I’m with you on having faith in the market. We think it is essential for all stock market investors, no matter what strategy you choose.

In general, we have found that investors that have faith in the market tend to do well, and those that are fearful do not.

It can be difficult to have faith in the market in big bear markets, like 2008, but the market has always bounced back. People forget that it is made up of a bunch of big, solid companies, that generally can adjust their operations to increase their profits, which is why the stock market goes up long term.

Perhaps it is a result of experience, but we have found it easier to be confident in a human being, like a fund manager. Once we have done our research and are confident in a fund manager and his methods, experience integrity, track record, etc., we get a more confident feeling from confidence in the fund manager than an impersonal market.

Many investors seem to have almost all their information from impersonal sources, like the internet, without seeing the importance of knowing the people behind it. We have avoided some major catastrophes by not investing with fund managers with exceptional track records, but where we were not comfortable with the person.

When the market crashes, there is a major comfort level from knowing that there is a really good fund manager hard at work making sure that anything that needs to be done is being done.

We still have faith in the market generally, but find a greater confidence from also having faith in a human being.

By the way, Michael Lee-Chin’s fund was a financial services sector fund. It is aggressive, since it is all in one sector.

Ed

Having said that, I just don’t like to be at the mercy of a manager. And yes, I prefer to have faith in the market. Weird, I know, but the market is many factors, the mutual funds are one factor, the manager.

Of course they took risk level into account. The 2 figures showing the amount by which the high Active Share fund managers beat their index are not a range. They are 2 different methods. The second figure (1.15% outperformance) is based on Alpha, which is “manager value added” taking into account risk level.

The study found that adjusting for risk did not change the results.

There is a follow-up article to come shortly that explains more details from the study. It lists a bunch of the factors that they all adjusted for, such as risk, size of holding, size of fund, etc. My first thought was that perhaps they were buying small caps, but they allowed for that as well.

The study clearly shows that these are superior stock pickers. They beat the market without more risk.

Ed

In our opinion, most of the managers you mentioned were just the aggressive manager in a hot sector. They were temporarily industry darlings (as you put it), but that is completely different from being an All Star.

What is so scary about a manager leaving? It’s like having Sidney Crosby on your team. What if he leaves? Just find someone else.

There is a deeper point here, though. The most skilled fund managers almost never leave their company. The reason for this is that top fund managers are rarely employees of someone else’s company. It is far more lucrative to have your own investment firm and get contracts managing certain funds, or have a fund company create a fund for you.

All Star Fund Managers, in our opinion, nearly always own their own firm, so they never leave it. Even when the retire, in some cases they have a good succession plan and the firm continues to outperform the same way.

Having a manager leave has almost never happened to us and is not something we worry about.

Ed

@Ed (#24) – I’ll read your #15 in more detail in the morning. In the meanwhile, your #24 and Canadian Capitalist’s blog entry today on Legg Mason make me wonder: did the study correct for level of market risk (beta) and level of volatility overall? One of the ways institutional investors sometimes claim to generate alpha (which is effectively what this study is saying about the active retail market) is by actually taking on more risk. You would expect more risk to mean more return, so if the High Active Share funds are also on average more volatile than their benchmarks, one could presumably duplicate the effect by biasing ones’ asset allocation and yet staying a passive (index) investor.

Remember the Altamira Equity fund of years passed? The darling of wall street and it put Altamira on the map.

What about AIC? That fund was untouchable. Didn’t he sell his company?

PH&N. Now that RBC has them, will they still perform?

Sorry, but Mutual Funds are quite dependent on the Manager and if the Manager leaves, then what?

I’ll stick to my ETF’s, thank you.

Thanks for the support!

I’m just trying to provide some general education. Too many people believe all the index propaganda that nobody can predictably beat the market and too many people buy closet indexers!

Your question about whether this will hold long term is interesting. We think so.

The study did show that fund managers with high Active Share tended to continue to have it in the future. This makes sense. A good stock picker that has holdings very different from the index is unlikely to suddenly have very similar holdings in a future year, except for possibly the odd year.

Our experience with what we call “All-Star Fund Managers” is that the all-stars tend to remain all-stars. In fact, our experience agrees with the study that experience makes a big difference.

Wisdom (for lack of a better word) makes a big difference. Young fund managers usually get into heavy computer analysis of public information, graphs, and many ideas that sound logical, while the more experience fund managers learned years ago that most of these don’t work.

Having said that, all top fund managers also have significant periods of underperformance. I saw one article that listed the top fund managers since 1950 and the percent of the time they underperformed. Most underperformed about 30% of the time and still beat their index by wide margins over the long term.

One striking example is Rick Guerin who managed a fund from 1965-83. This was a very unfortunate time – the period of time many people wrongly claim the index had no growth at all.

He beat the index by 32.9%/year compounded during his career! And yet, he underperformed the index in 7 of 19 years (37% of the time).

Having a high Active Share means you are very different from the index, so it inevitably means there will be periods of underperformance. This is part of why All Star Fund Managers are not that easy to identify.

Our big question is whether this fact will result in periods of time when an updated Active Share article would not confirm outperformance. The updates from the existing study already show the degree to which the top fund managers beat the market goes up and down.

In general, though, we would be surprised if future studies would not continue to show that top Active Share fund managers generally would beat their index.

Ed

This study actually proves the top funds do outperform indexes, not just “benchmarks”. It compared every fund to 19 different indexes and used the one that it was most similar to. So, it did not accept the fund’s category or benchmark.

This study compares every fund to a specific index – whichever index of 19 the fund most closely resembles.

The benchmark for a mutual fund is always an index or a combination of indexes, but this study was far more in-depth than to just accept their benchmark.

This is just one example of how in-depth the study is.

Ed

We just added the article to our web site so that we could make sure we had a reliable link from this blog for anyone that wanted to read the original study. But we did find it very useful.

Active Share does not by itself identify a top fund manager, but it can help you rule out those that are not.

Think of it like being a great skater in hockey. An all-star hockey player will most likely be a great skater, so you can rule out most of those that are not great skaters, but just because someone is a great skater does not make him a all star.

This study was very comprehensive and clear shows that some fund managers CAN predict which companies will do well.

The belief that nobody can beat the market is rampant, but makes not sense to us and does not fit with what we observe in studying fund managers. I can also identify the top people in many other fields, especially areas that publish a lot of stats – like sports.

Granted the markets are much more complex than hockey, but I can tell you who most of the most talented hockey players in the world are. There are people with superior talent, intellect, and work ethic in every field.

Why do people think that would not also apply to investing???

For example, the market is very manic with most investors, including most professionals, being sheep – essentially following the herd. This creates large systematic mispricings, which is why the majority of the all-time best fund managers are value style. There are many styles of investing, but one style has most of the top investors.

The top investors, not surprisingly, are not sheep, and are not following the herd. This study just confirms what makes intuitive sense to us – the top investors have holdings very different from the index.

Ed

The problem with the net fees in that report is that they are for American mutual funds which have much lower fees than in Canada. You just can’t ignore that.

The “net of fees” results in Canada will likely be much lower.

This stat is publicly available on Morningstar, etc. and can help you identify funds that might be closet indexers, but the study showed that funds with a high “tracking error” did not outperform funds with low “tracking error” or their index the way funds with high Active Share do.

Index funds and the most blatant closet indexers did have a low tracking error, so you can identify them. However, most of the closet indexers did not track quite that closely.

The difference is that Active Share compares the fund holdings to the index, while tracking error and coefficient of determination measure correlation.

Where this makes a difference is that:

- Funds with similar holdings to the index but focused in different sectors tended Not to outperform, but still have a low correlation to the index. The study called these funds “factor bets”. They are essentially sector market timers, which seems to be a strategy that does not work.
- Funds with more holdings, but very different from the index tended to outperform, but still be more closely correlated to the index. The study called them “diversified stock pickers”.

Ed

For now, the Active Share results of specific funds is not public information, although we have it for most of our fund managers. S&P and MSCI charge fees for the detailed index data that you would need.

You can estimate it by looking through the holdings and comparing them to the index. Funds publish their detailed holdings only once or twice a year, but you can see the top 10-15 every month, which can at least give you a clue.

We have heard that Morningstar might start including Active Share in their data later this year, but that is not confirmed.

Most of the tons of data produced about mutual funds is not very helpful in identifying the top fund managers, so adding Active Share would be helpful.

High Active Share alone does not tell you that you have an “All-Star Fund Manager”, but it can be very useful in ruling out all the closet indexers.

We think that the best benefit of publishing Active Share stats would be that the public could become educated and stop buying closet indexers!

Ed

By the way, we just found out that Cremers and Petajisto are now working on 2 new papers – one with US data through to the end of 2009 and one with data on Canadian funds.

We are really looking forward to those. We are speculating on the Canadian results. We have higher MERs in Canada, mainly because they include the cost of advice here, so can fund managers make that up? We also believe that the percentage of closet indexers is probably much higher in Canada. It is hard to find a fund that does not have several of the big banks and the largest insurance, gold and oil companies in their top 10.

The TSX is also not diversified, which should make the results interesting. The hottest sectors lately have been the banks and resource companies that make up 80% of the TSX. In the late 90s, the TSX was mostly technology, with Nortel actually making up 48% of the index by itself at the peak. So, the index rode the hot sector both up and down.

The Canadian results and the updated US results, including 2008 and 2009 (the worst and one of the best years in 80 years), should both be interesting!

Ed

What stock is this?

This study has actually been published several times with small revisions in the results.

The original (I believe) version was published in 2006 and showed the fund managers with the top 20% Active Share beat their index by 1.39-1.49%, while the bottom 20% underperformed by 1.41-1.76%.

The latest version is the one we just put onto our web site that shows the top 20% outperforming by 1.13-1.15%, while the bottom 20% underperformed by 1.42-.1.83%.

All these figures are after all fees.

We tried calling them and talked to one of their research assistants for an explanation. Apparently, it is common with university studies to publish updates based on their latest knowledge. They say it is a slight change in methodology, not any change in the data.

The 2 different numbers are not a range. The first number (1.13%) is just the amount the average fund manager in that category beat the index by. The second number (1.15%) is a formula calculating Alpha based on 4 factors. Alpha is a statistic measure of “manager value-added”.

The study specifically looks at MER and does not show an advantage for low fees, except for index funds and closet indexers (of course). For the managers with high Active Share, the fund managers with the highest Active Share actually had slightly HIGHER MERs – not lower.

This makes sense. They are truly active fund managers and more than justify their higher fees.

The study has data for 23 years, but the fund analysis is only for 13 years from 1990-2003. The reason for this is that nearly all funds were truly active back in the 1980s. They wanted to compare high Active Share to low Active Share funds, but there were hardly any low Active Share funds in the 1980s.

In the last 20 years, many people have come to the mistaken conclusion that nobody can beat the index. This seems to have led to a deterioration of the quality of fund managers, with 30% now being closet indexers.

Ed

You raise some good points in #2 & #14. The distribution of the returns for the high Active Share fund managers was quite wide. The study answers the question with statistics, but not an actual distribution.

In short, the returns of the high Active Share fund managers was quite different from the index – a relatively wide distribution. On average, they top group’s returns averaged 6-8%/year different from the index, with most of these returns being higher.

It is a bit complex to explain, but of the 694 fund managers with an Active Share of 80-100%, the largest group (245 of them) had a standard deviation of their returns of 6-8%/year different than the index.

That explanation may not be too helpful, but that is the interpretation from the study. This means that 2/3 of the time, the high Active Share fund returns were 6-8%/year different from the index and 1/3 of the time they were more different than that.

To try to explain the concept, this doesn’t quite mean they beat the index by 6-8%/year. These were the funds that mostly beat the index, but their returns were typically 6-8% different from the index each year (some lower and some higher, but mostly higher).

It is also interesting that the period of the study from 1990-2003 includes both the tech boom and the tech bust, as well as the 1992 recession, the “Asian Contagion” and a few other unique market events.

I doubt that the tech boom and bust affected these results, though. They used only broad-based funds (not the sector funds) and they filtered out many possible explanation factors, which would imply the they would likely get similar results with other periods of time.

Ed

One of the main points of the study is that you don’t need a crystal ball. Funds with high Active Share tended to keep it and tended to outperform. They specifically stated that choosing funds with a high Active Share is a good way to predict which funds will beat their index.

Also, of the funds in the highest category, the AVERAGE fund beat the index.

Ed