New Essay - Evaluating Performance

[racemize]

I've just wrapped up my latest essay, which I think will be of particular interest to this forum.  It revolves around how performance is reported and proposes a different statistical method.  I then used the method on 19 famous value investors: Warren Buffett (1957-1969), Walter Schloss (1956-2002), Charlie Munger (1962-1975), Tweedy Browne Partnership (1958-1983), Sequoia Fund (1970-Current), Davis Venture Fund (1970-Current), Tom Russo (1984-Current), Francis Chou (1984-Current), Longleaf Partners (1988-Current), Bill Nygren (1992-Current), Chuck Akre (1993-Current), Yactkman Fund (1993-Current), Francis Rochon (1993-Current), Tweedy Browne Value Fund (1994-Current), Guy Spier (1997-Current), Mohnish Pabrai (2000-Current), Bruce Berkowitz (2000-Current), Arlington Value Capital (2000-Current), and Austin Value Capital (our record) (2010-Current).

 

It has definitely shaped my opinion about which investors I want to follow in the future. 

https://drive.google.com/file/d/0BxTPR9eP5nWeNTJWOF9lUm1uOFU/view?usp=sharing

 

[LC]

Thanks for sharing all your work! I was looking forward to this after you were asking for all these guys' records.

[JayGatsby]

Nice performance. Well done.

[Jurgis]

Nice essay. Perhaps you should persuade someone like Morningstar to adopt this methodology for performance evaluation.  ;)

 

Edit: I probably should look at my own performance based your methodology. It's just seems like a lot of work. Plus looking at rolling window returns in private account will probably raise up the issues of capital inflows/outflows even more than looking at a (single) very long term IRR number. Also in my particular case at least, I used quite different investment approaches 5-10-15 years ago than I do now. So the data is dirty, the older data is not so useful and therefore there's not much of 10-y rolling windows overall. So, not sure I'm gonna do it. It seems interesting though.

[racemize]

Nice essay. Perhaps you should persuade someone like Morningstar to adopt this methodology for performance evaluation.  ;)

 

Yeah, I was thinking I could send it to them.  There's a ton of really cool stuff that could be done, but it requires a lot of decent web programming.  Of course, then I'm just handing it to them, but oh well.

[Jurgis]

Nice essay. Perhaps you should persuade someone like Morningstar to adopt this methodology for performance evaluation.  ;)

 

Yeah, I was thinking I could send it to them.  There's a ton of really cool stuff that could be done, but it requires a lot of decent web programming.  Of course, then I'm just handing it to them, but oh well.

 

Well, if you think that it's something you can make money off, you can do what oddball did and roll out your service. Or just talk to oddball and make him roll it out and you get percentage.  ;) (and you guys can send me a beer for getting you together  ;D ).

(Edit: there might be other guys here who would be interested to try to roll out a service. oddball just seems like top choice since he kinda knows the market, possible business size, probably has a lot of software that can be reused from his other project, etc.)

 

If you are not gonna do it commercial, then IMO making it public and promoting it to Morningstar or whoever (Bloomberg?) is a way to go. They'll probably not gonna do anything about it, but c'est la vie. If they do, then good free publicity to you and some benefit for investors. 8)

 

Take care.

[racemize]

Edit: I probably should look at my own performance based your methodology. It's just seems like a lot of work. Plus looking at rolling window returns in private account will probably raise up the issues of capital inflows/outflows even more than looking at a (single) very long term IRR number. Also in my particular case at least, I used quite different investment approaches 5-10-15 years ago than I do now. So the data is dirty, the older data is not so useful and therefore there's not much of 10-y rolling windows overall. So, not sure I'm gonna do it. It seems interesting though.

 

For anyone who wants to do this analysis, I can provide the spreadsheet.  It's currently a google doc.  Moreover, looking at the data itself gives some more insight on how things actually developed rather than overall statistics, but I didn't want to make this essay 50+ pages so I didn't include all of it.

[racemize]

Nice essay. Perhaps you should persuade someone like Morningstar to adopt this methodology for performance evaluation.  ;)

 

Yeah, I was thinking I could send it to them.  There's a ton of really cool stuff that could be done, but it requires a lot of decent web programming.  Of course, then I'm just handing it to them, but oh well.

 

Well, if you think that it's something you can make money off, you can do what oddball did and roll out your service. Or just talk to oddball and make him roll it out and you get percentage.  ;) (and you guys can send me a beer for getting you together  ;D ).

(Edit: there might be other guys here who would be interested to try to roll out a service. oddball just seems like top choice since he kinda knows the market, possible business size, probably has a lot of software that can be reused from his other project, etc.)

 

If you are not gonna do it commercial, then IMO making it public and promoting it to Morningstar or whoever (Bloomberg?) is a way to go. They'll probably not gonna do anything about it, but c'est la vie. If they do, then good free publicity to you and some benefit for investors. 8)

 

Take care.

 

Yeah, all of those ideas sound decent.  I don't think I have the time/energy to invest in making it myself, but I know how a really good implementation would look (which would be particularly nice for brokerages I think).  Anyone have any contacts at Morningstar or Bloomberg that would actually look at this?  I'll send Nate an email too.

[racemize]

FYI, for Chou, I had to go find the S&P 500 in CAD myself, and the data isn't quite lining up with how he reports it (it looks better on his semiannual report for trailing periods than what my spreadsheet says).  I'm reaching out to him about it, so there's some possibility that results will be higher/better for him.

[rishig]

As usual, excellent essay Joel. This kind of synthesis is unprecedented - Congratulations on the great work.

[Munger_Disciple]

Good work, racemize.

 

Buffett in fact included rolling 5-year BRK performance results in the 2010 annual report. It serves as a validating point for your thesis.

[racemize]

Good work, racemize.

 

Buffett in fact included rolling 5-year BRK performance results in the 2010 annual report. It serves as a validating point for your thesis.

 

Yes, I had forgotten he had done that in the annual letter.  I should mention that, in case it isn't clear, this wasn't a new idea, but it seems to not be used very often to evaluate managers' performance.  A friend of mine who was a quant just told me this is the normal way of evaluating quant strategies and they use a really nice heat map to visualize the results.  Hopefully, I can figure out how to generate those.

[investmd]

Racemize,

Thank you for sharing this important piece which has been a tremendous amount of work. Well done indeed! Commendable you would do that and compare your own firm's results.

After reading this essay, my take for someone looking to invest $s today with a money manager, is that one would likely be well served going with almost any of the current members on the list. However, the "best of best" proven value investing options today probably include Braddock Partners, Giverny Capital & Arlington Value Capital and ? Austin Value Capital. I don't know the funds, but will look into them.

Does anyone with more experience than me, have thoughts on above or specific insight into Braddock, Giverny, Arlington Value?

Thanks,

[Cigarbutt]

investmd,

"Does anyone with more experience than me, have thoughts on above or specific insight into Braddock, Giverny, Arlington Value? "

 

I have never invested in Giverny Capital but have followed them. 

Really a fine operation with a consistent and candid approach.

It's fairly easy to access the annual reports (some info is hidden for outside investors) if you want to understand their philosophy.

My understanding is that they follow a GARP approach and do not look at macro stuff at all.

 

racemize,

 

Enjoyed reading your report and thank you for sharing.

 

From Benjamin Graham, who does not appear explicitly on your list but who, somehow, is the "father" of all of them.

 

"Common stock selection is a difficult art, naturally, since it offers large rewards for success. It requires a skillful mental balance between the facts of the past and the possibilities of the future."

 

Congratulations on your results and good luck going forward.

 

 

[racemize]

Regarding Graham, I actually tried to get him on the list.  Unfortunately, there are a few holes in his record I couldn't fill in.

 

Moreover, it appears that his record is mostly better than the S&P 500 because of Geico.

[John Hjorth]

Thank you a lot for sharing your work, Joel,

 

It's much appreciated. Impressive, what you have been able to dig up in so short time - and to process. Most likely, you have made my Saturday afternoon to come.

 

I have never invested in Giverny Capital but have followed them. 

Really a fine operation with a consistent and candid approach.

It's fairly easy to access the annual reports (some info is hidden for outside investors) if you want to understand their philosophy.

My understanding is that they follow a GARP approach and do not look at macro stuff at all. ...

 

Thank you to both Joel & Cigarbutt for guiding me to Giverny Capital and Mr. Rochon. I did a search on here, and found quite some posts - They have skipped my attention, totally. Very interesting stuff.

 

- - - o 0 o - - -

 

I look forward to an enjoyable and instructive Saturday afternoon - thank you.

[thowed]

Thanks - this is very interesting stuff and beautifully presented.

 

Also, thanks for mentioning Braddock - I had no idea Akre had a L/S strategy.  Couldn't find any details, though I presume there's a fair bit of overlap on the long side.  He's a hero.

 

And obviously I'm now fascinated to know more about Austin's strategy.

 

@investmd

 

Giverny, you've heard about, nothing to add, they're a great example of holding quality GARP.

 

Arlington - they're closed to new money, the track record is insane, I wouldn't know how to describe them, but search this site for more info.

 

Braddock - look up the Akre Focus mutual fund for details - more outstanding, long-term, quality GARP.

 

[racemize]

Thanks - this is very interesting stuff and beautifully presented.

 

Also, thanks for mentioning Braddock - I had no idea Akre had a L/S strategy.  Couldn't find any details, though I presume there's a fair bit of overlap on the long side.  He's a hero.

 

And obviously I'm now fascinated to know more about Austin's strategy.

 

@investmd

 

Giverny, you've heard about, nothing to add, they're a great example of holding quality GARP.

 

Arlington - they're closed to new money, the track record is insane, I wouldn't know how to describe them, but search this site for more info.

 

Braddock - look up the Akre Focus mutual fund for details - more outstanding, long-term, quality GARP.

 

Well Austin's strategy is just what I'm doing plus my partner, who posts on here occasionally.  The outperformance mostly came from 2010-2013 and 2016, it's been tougher in the last few years.  We did a ton of investing in financials from 2011 to the trump run-up and also invest in the usual compounders, now more than ever, given the type of environment we are in. 

[crastogi]

I've just wrapped up my latest essay, which I think will be of particular interest to this forum.  It revolves around how performance is reported and proposes a different statistical method.  I then used the method on 19 famous value investors: Warren Buffett (1957-1969), Walter Schloss (1956-2002), Charlie Munger (1962-1975), Tweedy Browne Partnership (1958-1983), Sequoia Fund (1970-Current), Davis Venture Fund (1970-Current), Tom Russo (1984-Current), Francis Chou (1984-Current), Longleaf Partners (1988-Current), Bill Nygren (1992-Current), Chuck Akre (1993-Current), Yactkman Fund (1993-Current), Francis Rochon (1993-Current), Tweedy Browne Value Fund (1994-Current), Guy Spier (1997-Current), Mohnish Pabrai (2000-Current), Bruce Berkowitz (2000-Current), Arlington Value Capital (2000-Current), and Austin Value Capital (our record) (2010-Current).

 

It has definitely shaped my opinion about which investors I want to follow in the future. 

https://drive.google.com/file/d/0BxTPR9eP5nWeNTJWOF9lUm1uOFU/view?usp=sharing

 

Thank you! 

 

Very nice

 

 

[John Hjorth]

... And obviously I'm now fascinated to know more about Austin's strategy. ...

 

... Well Austin's strategy is just what I'm doing plus my partner, who posts on here occasionally.  ...

 

I'm not trying to be rude, nor condecending, or anything else here [i have no reason or basis to be so], but this - at least to me - sticks a lot deeper. For your part, [naturally, I'm not referring to Joel [racemize] or Jeff [rainforesthiker] here] please take your time to make up your own mind. Alone doing that, will be quite time consuming, here on CoBF, by reading the board.

It has absolutely nothing to do with a "strategy" - in its ordinary business understanding - [in the meaning: "We are at "A", and we want to get to "B", in a certain sense].

 

In short, it's an investment framework, that you can relate to [no matter what's on your watch list, or in your portfolio], or you can't.

 

Austin Value Capital - Philosophy.

 

CoBF book topic : here.

[Packer16]

Joel,

 

Given your return pattern you may want to use a small cap value benchmark like Vanguard Small Cap Value as it has a similar pattern of returns as you have.

 

Packer

[Cigarbutt]

Interesting question then.

Should a benchmark be chosen based on previous returns or based on the philosophy of the fund manager?

 

 

 

[LC]

Philosophy IMHO. How else can you tell if you are skilled or if you suck? Otherwise my benchmark would be the 30 day treasury.

[racemize]

... And obviously I'm now fascinated to know more about Austin's strategy. ...

 

... Well Austin's strategy is just what I'm doing plus my partner, who posts on here occasionally.  ...

 

I'm not trying to be rude, nor condecending, or anything else here [i have no reason or basis to be so], but this - at least to me - sticks a lot deeper. For your part, [naturally, I'm not referring to Joel [racemize] or Jeff [rainforesthiker] here] please take your time to make up your own mind. Alone doing that, will be quite time consuming, here on CoBF, by reading the board.

It has absolutely nothing to do with a "strategy" - in its ordinary business understanding - [in the meaning: "We are at "A", and we want to get to "B", in a certain sense].

 

In short, it's an investment framework, that you can relate to [no matter what's on your watch list, or in your portfolio], or you can't.

 

Austin Value Capital - Philosophy.

 

CoBF book topic : here.

 

John has done a much better job responding than my lazy answer below—more evidence that we should have him on our non-existent payroll.  The book, essays, philosophy, and letters lay out the framework fairly well for anyone interested. My answer related more to how the returns have been generated than what framework or strategy we are using. So, the framework is fixed but the strategy is opportunistic.

[Tim Eriksen]

Should returns be shown gross of fees or net?

As an investor we care about net, but in terms of evaluating the manager, his skill is best represented in gross returns.

For example a mutual fund manager with a 12% net AAR is not actually better than a hedge fund manager with 11% AAR that has standard 2 and 20.

Or someone with the Buffett partnership averaging 9% is really only 10% gross while someone with 2 and 20 reporting 9% net AAR is really at about 13% AAR gross.