The greatest enemy of a good plan is the dream of a perfect plan. - Karl von Clausewitz, Prussian General.

I'll start this article off with a couple of interesting numbers:

17% - That is the portion of money managers that outperformed the US S&P500 index in the decade 1968 through to 1978.

95% - That is the portion of these same money managers that expected to outpace the S&P500 index over the decade after that.

Source: A poll conducted by Institutional Investor, 1978.

21% - The number of these same professional money managers that remained in business and actually did outperform the index over the next decade.

Source: Common Sense on Mutual Funds, John C. Bogle

Index funds

First of all, are they an index fund with really low fees? If so you can rest assured that the returns will be superior to the great bulk of managers and you won't have to worry about picking funds at all. Simplicity is a virtue, and the choice to index rather than worry about picking funds makes your job a whole lot easier.

Want higher returns? First of all, consider buying a value or small company index fund, like the ones Dimensional Fund Advisors offer. Their value index strategies have beaten the large cap stock indexes by several percentage points per annum since their inception almost two decades ago - a remarkable performance and one equalled by very few elite active managed funds over the same period, not bad for a "dumb" passive strategy! You can also consider conservatively gearing your portfolio, which can boost long term profits if you are aware of the risks. Just make absolutely certain that you don't put yourself in a position where margin calls are even remotely possible.

Index trackers have higher returns than the majority of funds, they have lower turnover (and hence less capital distributions, therefore less tax to pay) than the majority of funds, and usually much lower fees. Unless you have the highest confidence in a managed fund, an index tracker fund is usually a much better bet.

The quote at the head of this article from General Karl von Clausewitz is to highlight what I consider to be perhaps the futility of trying to select an actively managed fund. Given the great difficulty of trying to figure out which managers are going to be part of the 15-30% of active managers that beat their benchmarks over the long term one might ask, "Why are you even bothering?".

Of course everyone knows that indexing is a "good" plan, but we all aspire to beating this strategy, and what would be perfect would be to identify in advance those rare managers that will give outstanding long term performance - a "perfect" plan. Do you or I have any right to presume that we will do any better in this field where millions have failed before? What evidence do you have, aside from confidence (or overconfidence perhaps) in your own intelligence, to believe that you will be able to select such managers in advance?

While I think an index fund is a perfectly fine choice I still think that active investment has its place. I am personally a value investor and am keenly aware of the mountain of evidence that value strategies work - both because of the practical success of top value managers and because academics have found that cheap stocks appear to outperform expensive ones in historical analysis. The article I have written below is textbook value investment. I pretty much found that the best professionally managed funds are the ones that invest "properly", using a disciplined value approach without resorting to index hugging and keeping costs and turnover low.

Another interesting article on funds that have outperformed is published by Tweedy, Brown Company, one of the world's great value managers. You can read it by clicking on the link below:

http://www.tweedy.com/library_docs/papers/tenways.pdf

After reading that you will understand much of what I have written below (though this article is not just plagiarism of the Tweedy, Brown article, I came up with this "system" of mine by considering that the best managed funds would have all the characteristics I find virtuous in a value fund, it appears though that after reading Tweedy's article I just reinvented the wheel!)

MER

An ideal fund has a low management expenses ratio (MER). There is absolutely no correlation between fees and pre-fee returns, so if all else is the same, you might as well just go for the cheaper fund out of the ones you are considering.

Some of the worst fund managers in terms of pre-fee performance also charge the most. Managers also tend to raise their fees if historically their performance has been good. As I'll elaborate on shortly, past performance is a poor guide to future performance and so when a fund manager opportunistically raises their fee they are just cashing in on past performance due to better marketability, not due to any promise of better performance in the future.

Good investment advice is nothing like good psychological advice, you really should worry a lot about little things in the funds management industry; "If you pay the executives at Sarah Lee more, it doesn't make the cheesecake less good. But with mutual funds, it comes directly out of the batter." Don Phillips (Morningstar President) in U.S. News & World Report article on July 8, 1996 commenting on whether paying higher fees for mutual funds results in higher returns.

John C. Bogle in Common Sense on Mutual Funds also presents a compelling case. According to his research, the biggest single factor in determining future after-fee performance is the cost of the fund. The cheaper the fund manager, the better returns. There was an almost perfect negative correlation of fees to after fee performance.

If a fund manager is really serious about their task and still wants to make a lot of money, you can accept performance based fees. The ideal fee structure for an actively managed fund is a very low standard MER and a hefty performance bonus. Performance bonuses should be subject to a "high water mark" where they are based on outperformance of the index based on a previous high, not the more common scenario where the manager has terrible performance one year and great performance the next, which equates to ordinary performance not exceeding the benchmark, but still takes a performance bonus for year two.

In an ideal world we would have two types of fund. There would be a few global managers like Vanguard and Barclays and State Street Global Advisers and Dimensional Fund Advisors who specialise in producing ultra low cost indexed funds, and a bunch of very active boutique managers that adopt specialised and aggressive stock picking strategies and take their remuneration from performance bonuses. This is nothing at all like the way the funds management industry works these days.

Portfolio turnover

I prefer a fund to have a very low turnover of stocks. Often this goes hand in hand with having a low MER, but not always. Some funds buy and sell the entire portfolio half a dozen times a year, hardly what I would call long term investing. In addition to adding hugely to expenses, for dubious benefit, they also put a lot of capital distributions into the hands of investors and cut down on dividends, which are often franked and hence provide a great tax advantage over capital gains, especially short term ones.

In a review of 3,560 American stock funds, Morningstar discovered that funds with low turnover ratios generate superior long term returns to funds with a high turnover. Over a ten year period, funds with an annual turnover below 20% outperformed funds with an annual turnover above 100% by a margin of 14%! High turnover funds gained only 11.29%pa vs low turnover funds which gained 12.87%pa, a difference of 1.58%pa. Additionally, high turnover funds were calculated to require at least 8% higher gain than the market index just to pay for their tax inefficiency. (Source: The Essential Buffett, by Hagstrom, reported originally in the New York Times on November 21, 1997 "The Price of Turnover" by Carole Gould.)

Restricting yourself to funds with a turnover below 40% will drastically narrow your choice of funds, of the funds I have examined there must be one low-turnover fund to every six high turnover ones, and there are less than half a dozen truly low turnover funds in Australia.

Why is turnover such a big deal? For a start high turnover wrecks tax efficiency, but even before tax turnover makes it hard to perform well. John C. Bogle Jr (son of John C. Bogle, Chairman of Vanguard, I'm quoting from his book again) is portfolio manager of Numeric Investors, a quantitative manager in the United States. He examined 20,000 trades and figured out how much turnover costs. To purchase "value" stocks you lose approximately 0.6% per purchase or sale due to liquidity, spreads and other costs. Small growth stocks cost about 1.8% to buy, trades that equal an eighth of daily volume another 0.5% and trades that equal approximately two day's volume cost a whopping 2.3%. Remember, this is one side of the trade only, to buy and sell costs twice as much.

Data by the Plexus Group gave a more general estimate of turnover costs and say that on average institutions lose approximately 0.8% of the transaction value. A fund manager that turns over 50% of their portfolio has to purchase 50% of their funds under management, and sell 50% of their funds under management. The cost of 50% portfolio turnover is therefore approximately 1.6%pa, and some of the really aggressive trading funds that turn over 200% of their portfolio may lose 3.2%pa! This is not included in the management expenses ratio, this is a "hidden" cost. So one would hope that an active manager would be very very good at stock picking because they are going to have to identify stocks that will out perform the index by several percentage points just to overcome their trading costs, and a couple more percentage points to overcome their management expense ratio, and a couple more percentage points again to make up for their loss of tax efficiency. I've never said that fund managers are stupid people who are incapable of managing money, but the massive cost hurdle they must overcome to beat a passive strategy makes their task virtually impossible.

Trading too much is a plague that the funds management industry just doesn't seem to comprehend. I have actually heard fund managers talk about their high turnover as a really good thing, claiming that in today's volatile markets only a very "active" fund manager can add value. (Clue: if the fund managers (as a whole) cut back their trading, volatility would decrease. Second clue: over the last couple of years since the market got "volatile" there seem to be less funds than ever outperforming their benchmarks, especially really "active" ones, and one of Australia's top managers is Maple-Brown-Abbott, who have a very low portfolio turnover yet have shamed most trading funds over a very long period of time).

I am still trying to understand how a fund that markets itself as disciplined, unemotional and patient, with long term holding periods can justify a portfolio turnover of 170%pa. Don't take an adviser's word for it that a fund has "long term" holding periods, I've seen funds with annual turnover above 200% call themselves long term holders of stock, and nobody challenged them on this.

I would be slightly more flexible on the low turnover criterion with growth funds than value funds. If a value and a growth fund have performed equally well, and all else being equal, I'd go for the value fund for its likely superior defensive nature in bear markets and lower transaction costs with low portfolio turnover.

The greatest value investors in the world (Warren Buffett, John Marks Templeton, John Neff, Tweedy, Brown Company and others) are well known for their low portfolio turnover. There are few good value investors that also do much trading.

A low portfolio turnover may be mandatory for value funds, but if this is a particularly good looking growth fund I'd loosen up my turnover criteria.

Most of the really famous growth investors (like T. Rowe Price and to a much lesser extent Peter Lynch - though Lynch is more a GARP guy than anything) have a reputation for somewhat higher portfolio turnover, growth investment relies more on riding trends than value investment, which is a more systematic pursuit.

A fund with high turnover will very likely have higher costs, less tax efficiency because of constant realisation of capital gains, and often less franked dividends.

Tax management and block trading

The better fund managers strive to achieve not just high performance as such, but high after tax performance. Tax is a major consideration, and while it doesn't show up in the past performance figures in the back of Money Management, it certainly does show up in the investor's account balances.

First of all the more tax efficient fund managers strive to keep portfolio turnover down, all things being equal the less a manager sells at a profit the fewer capital distributions will be made. Several studies have shown that turnover by itself is not always a major influence on after-tax returns, in fact when a portfolio manager sells stocks that are trading at a loss the realised capital loss can actually improve tax efficiency.

Some fund managers deliberately realise losses in order to improve tax efficiency. You can't judge a fund's tax efficiency just by looking at turnover, you need some idea of how much of this activity is profit taking vs how much is loss realising.

Another strategy is for the fund manager to concentrate its buying activities only after the stock goes ex dividend, thus reducing the amount of income the fund produces and increasing capital growth.

You can assess the tax efficiency of a fund by looking at the percentage that distributions are franked, as well as looking at the ratio of capital appreciation to distributions.

Another strategy is block trading. Some fund managers try to bypass the market in order to improve their performance. If one of the larger managers wants to exit a position they can either dump the lot on the market immediately and cause the price of that stock to plummet, or they can spend the next three months drip-feeding the market as they unwind their position.

A third option is for fund managers to call one another and find out if anyone is interested in buying whatever that manager wants to get rid of. Some fund managers set themselves up as full time block traders, usually offering to buy stocks others want to get rid of.

Just say fund A has a huge block of Telstra shares to get rid of. This block is so big that even with a liquid stock like Telstra it could take months to get rid of. They approach fund B, who are well known as being interested in taking the other side of block deals. Fund B agrees to the deal, but only with a few percent discount to the market price. Depending on how keen A is to get rid of the Telstra shares, and how disinterested B is, A may agree to a substantial discount.

One fund manager that really stands out as being the ultimate tax manager and block trader is Dimensional Fund Advisors, passive quantitative specialists running a variety of index funds. Block trading has made such a difference to Dimensional's results that their index funds have historically beaten their benchmarks after fees, particularly in smaller cap markets where they can buy with the biggest discounts.

Tax planning and block trading would seem to be obvious ways for fund managers to improve their performance, but surprisingly only a relatively small proportion of funds exploit these strategies to a really significant degree.

Diversification and closet index tracking

If the fund is broadly diversified and mostly resembles an index tracker, you might as well stop mucking around and actually buy an index tracker. If managed funds are to add value they really need to do more than outperform the index by half a percent, thus I prefer very concentrated portfolios where the majority of funds are held in a limited number of stocks. If they do well, they will outperform the index by a large margin.

The attitude of most active funds is disappointing. They expect investors to pay a higher fee for having their portfolio managed, and yet most managers are, not to put too fine a point on it, gutless.

It takes a lot to produce consistently higher returns than an index, so to make up for their costs managers need to make some hard decisions. Few actually do though, preferring "index hugging", which means they try to get away with making only small changes to their portfolio away from index weightings (while still charging for "active" management anyway). The problem is that fund managers place a greater emphasis on not underperforming the benchmark than they do on actually beating it. Managers would rather remain a part of the pack (a pack that get beaten by the index funds after fees) instead of really taking any relative risks. If you are choosing an active fund, try to avoid ones that indecisively want to sit on the fence and hedge their bets.

The thinking of an indecisive manager goes along the following lines, "I think XYZ Corp is way overvalued and far too volatile a stock to hold in my portfolio. There is no way any sensible investor would buy it at this price. On the other hand, XYZ Corp might go up after I sell it, and since I don't want to risk falling behind and getting bad press for our poor relative performance, I'm going to buy it anyway. I'll reduce my exposure to this stock by a quarter, so when the stock takes the dive that our analysts strongly expect it will, we'll marginally improve our performance relative to the index." And for employing these people you pay as much as ten times as much in fees for many active funds compared to the least expensive index funds.

If a concentrated or otherwise highly modified portfolio does badly, it will probably do very badly, at least in the short term. Every single investor in history that has achieved the goal of long term outperformance of the index has gone through periods when their portfolio lagged the general market for a while. Each and every one of them got bad press as financial journalists gloated over the apparent demise of a once great investment genius. Historically the record of the world's great funds is that the best portfolio managers lag their benchmarks about 40% of the time, occasionally by a lot.

To be a good manager you need to have the discipline to stay with a workable strategy even if the rabble start to catch up or even overtake you. Unfortunately, most active funds would rather be a part of the rabble than be an individual, because they are not in the investment business to produce good returns for their investors, they are in the business to build and retain funds under management and hence make high fees. Because of these priorities, which are of course perfectly understandable in a capitalist country, most fund managers are mediocre.

The usual business strategy is to focus on boasting loudly about some past period of (probably accidental) outperformance and the rest of the time trying to make sure they never underperform by a wide enough margin to make skittish investors take their money elsewhere. It isn't about maximising absolute performance, it is about minimising different performance.

The crucial problem is that many funds try hard to diversify widely and mimic the indexes and thus get index performance, but still have the cheek to charge active fees. I want the funds I use to make up their minds, they will either be a very active fund where individual merit, and not index weightings, make the criteria for purchase, or they will be an index fund with very very low fees.

On the other hand, if the fund buys a lot of very small issues that aren't in major indexes, broad diversification is perfectly fine. Peter Lynch had an enormous number of stocks in his portfolio. If the fund is buying 250 stocks from the S&P500 though, the chances of them outperforming the index after fees and taxes would be quite minimal.

Many fund managers take a conservative stance that limits their ability to add value. They worry about "tracking error" and "diverse performance", so they end up holding a portfolio that includes most of the major index stocks in roughly the same weighting. I take the view that when you decide to pay the higher fees of an active fund you should expect the manager to have an active approach, not a pseudo active approach.

If you don't feel you are willing to take an active management risk then use index funds only. If you want to take a small active risk then buy an index fund and an active fund, but put more money into the index. This will be a more cost effective and diversified approach than buying a semi-active manager.

Past performance

"There is some evidence that last year's winners tend to repeat next year. But it is very slight. Mostly the effect comes from the fact that really bad funds stay bad. Their expenses are high, and their choices stay haphazard." - Paul A. Samuelson, "The Long-Term Case for Equities," The Journal of Portfolio Management, Fall 1994.

Long term past performance is something to look at, but managed funds have a fairly high turnover of staff, everyone recruiting from their rivals in an attempt to poach some new talent. As funds tend to get new portfolio managers on a fairly regular basis, it is doubtful that past returns have all that much bearing on what is to come, unless the manager really does undertake to stick to a consistent strategy.

The best performing fund from last year will probably be either a hedge fund, a special situations fund or a fund that was simply involved in last year's "hot" sector. At all costs avoid last year's top fund if last year was significantly higher than its average return, you'll probably buy right at the peak of their success if you do, some funds get fund of the year from luck as much as skill. They also get an influx of new money from all the people who read about last year's result in some Morningstar list, with the sudden rush of money to invest the manager will find it that much more difficult to repeat their previous success.

A study by Boston outfit Financial Research Corp has found that putting your money into top performing funds is often a recipe for disaster. The top 10% of funds in every year since 1988 on average fell to the 48th percentile the following year and the bottom 10% often rose to the top half. This is a classic example of regression to the mean.

Unfortunately, investors seem oblivious to this phenomenon. Top performing funds are usually deluged with cash, which partly accounts for why they have such difficulty following up their high performances. According to Financial Research figures, the funds ranked highest by Morningstar, which make up just 18 per cent of the 7,075 stock and bond funds sold in the US, took in almost three quarters of the $340 billion of new investments last year. (For the record, this study was released in 1999, so "last year" = 1998).

The trouble is that every fund has a style, and markets rarely reward the same style year in and year out. One year large cap growth managers will dominate, the next small cap value, then financial stocks, then resources. You will almost never be well rewarded by for ever chasing performance in this way. Even in longer time frames performance chasing is a dud, even by relying on five year figures you have no guarantee that you aren't simply buying at the tail end of a medium term cycle. Chasing high past performance is a sure recipe for instilling a "buy high, sell low" discipline.

Consistency is important. Index trackers do well because they represent disciplined, long term investment in a diversified portfolio of large capitalisation stocks with low costs. (Read the "What Works on Wall Street" summary in the stocks section) Fund managers that keep changing their style are hard to analyse and it shows that the fund itself has little faith in its own techniques. Humans are often the weakest link in the chain and the more the humans have a chance to inject their opinions into the research, the more likely the fund will be biased toward chasing the stock du jour.

Funds that attempt to time the market are bad news, it is well documented that fund managers are never more bullish than at the start of a bear market and never more bearish than when the market is going to rally. If you have made a decision to invest your money in the stock market for the long term then you are entitled to expect that this is what the manager will do for you. Funds that keep 40% in cash waiting for a crash to happen or regularly sell off stocks when someone makes a dire prediction should be avoided.

On the other hand, if you are choosing a fund there are some things you can do to improve your market timing. It has been established before that contrarianism can be a useful way to achieve a degree of market timing success.

Robert G. Kirby, a lead portfolio manager for Capital Group and a member of the faculty at Stanford Business School had this to say about past fund performance when he presented a paper to the Institute of Chartered Financial Analysts and the Financial Analysts Research Foundation in Chicago, Illinois on April 2, 1976:

I won't state that short term, quantitative performance data are totally useless. On the contrary, such data are often quite useful as an inverse indicator. If I were looking for an outside money manager to manage my company's pension or profit sharing fund, I would go through a procedure something like the following: I would first look for an organisation that had been around for a while and that had produced good, long-term performance records with a variety of portfolios. I would make sure that the individuals in the organisation were experienced and talented. I would determine that the good, long term record was the result of a consistent application of a clear investment philosophy. I would satisfy myself that the organisation provided an environment in which it was rewarding to work so that good people would stay. Then, after I had identified the organisations that met all these specifications, I would hire the one who, for the past two years, had had the worst record..... Short term investment performance data are a far better inverse indicator of what to expect in the immediate future than anything else.

Unrealised capital gains

Avoid funds with large unrealised capital gains. I know this entirely contradicts the recommendation that you go with a fund with low turnover, but every time a fund sells a stock that it bought before you invested in the fund you pay the capital gains tax of the person who sold out before you got in.

It is hard to know where to stand on this one, since a fund with low turnover will most likely have the highest capital gains overhang, where an aggressive trading fund will not.

One solution would be to invest in new funds, though of course this itself is risky since you probably know nothing at all about the fund strategy, it has no proven track record to build on, the fund manager may be unknown and the ratings services still haven't had a look at it.

No one said it was going to be simple! Another method of course is to buy a fund that has just taken a beating in the short term but otherwise in which you have great confidence. This way you will be buying into the last guy's unrealised capital losses, which means you pay less tax than you have to when the fund bounces.

True to label

Funds should be "true to label". There are many fund managers that do particularly well in one specific sector, but badly in others. For example, a manager may have an excellent imputation fund for "high growth". When you buy their more diversified funds you get something that is 70% Imputation Fund and 30% cash, which is silly. Worse than this they charge a higher MER for their more conservative funds, even though they did nothing more than a dilute their imputation fund. If you are looking at a multi-sector fund, make sure it really does diversify across multiple sectors. You don't need the exact same allocations I show in the asset allocation article, but if you are paying a high MER for a diversified balanced fund then see that this fund does what it is supposed to.

True to label also applies to the investment style. A value fund should buy stocks that are trading at below value, if you see them start to chase short term gains by buying growth stocks then they are not doing as they are supposed to do. Many funds covertly change style when they want to try to beat out the competition. Don't give your money to a fund manager that lacks the discipline to stick to the strategy it nominated itself to follow, that is tantamount to false advertising.

People

Find out the names of the portfolio managers and see what else they have run. Don't buy a fund just because you think BT or Merryl Lynch or Perpetual or Rothschild or ABN AMRO are good managers, look at the past track history of the particular staff member who runs the show. This information is often in the bios provided in the prospectus, but if not you can find this information by calling the fund manager directly.

It is not uncommon for fund managers to poach staff from other institutions, so if their CV says they worked with another fund, you might want to check out the past performance of the funds they used to manage.

Peter Lynch wrote a chapter in his book One Up on Wall Street called "The Wall Street Oxymorons". In this he described clearly what the problem is that prevents most fund managers from adding any real value. He says the problem is that few of them are willing to stick their neck out. Every manager wishes to avoid "different performance" and work in a relative sense to benchmarks, as opposed to on an absolute sense. There is an unwritten rule, "you'll never lose your job losing your client's money in IBM."

He says 90% of funds adhere to an "approved stock list" that includes a fairly limited range of stocks. Before a fund will buy a stock, the stock must first be well known, respected and watched by many analysts and put on the recommended list. Lynch calls this waiting "Street Lag", and if you read Lynch's books you will notice that he makes a very big deal out of this, Lynch's ideal stock would have little or no institutional ownership. If the institutions own large chunks of it the stock may already be too expensive, over analysed and probably worth giving a miss.

Lynch finds it ironic that the rules of many funds that state that they will not deal in stocks below a certain capitalisation preclude them from buying small growing companies. It is not until a product is sold in every shop all over the country and has grown as big as it will ever get, and is soon to lose its "growth premium" in the stock market that funds start to take an interest and begin buying that stock.

So if you want to be in with a chance of having excellent performance well above the index you will need to avoid fund managers with an unwieldy and bureaucratic management style. If a fund is ever going to make it onto a recommended product list it must appeal to one of these committees, and for a stock to appeal to every member of a committee it must be an "obvious" pick, no committee is going to sign off on a small untried growth company or deeply undervalued contrarian stock, but of course these two classes of stock are where the biggest money is made.

You'll find that the big fund managers are all bound by committees that restrict their manager's ability to act independently. This is one reason why I prefer smaller fund managers and try to avoid any manager large enough to regularly advertise on television.

Wall Street Journal columnist Roger Lowenstein had this to say about fund managers and committees.

Picking stocks, like writing stories, is a one-at-a-time endeavour. It is done best by individuals or small groups of people sharing their ideas and buying only the very best. A small fund family managing selective portfolios ... can succeed as a group, but no large institution ... can order dozens of managers to outperform. The image can be branded, but not the talent. The people matter more than the name.
Another consideration that I very much like to see is the fund manager investing significant amounts of his own money in his own funds. I don't know of terribly many large institutional managers who do this, but I do know quite a few boutique managers that put most of their net worth into their own products.

When you see a fund manager that operates with a performance bonus system (they earn a percentage of the outperformance of their benchmark, not just a flat fee) then you know the manager will strive to perform really well. When you find that a significant number of staff members from that fund manager have all their super and most of their personal portfolio invested in their own funds, you know that they will strive for high performance but not take too many risks.

Fund size

It becomes increasingly difficult to manage money well when a fund gets too large, positions become unwieldy and they are forced to over- diversify into less desirable stocks. This is a particular problem with domestic or niche funds because global funds obviously have a much larger universe to choose from. For a global fund I would probably in fact favour a large fund as there are unique problems of global investing that can only be solved by having a well funded research operation.

Have you ever seen a paper that made the claim, "A sure way to beat indexing: buy the biggest funds"? The most any large fund manager can do is assure investors that the large size of the portfolio does not come at a large performance penalty.

On the contrary, there is plenty of documented evidence to show that large funds have even more trouble beating the market than most. One of many papers on this subject went so far as to specifically advise investors to seek out smaller managers, having found strong evidence that funds, in this case specifically small cap funds (though there are plenty of papers saying the same of large cap managers), tend to perform worse the larger they get. (Christopherson, J., Ding, Z., Greenwood, P., "The Perils of Success", Journal of Portfolio Management, Winter 2002, 41-53.)

Even Warren Buffett, arguably the best investor in history has indicated that he has had more and more trouble managing his portfolio as it grew to be worth billions of dollars. He has claimed that if he were ever to start again from scratch he could make as much as 100%pa with merger arbitrage and buying small capitalisation stocks, but now that he has such a huge portfolio he claims he can do no more than aim to beat the index by a couple of percent a year. Now very few fund managers would come anywhere near Warren Buffett's enormous abilities, so why should we assume a much less able investor could do any better than what Buffett aims to do?

Fund managers love large portfolios because they can take higher management fees. Would you rather take 1% of $1,000,000 or 1% of $10,000,000,000? Fund managers are keenly aware of the difficulties of producing above average returns with such a large amount of money to manage, and yet still they advertise themselves on the basis that biggest is best. Every second advertisement I see talks about how big the company is and how much money they manage in how many countries. This should not be an impressive statistic for a more knowledgeable investor.

Research houses and fund ratings

The fund ratings houses are alright, but I don't get too excited about them. van Eyk research is the most respected ratings service in Australia, though ASSIRT, Morningstar, InvestorWeb and others do provide a decent service. I doubt many retail investors would benefit from having a subscription to any ratings service, though financial planners certainly make use of them, even if the ratings house does show a great fund I doubt you'll recoup the money in your subscription fees. Although I may be accused of having something riding on this advice, if you are unwilling to do any serious research I really do think that you should either try to find a really good financial planner or buy an index fund.

Those tables in the back of Personal Investor magazine from ASSIRT don't give you the kind of information you will be able to use to choose a particularly good fund, there is much more to it than that. I don't believe that there is any useful information provided in the standard data sheets or tables on funds. None, nada, zip.

A table of past results alone, even if it comes with extra dressing like Sharpe Ratios or snail trails is not something that you can use to pick a good fund. The kind of information that would help you pick a good fund from a bad one is never available in the popular literature. It must be extracted, often with some difficulty, from the fund managers themselves.

One word of caution though about the fund ratings services. Don't pay too much attention to their scoring systems. Many, including most famously Morningstar, use a star score system based to a very large extent on recent "risk adjusted" performance. Studies are starting to pile up that show historical volatility is a virtually worthless measure, it is not a particularly effective predictor of future performance and in fact it isn't even a particularly effective predictor of future volatility.

If you combine this useless measure with another useless measure (short term recent performance) you don't end up with a great scoring system. I personally give absolutely no weight to a Morningstar score, or in fact the scores given by any of the ratings houses except maybe van Eyk (to a limited extent).

According to Mark Hulbert, the Hulbert Financial Digest has been tracking Morningstar’s performance since the beginning of 1991. Over the subsequent nearly seven years, Morningstar’s top-ranked no-load equity funds have lagged the stock market by an average of nearly three percentage points per year. See Hulbert's "No Stars for Morningstar" in Forbes (12/29/97).

I mentioned earlier in this article that I favour a contrarian approach. If you are going to go by past performance (and I suggest this have only a very low prominence in your thinking), go with recent under-performers with good long term performance. This will in all likelihood mean you are buying funds with a really low Morningstar rating. The systems used to determine "risk adjusted performance" usually look at recent history, giving lots of stars to funds that have outperformed in the short term with low volatility and few stars to funds with poor recent performance. Thus, in all probability many of the funds most likely to return to outperformance will have poor ratings from the research houses. Many of the five star funds are trading right at their peak, and may well be dragged back when the pendulum swings the other way.

The ratings houses are usually staffed by very similar types of people to those that run the funds themselves, and have similar views on investment and risk. By taking the advice of people who think very much along the same lines as the fund managers themselves you'll hear a very conventional view. Remember: these people get their backsides whipped on a regular basis by passive index trackers that could be run by chimps pulling levers. A decent stock investor shouldn't have problems beating the index but these guys do. Bear that in mind before you praise and glorify them too much.

If a fund has an excellent past performance stretching back a number of years, that might be enough to get you interested in it to find out a bit more about how this was achieved. You can probably eliminate most of the funds with a protracted history of underperformance, but that is all you can do.

Lack of discipline, short term quarterly performance and the herd instinct

Why do funds trade the way they do? The blame lies squarely on consumers and I don't think financial advisers are helping much. Less sophisticated investors tend to be extremely myopic about their funds, ready to change them at the drop of a hat. Many advisers take advantage of this, generating repeated commissions by moving people around a lot. I'd say most of the advisers that do this are probably ignorant schmucks themselves who know not what they are doing, but the clients certainly don't challenge them on this.

If a fund underperforms their benchmark for a few months, the press starts running articles on the fund having lost its way and falling apart. Investors will stampede for the exits and roll over into any other fund with a more conducive recent record. Fund managers know that they can't afford to lose these investors and hence will do anything in their power to not underperform for short periods of time. This is why fund managers trade so desperately, why they buy stocks based on momentum rather than fundamentals, why they always sack their most talented staff members right at the bottom of the downturn, and why they turn their backs on methods that have worked brilliantly for years as soon as some other method moves ahead.

I'll finish up this article with a statement, and then ask you to make a logical judgement based on this.

It is not easy to outperform the index by a couple of percent per annum.

If the long term performance of the stock market is about 10%pa, this means that an outperformance of the benchmark by 2%pa means you will need to do 20% better than the index return. To get a return that is 20% higher you'll need to do more than just tilt your portfolio around by 10% on either direction of the benchmark weighting, because you will make at least some mistakes your correct calls will have to make a very large difference.

If you are giving away 2% due to turnover costs, 1.5% due to management and marketing expenses and costing an extra 10% in loss of tax efficiency, then how on Earth are you going to achieve your goal over the long term?

An active fund is by its very nature going to have to pick stocks that outperform the market not by 2%, which seems a modest goal, but to attain a return that is 20% + 20% + 15% + 10% = 65%pa higher than the index, ie to aim to achieve a pre-fee, liquidity adjusted, untaxed 16.5%pa in order to attain your stated goal. Very few people can boast of having beaten the market by 6.5%pa over a matter of decades, and if you are not Warren Buffett, Peter Lynch, Sir John Marks Templeton or a surprisingly small list of others you have no right to claim that you can.

Powerpoint presentation

Click here to have a look at a Powerpoint presentation "How to Pick Managed Investments" that I prepared for a seminar I gave in October 2002.