much further than cash (depending on what you believe about inflation) – why take risk when you don’t need to? But, if the forecast shows he needs a real return of 5% he should really be fully in equities and have to accept the risk or revise his objectives. Simply giving him a ‘balanced’ asset allocation and portfolio in both cases will be poor advice because it is likely to fail him either by taking too much risk or not enough.
My thoughts only of course!
Thank you very much for highlighting the difference James. As I said I’m not an expert and am guilty of hearing what I think people are saying not what they are actually saying.
I believe the BSB replaced actual stock selections with indices and found very little difference in performance (infact the indices performed better) thus the conclusion was it doesnt matter what the actual stock in the asset class is as long you are in the correct asset class.
This in turn means as you rightly point out that there is no optimum asset allocation and I think the work that I did only goes to highlight the difference between the ‘experts’.
Hi Richard,
I agree with your comments on Selestia and standard allocations and you make a very important point re the level of risk the client needs to take as opposed to what they are prepared to tolerate. I think this is one of the key differences between financial planning based on cashflow modelling and the more traditional kind of advice. I’ve still got no idea why different models have such completely different weightings though. I suppose as long as you have explained it to the client and they are happy with it, there’s no problem.
As James points out the asset allocation research in fact outlines that it determines portfolio volatility. Far to many people miss quote this, in fact asset allocation will effect returns but also big players in this are market timing (initiall(, period of investment and crucially the charges and tax have a big impact on returns so a low TER crucial, hence the importnace of ETF’s in a portfolio. In addition to thias if you head down the managed fund route smaler botique funds tend to out perform mpore consistently than the large fund houses and the longer a manager is in charge the better consistency of returns, so be careful when realying to much on the asset allocation modelling, this should be used for risk profiling !