An experiment with riding a mutual fund manager
 

The reasons mutual funds are constrained and often put in disappointing performances have been discussed here a few times. One of the bigger reasons is that if they are any good they usually they have too much money to manage, and can't come up with enough good ideas to employ it all, and end up with a lot of so-so holdings which drags down the overall performance.

So I'm playing a little game with a fund manager I respect and believe is a consistent outperformer (who will remain nameless for now). I'm taking five of his top weightings in a successful fund he runs and putting a sum of money just into those five, and removing the 150-200 other holdings in the fund. In other words I'm taking his top weightings (near 2%) and pumping them up to 20% each. I'm choosing the five based on a few criteria: his weightings (higher is better), some semblance of diversification among the five, and a few morsels of my own research.

I started this about a month ago and have had a lucky streak and am up 10% so far. Yes, I know I have taken on significantly more risk than owning the whole fund.

Just a simple little lazy person's idea I thought I'd throw out there. Ask me how it worked in a year or so.

eastbay

Re: An experiment with riding a mutual fund manager
 

Want to post the 5 stocks?

Re: An experiment with riding a mutual fund manager
 

Are you just holding the five you picked or are you going to be trading them if he also trades them or if he makes moves that reweight his fund?

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
I'm taking five of his top weightings in a successful fund he runs and putting a sum of money just into those five, and removing the 150-200 other holdings in the fund

[/ QUOTE ]

I still don't really see why professional fund managers carry so many stocks (OK - some have liquidity issues). I had a small cap funs that had 600 stocks. I sold it and bought a small-cap index that held about 400 stocks - many in the same weighting as the managed fund.

These guy managing scores of stocks are just charging for high-priced closet indexes.

Re: An experiment with riding a mutual fund manager
 

do you follow his re-weightings?

if one of those stocks moves to a below top 5 weight, do you sell it and buy the one that mvoed into the top 5?

Barron

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
[ QUOTE ]
I'm taking five of his top weightings in a successful fund he runs and putting a sum of money just into those five, and removing the 150-200 other holdings in the fund

[/ QUOTE ]

I still don't really see why professional fund managers carry so many stocks (OK - some have liquidity issues). I had a small cap funs that had 600 stocks. I sold it and bought a small-cap index that held about 400 stocks - many in the same weighting as the managed fund.

These guy managing scores of stocks are just charging for high-priced closet indexes.

[/ QUOTE ]
You learned so much in those few lines.

Re: An experiment with riding a mutual fund manager
 

Unless he's close personal friends with the fund manager, he probably won't find out until weeks (or months) after the fund manager has sold, that he's done so. Most funds only divulge their holdings every quarter.

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
do you follow his re-weightings?

if one of those stocks moves to a below top 5 weight, do you sell it and buy the one that mvoed into the top 5?

Barron

[/ QUOTE ]

It's not a simple top 5, but yes, I will move if he significantly moves his weightings.

It's true that I don't know the holdings or weightings real-time, but this fund in particular has very low turnover.

eastbay

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
I'm taking five of his top weightings in a successful fund he runs and putting a sum of money just into those five, and removing the 150-200 other holdings in the fund

[/ QUOTE ]
Have you done any research on whether funds with a strong performance have their five largest position outperform in the future? It doesn't seem like it would be too tough to see if the data matches the theory.

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
Yes, I know I have taken on significantly more risk than owning the whole fund.

[/ QUOTE ]

It's not just that you have fewer stocks, but you could end up diversifying in a different way depending on how you pick your five. For example, let's say that as he changes his weighting on financials, he always adjusts commodities to have the exact same weighting as financials. If you don't copy this, it can cause an additional divergence in performance. Maybe better, maybe worse, but different.

You probably already know that, just throwing it out there.

Re: An experiment with riding a mutual fund manager
 

Would you expect risk-adjusted performance to be better than the fund itself, or just performance overall? I would expect only the latter.

Re: An experiment with riding a mutual fund manager
 

Are you able to screen for stocks that might be weighted higher because of a recent run? For instance, he bought stock ABC at 1% weighting, and it went on a tear and is now 2% but possibly overpriced?

Re: An experiment with riding a mutual fund manager
 

Here is an interesting link regarding the number of stocks one should have in a portfolio:

The 15-Stock Diversification Myth

The conclusion:

[ QUOTE ]


If the O’Neal data are generalizable to stocks, and I believe that they are, then even 100 stocks are not nearly enough to eliminate this very important source of financial risk.

So, yes, you can eliminate nonsytematic portfolio risk, as defined by Modern Portfolio Theory, with a relatively few stocks. It’s just that nonsystematic risk is only a small part of the puzzle. Fifteen stocks is not enough. Thirty is not enough. Even 200 is not enough. The only way to truly minimize the risks of stock ownership is by owning the whole market.

[/ QUOTE ]

Re: An experiment with riding a mutual fund manager
 

canuck... this thread is about mutual funds.

why come in and post that everyone should only buy index funds? efficient market theory is clearly wrong... yet people keep trying to turn every other thread in this forum into a debate about it...

15 stocks is more than enough diworsification... why wouldn't i want to put as much money as possible into my BEST ideas, which have what I perceive to be the highest EV? Spreading it around into a bunch of mediocre ideas isn't appealing. Do you know anyone who got rich off of their 7th best idea? Or their 200th best idea?

Re: An experiment with riding a mutual fund manager
 

Canuck,


You minimize risks more by owning the whole market than you do by owning 15 stocks? That is amazing, I can't believe no one ever thought of that before!

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
Here is an interesting link regarding the number of stocks one should have in a portfolio:

The 15-Stock Diversification Myth

The conclusion:

[ QUOTE ]


If the O’Neal data are generalizable to stocks, and I believe that they are, then even 100 stocks are not nearly enough to eliminate this very important source of financial risk.

So, yes, you can eliminate nonsytematic portfolio risk, as defined by Modern Portfolio Theory, with a relatively few stocks. It’s just that nonsystematic risk is only a small part of the puzzle. Fifteen stocks is not enough. Thirty is not enough. Even 200 is not enough. The only way to truly minimize the risks of stock ownership is by owning the whole market.

[/ QUOTE ]

[/ QUOTE ]

note: i did not actually read the article, and my comments are going to be very nerdy/academic.

min {risk} is not our optimization. what we are really trying to do is something like max {E[return] - execution cost} subject to {risk <= c}, where {risk} is whatever arbitrary risk measure you want to use, usually std dev, but often VaR or something relating to portfolio drawdown.

The issue here is that the number of stocks owned impacts all 3 factors (return, cost, and risk) to varying degrees. The standard argument, I think, is that impact on tcost (for individual investors) is negligible at most levels, and that after 15 stocks or so, the reduction in E[r] outweighs the reduction in risk for most people's utility functions.

edit: ok, curiostiy bit me while i was writing this up and i decided to go read it. i have a couple of issues: first, "kurtoskewness" and "TWD" are completely foreign words (to me at least) being used to describe basic phenomena, which is my first clue that these guys are FOS. also, they randomly choose portfolios of 15 stocks for their analysis, while the entire argument behind using only 15 stocks in the first place was that you could make better choices than when you invested in more, and were able to boost your E[r] more than you increased risk.

Basically, imo, the guys who wrote this are at best slightly dumb and out-of-touch, but more likely publish-or-perish academics with no real ideas who are putting out something misleading for the sake of putting something out, which is much worse.

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
Would you expect risk-adjusted performance to be better than the fund itself, or just performance overall? I would expect only the latter.

[/ QUOTE ]

Define "risk adjusted performance."

eastbay

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
[ QUOTE ]
I'm taking five of his top weightings in a successful fund he runs and putting a sum of money just into those five, and removing the 150-200 other holdings in the fund

[/ QUOTE ]
Have you done any research on whether funds with a strong performance have their five largest position outperform in the future? It doesn't seem like it would be too tough to see if the data matches the theory.

[/ QUOTE ]

I am not investing in "funds" in general, so... no.

eastbay

Re: An experiment with riding a mutual fund manager
 

eastbay,

instead of performing this experiment wouldn't it be easier to look up the top five weightings this particular mutual fund has had for the last 10 years and measure their performance against the fund's overall performance?

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
eastbay,

instead of performing this experiment wouldn't it be easier to look up the top five weightings this particular mutual fund has had for the last 10 years and measure their performance against the fund's overall performance?

[/ QUOTE ]

well, now you realize he'd have to find the fund's weightings. obviously the top 5 now haven't always been the top 5...

but you are correct that a backtest of this strategy would prove as useful info.

Barron

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
[ QUOTE ]
Would you expect risk-adjusted performance to be better than the fund itself, or just performance overall? I would expect only the latter.

[/ QUOTE ]

Define "risk adjusted performance."

eastbay

[/ QUOTE ]

monthly excess reutns of fund / total fund standard deviation of monthly returns

vs.

monthly top 5 excess returns / top 5 standard deviation of monthly returns

the poster is asking whether you will get a larger increase in standard deviation (or whatever appropriate risk measure you choose) than in absolute returns.

Barron

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]

Canuck,


You minimize risks more by owning the whole market than you do by owning 15 stocks? That is amazing, I can't believe no one ever thought of that before!

[/ QUOTE ]

Please read the article next time before you respond. If you did read it then work on your reading comprehension.

His point was that by owning 15 stocks you can indeed eliminate nonsytematic portfolio risk, however you risk missing out on certain super stocks that increase several hundred times in value. These might contribute significantly to your portfolio's earnings.

Also, about the author:

[ QUOTE ]
William J. Bernstein is an American financial theorist, known for pioneering research in the field of Modern Portfolio Theory. Bernstein is also highly regarded for his self-help finance books for individual investors who wish to manage their own equity portfolios.

[/ QUOTE ]

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
eastbay,

instead of performing this experiment wouldn't it be easier to look up the top five weightings this particular mutual fund has had for the last 10 years and measure their performance against the fund's overall performance?

[/ QUOTE ]
I must have communicated poorly, this was sort of what I was asking in my earlier post. Basically, have you tested the hypothesis that you lay out in the OP? You could not only measure how the top 5 holdings have done against the rest of the holdings in the fund historically, but look into other successful funds as well and see how their top 5 holdings do against the rest of the fund as well as how those top 5 holdings do against the rest of the market.

Re: An experiment with riding a mutual fund manager
 

Has anyone tried to backtest this "system"? I am pretty intrigued, but as someone said most 5+ baggers tend to be smaller weighted stocks that wont be anywhere near the top 5 in a huge fund. Alot of time the stocks with the most volume are also the most stable/sideways moving stocks in the fund. I am a mutual fund noob though , so perhaps someone else can pick this apart better than me.

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
eastbay,

instead of performing this experiment wouldn't it be easier to look up the top five weightings this particular mutual fund has had for the last 10 years and measure their performance against the fund's overall performance?

[/ QUOTE ]

That would not be possible considering the fund has managed by this person for three years. Nor am I using the top five weightings.

So... no.

eastbay

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Would you expect risk-adjusted performance to be better than the fund itself, or just performance overall? I would expect only the latter.

[/ QUOTE ]

Define "risk adjusted performance."

eastbay

[/ QUOTE ]

monthly excess reutns of fund / total fund standard deviation of monthly returns

vs.

monthly top 5 excess returns / top 5 standard deviation of monthly returns


[/ QUOTE ]

This is not an interesting quantity to me, so the answer is: I don't care.

As a measure of risk and reward, it presumes that standard deviation is a good measure of risk, and furthermore it assumes that I should consider twice as much return to be "the same" if there is twice as high a standard deviation associated with it. Neither of which makes much sense to me and together renders the numbers pretty uninteresting indeed.

eastbay

Re: An experiment with riding a mutual fund manager
 

[ QUOTE ]
[ QUOTE ]
eastbay,

instead of performing this experiment wouldn't it be easier to look up the top five weightings this particular mutual fund has had for the last 10 years and measure their performance against the fund's overall performance?

[/ QUOTE ]
I must have communicated poorly, this was sort of what I was asking in my earlier post. Basically, have you tested the hypothesis that you lay out in the OP?

[/ QUOTE ]

The hypothesis is not backtestable, because it is too discretionary. I am choosing a particular individual, who cannot be generalized to a statistically significant sample by any systematic method.

A hypothesis which is tangentially related is whether or not outperforming fund managers tend to repeat their outperformance. And I have seen convincing evidence of this phenomenon, which I believe to be consistent with common sense. Some stock pickers are better than others.

eastbay