|
Research Basis of Our Portfolio Management Approach "Investment Philosophy"
The 'old style' of financial planning, where revenue is generated from commissions, is based on financial products. That's because the financial products pay the bills. So there is always going to be a focus on them. Many institutions own financial planners as 'distribution networks' for their products. AMP financial planners tend to distribute AMP product, Commonwealth Bank planners Colonial products (Colonial is owned by Commonwealth Bank), Suncorp Metway planners Suncorp Metway funds and Westpac financial planners BT products (BT is owned by Westpac).
A 'fee for service' financial planning structure, free of institutional ownership and all commissions provides a whole different challenge for a financial planner. The challenge is - out of all the investments in the world - managed investments, direct listed investments, internally geared investments, agricultural investments, private equity, high yielding fixed interest investments, unlisted investments, share based investments, Government bonds, term deposits and cash accounts - how do you to about building an investment portfolio that will provide the best returns for your clients? After all, when your only source of revenue is client fees, you have to provide quality solutions for your clients.
The purpose of this document is to articulate our philosophy on investment and portfolio management.
Everyone in the financial services industry will claim to have some sort of 'research driven approach' to investment analysis and portfolio construction. What will vary will be the quality of research.
Some approaches will be build on a share trading system that has been back-tested (researched) on 12 months of historical data - which they claim will give a strong guarantee of outstanding future returns. Others will tote a research approach built around internal research to identify the best fund managers to be investing your money.
When we talk about a 'research driven approach' to investment we are talking about an approach built on the highest quality research produced by academics over the past 30 years. This approach goes far beyond the often simplistic approach to managing money, which is to try and find managed funds that have had good historical performance in the hope that they can repeat this. Our approach looks at the careful management of risk, the use of various asset classes to reduce portfolio volatility and the importance of costs and taxes in investing. It considers the problems caused through overconfidence and how to avoid them. It is an approach that incorporates the work of Nobel Laureates such as Markowitz and Sharpe, along with many distinguished academic writers.
Markowitz is most famous for his article published in the 1952 Journal of Finance entitled 'Portfolio Selection'. The paper provided a rationale for investors to consider their investments as an overall portfolio, rather than as individual investments. Their focus should be on forming a portfolio that was properly diversified with attractive risk-reward characteristics. Shape built on this research in developing the Capital Asset Pricing Model (CAPM). This model was the building block upon which further investment pricing models were built, including the Fama and French model that we apply to our portfolio management process.
The academic articles cited here are parts of the evidence, not the whole evidence. These articles have all been published in 'peer reviewed' journals, often knows as scholarly journals. This means that articles have been assessed by a panel of experts prior to publication, providing scrutiny as to the quality of the work. Once published, these articles are further scrutinised and criticised (if appropriate) by academic peers who may challenge or disagree with what is being put forward.
Our Portfolio Management Approach
Step 1 - A Focus on Long Term Asset Allocation
Asset allocation refers to the decision related to which asset class to invest in and in which proportion to invest in them. Common growth asset classes include Australian shares, international shares, listed property investments and direct property investments. Defensive asset classes include fiixed interest and cash investments. There are also many alternative investments such as agricultural investments, hedge funds and mezzanine debt investments.
In the Financial Analysts Journal in 1991, Brinson, Singer and Beebower provided an update to their 1986 article 'Determinants of Portfolio Performance'. This study examined 91
The results were conclusive, with over 90% of the variation in returns explained by asset allocation. 4.6% of the variation in returns was explained by security selection and 1.8% by market timing.
For our purposes we use the study to justify a focus on asset allocation as being the key driver of portfolio performance. We also use it to justify why we are not going to try to use market timing, moving from asset class to asset class, to try to increase performance. The article shows that this is difficult, if not impossible to do. So our approach is to build an appropriate long term asset allocation, and then stick with it over time.
More recent studies by Ibbotson and Kaplan (2001) 'Does Asset Allocation Policy Explain 40, 90, 100 Percent of Performance?' published in the Financial Analysts Journal in 2001 and 'Another Look at the Determinants of Portfolio Performance' by Craig French and available at ssrn.com, found that asset allocation policy explains more than 90 per cent of the variation in total portfolio return. This supports the study of Brinson, Sinter and Beebower.
We apply this to our portfolio management approach by focusing on identifying and maintaining the appropriate asset allocation for all investors. The mix of share based investments, cash investments, property investments and fixed interest investments will be the biggest contributor to portfolio performance. At each review, asset allocation continues to be a key focus.
Step 2 - Using Low Cost, Passive Investments
It may well be the biggest debate currently in the world of investing. Should investors use an 'actively' managed fund or a 'passive' fund?
An actively managed fund is one that looks to select investments that will provide a better than average return. For example, an active Australian shares fund manager will put together a portfolio of Australian shares that he or she things will perform better than the average market. A passive fund manager simply seeks to own the average market. That is, they will buy all the assets in a market, in the proportion that they exist in the market. For example, an index fund based on the 200 biggest companies in the Australian share market will simply own those 200 companies, in the proportion that they are in the index. Often passive funds are known as 'index' funds because they seek to replicate the index related to their investments.
Actively managed funds tend to be more expensive, often significantly so, because they are researching investment opportunities. Actively managed funds tend to trade shares more often, as they attempt to add value through their trading. This trading makes actively managed fund less tax efficient, as each time a profitable trade is sold capital gains tax has to be paid by the investor.
So, the question is can actively managed funds overcome their higher costs and tax inefficiency to outperform passive index funds? The answer from academic researchers seems to be a clear no, with most studies not even considering the tax advantages of index investing in making their conclusion about the superiority of index investing.
In a study entitled 'Indexing versus Active Mutual Fund Management' published in the Journal of Financial Planning in 2002, Fortin and Michelson found that 'on average, index funds outperformed actively managed funds for most equity and all bond fund categories on both a before tax and after tax basis'.
Russ Wermers, in a paper entitled 'Mutual Fund Performance: An Empirical Decomposition into Stock-Picking Talent, Style, Transaction Costs, and Expenses' published in the Journal of Finance in 2000, found that while fund managers had some ability to select stocks that outperformed the market, the funds still underperformed the index, primarily because of expenses and transaction costs associated with an actively managed fund. This study only looked at before tax returns. The better performing funds considered in the study had a turnover rate higher than 100%, that is they bought and sold their entire portfolio every year, which would have meant higher taxes and reduced after tax returns for the investors in these funds.
In
As anecdotal evidence of this it might be worth looking at the Morningstar Managed Fund tables in the Australian Newspaper. We don't want to put too much stock in short term results like the 1 year return, however the 5 year returns are often interesting. For the table published with data to the 13 April 2006 (the more recent available at the time of writing) provides details for 77 managed funds listed in the Australian shares category. The average return for all fund managers was 13.31% a year. The return for the Vanguard Index Fund was 13.9% a year, above the average. Of the 77 funds only 19 performed better than the Vanguard Index Fund. So, the passive approach was well within the best 1/3 of managed funds with an above average return - all before taking into account its likely tax efficiency advantage over the other funds.
We apply this to our portfolio management by favouring low cost, tax efficient index funds as the 'core' of our portfolios. Much like index managers not having to spend money to research individual stocks, we can provide a lower cost service by not spending resources on researching fund managers.
Step 3 - Ignoring Historical Returns
You often see the disclaimer in financial services industry adds that 'past performance is not an indication of future returns'. Ironically this disclaimer is often in advertisements trumpeting a period of high performance for an investment produce in the hope that this will attract investors. It is often assumed that any outperformance by a fund manager over time is a sign of investment skill.
There are thousands of managed fund in
Mark Carhart, in his paper entitled 'On Persistence in Mutual Fund Performance' published in the Journal of Finance in 1997, found that 'The results (of his study) do not support the existence of skilled or informed fund portfolio managers'. He did find that there was some evidence of persistence in underperforming funds, that is that poorly performing funds tended to keep performing poorly. Overall he did not find that there was evidence that selecting a managed fund that had performed above the average would then lead to above average performance in the future.
In the Australian environment Drew, Stanford and Veeraraghavan, in an article entitled 'Selecting Australian Equity Superannuation Funds: A Retail Investor's Perspective' and published in the Journal of Financial Services Marketing in 2002, found that 'the results (of their study) suggest that pervious annual performance has little influence on future returns'. In fact, 'Selecting funds based on persistence strategy resulted in underperformance of industry and passive (index fund) returns'. The study also comments on the negative effect that high fees and trading costs have on the performance of actively managed funds.
We apply this to our portfolio management by not using historical return as a basis for selecting fund managers. By using passive funds we can focus on building your financial position without having to worry about researching fund managers, switching investments or worrying about changes in the teams of fund managers. Our investment approach does not require the identification of above average management skills to be successful.
Step 4 - A 'Tilt' to Small Companies and Value Companies
Gene Fama and Ken French are US based researchers who built a more sophisticated model of expected returns from stock (share) investments. In their paper, they found that expected returns from each investment should also be related to the size of the company and to a 'value' premium.
They found that small companies had a higher expected return than large companies. This is not difficult to understand as we can recognise that investing in small companies may be riskier, and therefore requires a higher return to attract investors.
The 'value' premium is more difficult to understand. Value companies are usually identified as those companies trading at lower prices compared to the value of their assets. This may be because they are companies under some financial pressure, and therefore investors have tended to ignore them, or treated them with some caution. The reality is that these companies tend to outperform over the long term - providing a higher return for investing in companies under financial pressure. In Australia the index made up of the 30% of companies with 'value' characteristics has returned 19.9% a year since 1980, whereas the sharemarket average return has been 13.1% a year over the same period.
Quite often you will see the Fama and French research described as 'controversial'. Certainly there were initial claims that their research was based on 'data mining', that they had found a sample set that supported their research, and that was all. However large scale 'out of sample' research has shown that both the 'small company' premium and the 'value' premium broadly exist.
We apply this to our research by using a series of low cost, passive funds that focus on investing in smaller company and 'value' companies. Building portfolios to include these funds allows us to 'capture' the small company and value premium as described by Fama and French, increasing our expected long term return from sharemarket investments.
Step 5 - More Research on the Hidden Costs of Active Management, and the Difficulty in Choosing Outperforming Actively Managed Funds
Ross Miller, in his paper 'Measuring the True Cost of Active Management by Mutual Funds', sets out to identify how much the returns from mutual funds, a US term for a managed funds, are a result of closet indexing and how much are a result of active management unrelated to the index. He then proportions a reasonable fee for the index fund management based on the Vanguard S&P 500 Index Fund (0.18%) to find out the true cost of the actively managed portion of the fund. That is, he assumes that the indexing investment management cost 0.18% for the portion of the fund managed this way, with the remaining management cost being attributed to the actively managed portion of the fund.
The results are very interesting. For the 152 'large company' mutual funds that formed the sample, on average only 15.55% of the total funds were actively managed. The average MER for the actively managed portion of the funds was 6.99%. On average more than 96% of the variance in the returns of the fund was explained by movements in the index. On average the 'value added' by the active management was negative 9%. This is an investment loss of 2% on top of the fees of 6.99% apportioned to the actively managed component of the fund, clearly demonstrating that in this sample active management destroyed value.
On an overall basis the 152 mutual funds underperformed the index by an average of 1.5%.
Generally we see active funds management as an expensive way to manage portfolios. Given the overwhelming evidence that active management does not add value to a clients situation, we will avoid active management for the cheaper, more tax efficient method of passive funds management.
Step 6 - Franking Credits, an 'Unpriced' Bonus!
Cannavan, Finn and Grey, in an article entitled 'The Value of Dividend Imputation Tax Credits in
Franking credits are usually paid as part of the dividends from Australian shares, and are able to be used fully by Australian investors to either reduce their tax or, if their franking credits are greater than their tax owing, to receive a tax refund for the value of the credits. A 'fully franked' dividend valued at $70 will include a further $30 in value from franking credits. The research notes that international investors often cannot use the franking credits and, as such, they have no value to them. If these investors are the price setting investors, then this explains why franking credits are not priced into the value of shares.
This 'additional bonus' can be significant. Even a share yielding around the market average of 4% provides an extra 1.7% of return through franking credits if the dividend is fully franked. An investment yielding around 6% (fully franked) provides an extra 2.6% return through franking credits.
We are prepared to use some high yielding direct securities in portfolios to benefit from this suggested non-pricing of franking credits. Direct securities are also inexpensive to own, with no ongoing costs, and the ongoing income can either be used to re-invest into portfolios, or to fund a person's lifestyle. Even though no research has been done on the benefit of tax free and tax deferred income from listed property trusts, we are also prepared to incorporate these into portfolios to increase the tax efficiency and income streams of portfolios.
Step 7 - Ignoring the Hype Around IPO's
Stockbrokers often pronounce as one of their benefits that they can provide clients access to Initial Public Offerings (IPO's).
However, a measured look at the academic research suggests little benefit in investing in IPO's. A paper by Dimovski and Brooks entitled ?Initial Public Offerings in Australia 1994 to 1999, Recent Evidence of Underpricing and Underperformance' published in the Review of Quantitative Finance and Accounting found that 'longer term analysis supports the finding of previous studies in that IPOs on average, underperform the market in the first year following their listing.'
A general summary of the literature, supported by this specific study, is that IPO's tend to provide a reasonable return on the first day of trade, and then underperform. We do not believe in investing in IPO's to try and profit from the first day's price movements, as we see this as a speculative attempt to try and forecast prices.
Another factor to bear in mind is what is often described as the 'winners curse'. That is the situation where all the stock in a good float will be snapped up by the institutional investors and private broker clients. So, if stock is being offered to 'retail clients' of brokers, then it is likely to be poorer quality, hence the term 'winners curse'.
We generally ignore investing in floats as being a largely speculative activity. The exception to this is the case of a large, profitable company being floated where a strong and reliable financial history is available.
|