Corner of Berkshire and Fairfax Fund

[Ross812]

The idea of starting a crowd based portfolio to track the best ideas of the board has been tossed around a bit.

 

Above are the 60 topics in the investment ideas forum with the most posts. Some topics were removed if the stock no longer trades or there where no posts on the topic in the past 6 months.  Each member gets up to 5 votes. Choose the companies you believe will have the best total return. The 10 companies (cash is an included asset class) with the most votes will be bought in a portfolio with a $1,000,000 starting balance. Their proportions will be determined depending on the number of votes they receive.

 

The portfolio will be public for anyone to see at the following link:

 

https://docs.google.com/spreadsheet/ccc?key=0AivVdWOTQE2JdGQzYkFSQUg2eUR3UVRTcThFYWpBelE#gid=21

 

I will start a new pole which will stay open for 10 days with the 60 most discussed companies at the beginning of each quarter and re-balance accordingly.

 

If anyone has a better idea for a publicly shared portfolio that looks better or has more capabilities than the google spreadsheet that would be great. Let's see where this goes...

 

-Regards,

 

Ross

[moore_capital54]

This is a good idea. Except instead of providing a list of securities you should supply 4-5 empty fields and let us enter the securities. That way we are not limited by the ones you pre-selected.

 

 

[constructive]

I already made my 5 picks (AAPL, DTV, INTC, JPM, LUK) but here's a write-in suggestion: Short CRM.

[twacowfca]

You've got a great idea, but the methodology gives too much weight to "has been's".  You might also sort out a list of the twenty most popular stocks over the last three months that don't appear on the list that goes back farther in time and have a separate poll for them.  Then we can see if the best of the new kids on the block beats the old gang.  :)

[ASTA]

I totally agree twacowfca. Please everyone vote for my stock for parents ideas :D They are all plus :D

[Packer16]

It will be interesting to see how this will turn out because the most popular value stock may already have some aspect of popularity factored into its price.  Another idea is the value rank the stocks by a metric such as EV/EBITDA or MktCap/FCF or P/BV, chose the cheapest and value-weight the portfolio by how cheap they are.  This will at least reduce the amount or popularity bias in the fund.

 

Packer

[Ross812]

You've got a great idea, but the methodology gives too much weight to "has been's".  You might also sort out a list of the twenty most popular stocks over the last three months that don't appear on the list that goes back farther in time and have a separate poll for them.  Then we can see if the best of the new kids on the block beats the old gang.  :)

 

I like the idea. Do you have an idea to figure our which ideas have been "trending" in the last three months?. My original idea was to only list stocks that had been mentioned in the last three months but the board is so active that the list comes to 210 names! There has to be a cutoff somewhere. Do we do the last 60 names mentioned? last 100? 60 names gets us back about 4 days... 100 gets us back about 2 weeks...

[jay21]

You've got a great idea, but the methodology gives too much weight to "has been's".  You might also sort out a list of the twenty most popular stocks over the last three months that don't appear on the list that goes back farther in time and have a separate poll for them.  Then we can see if the best of the new kids on the block beats the old gang.  :)

 

I like the idea. Do you have an idea to figure our which ideas have been "trending" in the last three months?. My original idea was to only list stocks that had been mentioned in the last three months but the board is so active that the list comes to 210 names! There has to be a cutoff somewhere. Do we do the last 60 names mentioned? last 100? 60 names gets us back about 4 days... 100 gets us back about 2 weeks...

 

Maybe a google spreadsheet where each participant types in five tickers.  Then use some type of count function?  That way there might not have be too much work on the organizers end. 

 

One question:  How do we vote to sell?  Do we update our picks every month, day, whenever? And then the portfolio is rebalanced?

[Ross812]

It will be interesting to see how this will turn out because the most popular value stock may already have some aspect of popularity factored into its price.  Another idea is the value rank the stocks by a metric such as EV/EBITDA or MktCap/FCF or P/BV, chose the cheapest and value-weight the portfolio by how cheap they are.  This will at least reduce the amount or popularity bias in the fund.

 

Packer

 

So select the top 10 by popularity then value rank the companies and by a metric. The problem is what metric do we use? There is not really one metric that measures all companies equally. What about (P/BV) / (Company's 5 yr Avg. P/E) ?

 

 

[rkbabang]

Another problem is weighting.  I might pick 5 stocks thinking that my #1 pick is 10 times better than my #5 pick, but you are going to weight them equally.

 

You might assign each person 100 points and let them distribute those points to as many ideas as they wish.

[Ross812]

Maybe a google spreadsheet where each participant types in five tickers.  Then use some type of count function?  That way there might not have be too much work on the organizers end. 

 

One question:  How do we vote to sell?  Do we update our picks every month, day, whenever? And then the portfolio is rebalanced?

 

The portfolio will get rebalanced every quarter with another poll. If a company falls out of the top 10 it will be sold.

 

I like the google spreadsheet idea, I'm just not sure what kind of participation we would get. The write in would be a popularity contest of sorts as well. In the next round rebalance (End of September beginning of October) we will add in some sort of write in. There will be some more discussion on how to accomplish the write in I'm sure.

[Packer16]

I think the metric us depends upon the type of firm.  Financials P/BV, mature firms MktCap/FCF, growing firms EV/EBITDA.  With this you can compare across firm types but between firm types you may need to estimate default value for each multiple.

 

Packer

[cofabmd]

If a CBF portfolio was housed on Covestor, board members (and others) could invest in it.

 

Something like this: http://covestor.com/mco-investments/margin-of-safety

 

 

[compoundinglife]

If a CBF portfolio was housed on Covestor, board members (and others) could invest in it.

 

Something like this: http://covestor.com/mco-investments/margin-of-safety

 

I tried out co-vestor a few years backs and the issue I had is that did not track some of my more esoteric securities. Prefs that were not on the NYSE, some OTC stuff etc... They may be better about that now.

[Ross812]

If a CBF portfolio was housed on Covestor, board members (and others) could invest in it.

 

Something like this: http://covestor.com/mco-investments/margin-of-safety

 

I thought about Covestor actually. If this works well, I'd be open to starting a model portfolio on covestor. Fees collected would be donated.

[SharperDingaan]

Couple of suggestions:

 

A Standard Investment Policy outlining objectives, risk tolerance, max/min portfolio weightings, permissible  instruments, leverage, strategies (long-short), etc. Addendum outlining administration mechanics, selection procedures, fees, etc. Make the whole thing a charity foundation & select a manager to run it. Fund through one of the crowd sourcing sights, & make all contributions donations. Allow the foundation to issue medium term debs to raise seed capital & achieve the scale to cover its costs. 4-10 patrones with multi $M injections. Commercial terms. Set a minimum foundation payout/yr from day 1 & fully index to inflation. Fund a lump-sum investment 'X' Prize with the medium term growth, & make it a game-changer.

 

Recognize that this is all new frontier. It could not exist without the internet, crowd sourcing technology ($, capacity, communication, competitive co-operation), the modern approach to charity (results measured by business metrics), & a large enough group of wealthy like minded patrones schooled in value investment. It is a modern version of the old masonic guilds, a successful result will set the template for others, & it is highly likely that the industry will go out of its way to see it succeed.

 

It is generally accepted that X prizes spur break-through innovation, & that it is better suited to NA personalities. This is not a time for the kaizen approaches of Asia, & is too early for the precision or miniaturization genius of Germany & Japan.

 

Micro-financing used to be an 'impossible' idea - as it was just not 'done'; now millions around the world benefit from it. So why is it that supposedly this cannot be done with the investment industry?

 

A little rebellion is great for the soul!

 

SD

 

[Valuebo]

Great post SD. I was thinking about something in the same lines after this topic got posted but you clearly have some great insights already. This could really be a long term winner for the board, its members and the chosen charities. I hope some are willing to run with this idea.

[Guest deepValue]

Why not make it a 501©(3) and free the fund from those pesky taxes (and incentivize seeders to seed more). It usually takes 6-18 months to get c3 status, though.

[jeffmori7]

I think there are two different things here, one is a fictionnal portfolio to see how it wold perform, and then there is the idea of launching a real fund, be it charity or not. In the short term, we have to find a way to implement an easy mechanism to track the overall performance of the ideas generated by the board. In the same time, we could look at a long term project of starting a real thing.

[ASTA]

I would gladly help but the problem is I am a European investor so how does that affect the system. Are the trades reported to investors so they can copy them in his or her country?

[mikazo]

The portfolio will get rebalanced every quarter with another poll. If a company falls out of the top 10 it will be sold.

 

How would you vote for new ideas while maintaining vote popularity for your old ideas so that they have time to play out and realize value/growth? I wouldn't want to have to vote for the same 5 time after time, just because the market hasn't recognized their value yet.

 

Awesome idea though, I love this whole concept.

[Ross812]

The portfolio will get rebalanced every quarter with another poll. If a company falls out of the top 10 it will be sold.

 

How would you vote for new ideas while maintaining vote popularity for your old ideas so that they have time to play out and realize value/growth? I wouldn't want to have to vote for the same 5 time after time, just because the market hasn't recognized their value yet.

 

Awesome idea though, I love this whole concept.

 

This whole idea is a test on crowd sourcing a portfolio. If a poster still believes his (or her) five picks are the most undervalued or have the best chance for total return it would not make sense to choose any differently. The next poster is going to choose their best to pick 5 stocks from the list. Is the collective effort of everyone on here presenting ideas and debating their merits sufficient to build a portfolio with a positive return?

 

I see a lot of votes for SHLD right now. If SHLD was at $80 right now would it have as many votes? It may not even make the top 10 at $80. The thinking is if SHLD goes down to $30 even more members would be willing to select it and the allocation would increase.

 

I'm starting to think 5 picks may be too many. Maybe only 2 picks would be better? Also the Q4 Poll will have 100 names. Packer talked about picking a metric to base portfolio allocation. Maybe an identical sister spreadsheet with a different allocation methodology can be set up to compare the best approach. The best ideas still may float to the top via polling. 

 

Taking SharperDingaan suggestion. A crowd sourced portfolio based on the message board has the potential to go a long way. 

[constructive]

I'm starting to think 5 picks may be too many. Maybe only 2 picks would be better?

 

I like 5 picks per person. I suspect top 2 picks would result in more polarizing, risky bets whereas top 5 picks builds more of a consensus.

[racemize]

There's also some question of what we are picking.  Are we picking what we would choose for our own portfolio if we only had 5 choices?  Are we choosing what 5 choices will likely have the highest return?  Are we trying to balance the risks in the 5?

 

For example, I picked WFC among some others, which I would not do if I were going for highest returns (e.g., it was one of my conservative picks).

[constructive]

There's also some question of what we are picking.  Are we picking what we would choose for our own portfolio if we only had 5 choices?  Are we choosing what 5 choices will likely have the highest return?  Are we trying to balance the risks in the 5?

 

For example, I picked WFC among some others, which I would not do if I were going for highest returns (e.g., it was one of my conservative picks).

 

I compared the list to my holdings, and picked my 5 largest holdings. (Actually, I only own 4 stocks on the list so added JPM from my watchlist.) I assume others do something similar.