Value of selecting mutual funds based on past performance and more...

[Valuebo]

Hi all,

 

I was looking into some mutual funds (not for me, at least not now) and was wondering about valid ways to identify one that will beat the market going forward. It should be that impossible to pick winners when you know what you are looking for.. right?

 

Specifically, I'd like to know exactly what the odds are that anyone outperforms against their benchmark given both the length of their track record and specific degree of outperformance, all after expenses and performance fees. I would then combine this data with other criteria like:

 

- value approach

- level of diversification and what the fund invests in (now and past)

- track-record & background before starting the fund

- turnover rate

- R-squared

- AUM if relevant

- expense ratio (kinda useless if you covered outperformance including all costs?)

 

I assume the combination of both past outperformance and the criteria above would give me a solid list of value investors with real skills and future potential, cancelling out the ones that were mainly lucky.

 

Obviously the odds of plain luck are close to zero for a fund manager with 5% outperformance after expenses and a track record of 10+ years. (Unless maybe he rode a big bubble, didn't have any diversification and got really lucky or just had the wrong benchmark? Or a combination...) But what about managers with a track-record of only 5-10 years? Can you really say something useful about their skills in that time frame? Can you put an accurate mathematical chance on it without extreme effort while calculating it? Just wondering...

 

Anyone who can provide insight/numbers or has interesting research on this? Am I reaching to far and should I simply focus more on the qualitative criteria? A broader discussion ('Can mutual funds have a place in the portfolio of someone who consistently beats the market? What criteria are of main importance when selecting funds?...') would also be appreciated.

 

TIA

 

Tom

[vinod1]

I would look first at the qualitative criteria

 

1. Are they following the value approach i.e. do they really believe in it and does their portfolio reflect this? Look at the portfolio they are holding in 1999, 2005-2007, and mid 2009. Read the shareholder letter to find out their reasoning for the portfolio. For me a badge of honor is negative returns in 1999 and low returns in 2006 and 2007. We know from hindsight and it should have been clear to any good value investor in 1999, that the most attractive and safe stocks in 1999 had a near zero to a small loss.

 

2. Look at turnover and expense ratio. If they are really following value investing they should have turnover in the range of roughly around 15-25% implying a holding period of 4-7 years. But this is a very rough guide and it should only be used over a number of years. Also they should have low expense ratio indicating that they are not greedy.

 

Vinod

[scorpioncapital]

For managers less than 10 years, I would track

 

-incremental improvement. Say 2% above average in first 3 years, then 3% in next 3 years, then 4% in next 3 years. Actually it doesn't matter as long as the number is going up.

 

-I would track number of decisions made because one can be lucky and just "play once" and outperform big. So set a minimum number of investment decisions made that you would like to see that you think would cut out the noise.

 

- I would track a third variable, not related to investment selection but impacting on it significantly which is (mental) concentration ability and this can best be seen if you can somehow meet the person or read their literature to see how they interact with partners and how they think about the investment process.

[SharperDingaan]

Simplest to just buy the fund class with the worst 1yr record (Greenblatt), & hold it for 1-2 Yrs.

 

You're really betting on mean-reversion. Today's star being replaced with another, the popularity order getting re-arranged, & todays unpopular choice rising a rung or two on the backs of tomorrows dogs.

 

You're also buying the class (ie: todays global equities) on the expectation that all boats in that class will eventually rise with the tide. Even the dog in the class will rise with the tide.

 

SD

 

 

 

[Shane]

David Swenson outlines some guidelines in his book "Unconventional Success"

 

I do not have the book on hand but you might want to check that out.

 

[frog03]

I like to have at least 8-10 years track record outperformance (>=5+ points vs relevant index dividends reinvested) and reasonnable AUM.

This basically does not leave anything left in the US (where the best managers typically go the hedge fund way due to better compensation and more flexibility).  It does however keep some very interesting options in some other parts of the world like Canada or Europe.

[Shane]

frog - mind giving out some examples?

[frog03]

Some examples:

 

Sprott Canadian Equity in Canada (about 12 points of added value vs. the index over 14+ years).

Flinvest Entrepreneurs in France (about 8 points of added value vs. the index over 8+ years and the manager has a similar outperformance for 5 years before founding Flinvest so about 8 points for 13+ years).

 

 

[Valuebo]

I would look first at the qualitative criteria

 

1. Are they following the value approach i.e. do they really believe in it and does their portfolio reflect this? Look at the portfolio they are holding in 1999, 2005-2007, and mid 2009. Read the shareholder letter to find out their reasoning for the portfolio. For me a badge of honor is negative returns in 1999 and low returns in 2006 and 2007. We know from hindsight and it should have been clear to any good value investor in 1999, that the most attractive and safe stocks in 1999 had a near zero to a small loss.

 

2. Look at turnover and expense ratio. If they are really following value investing they should have turnover in the range of roughly around 15-25% implying a holding period of 4-7 years. But this is a very rough guide and it should only be used over a number of years. Also they should have low expense ratio indicating that they are not greedy.

 

Vinod

 

For managers less than 10 years, I would track

 

-incremental improvement. Say 2% above average in first 3 years, then 3% in next 3 years, then 4% in next 3 years. Actually it doesn't matter as long as the number is going up.

 

-I would track number of decisions made because one can be lucky and just "play once" and outperform big. So set a minimum number of investment decisions made that you would like to see that you think would cut out the noise.

 

- I would track a third variable, not related to investment selection but impacting on it significantly which is (mental) concentration ability and this can best be seen if you can somehow meet the person or read their literature to see how they interact with partners and how they think about the investment process.

 

Those seem like very good tests, thanks. How can I obtain the letters from smaller funds tho? Do they even all have one?

 

David Swenson outlines some guidelines in his book "Unconventional Success"

 

I do not have the book on hand but you might want to check that out.

 

 

I think I will. Seems like the perfect book to get a better grip on the industry. Thanks.

 

If anyone knows more books like this or books with a good chapter on picking mutual fund winners, let me know. :)

 

I like to have at least 8-10 years track record outperformance (>=5+ points vs relevant index dividends reinvested) and reasonnable AUM.

This basically does not leave anything left in the US (where the best managers typically go the hedge fund way due to better compensation and more flexibility).  It does however keep some very interesting options in some other parts of the world like Canada or Europe.

 

What would you call reasonable AUM? Or would it totally depend from the manager's playing field (small, mid, large cap, specific sectors, ...)?

 

Some examples:

 

Sprott Canadian Equity in Canada (about 12 points of added value vs. the index over 14+ years).

Flinvest Entrepreneurs in France (about 8 points of added value vs. the index over 8+ years and the manager has a similar outperformance for 5 years before founding Flinvest so about 8 points for 13+ years).

 

I've learned about Flinvest before through you from another topic and think the manager is doing a great job. I love that they are European and am impressed with their performance in this flat market. Volatility also seems to be a lot lower. I know we shouldn't look at volatility as risk taken, but can constant lower volatility (dropping a lot less in down markets) be a good indicator as well?

 

Could you tell me more about his performance before Flinvest? Where can I find out more about his long-term track record?

 

I'll look into Sprott Canadian Equity, thanks!

 

 

Thanks again all and sorry for the wall of text.  ;)

[Packer16]

The key is to find folks that are willing to make concentrated non consensus contrarian bets.  This is hard to find in mutual funds as they are dealing with larger amounts of money (thus larger more well researched stocks), have diversification rules, the buy/sell decisions can be determined by flow of funds and some of the picks can become value concensuses.  If you want to get average returns, they are great tool but above average returns are much harder due to above mentioned constraints.

 

The key for me is to examine the undelying holdings and see if the thesis is sound and if the holding is truly disliked.  You can compare put/call volatility to get some sense of the consensus.  If you find stocks with higher put volatility than call volatilty that is a start.  Also having a value discipline is important.  The most you can hope for from these funds is beat by 5% and if you are going to spend the time and effort to do this much research going a step further to individual stocks is not that much more research and you can control the buy/sell timing.  Just some thoughts. 

 

Packer

[scorpioncapital]

The difference between investing and other games is that in other games you actually have to work just to get average - since everyone else is trying to do the same, you have to think just to tread water and not do anything stupid like jump in and out of stocks cause you're panicking. What makes investing an easier game is that to be average you just have to buy an index and do nothing. I know of no other game that is quite that easy - to get average that is.

 

As for talking fund managers I find many do have open houses and you can just go up to their office and talk to the manager, don't be shy.

 

 

[frog03]

I've learned about Flinvest before through you from another topic and think the manager is doing a great job. I love that they are European and am impressed with their performance in this flat market. Volatility also seems to be a lot lower. I know we shouldn't look at volatility as risk taken, but can constant lower volatility (dropping a lot less in down markets) be a good indicator as well?

 

Could you tell me more about his performance before Flinvest? Where can I find out more about his long-term track record?

 

*****************************

Flinvest E. returns are MUCH higher than the market (despite hedge fund like fees!) with MUCH lower standard deviation of yearly returns!

 

Before founding Flinvest, Flecchia managed Oddo Avenir and Oddo Europe Midcaps.  I don't have the Europe returns but I do have the Avenir returns.

 

1998 (Jul-Dec) -3,2%

1999 57,9%

2000 23,0%

2001 -13,0%

2002 -7,6%

2003 (Jan-Jun) 5,7%

 

You can also google Flecchia Oddo for some information including interview.