Notes from the 3rd Annual New York Value Investing Congress by Marcelo Lima

Notes from 5 speakers: http://blog.valueinvestingcongress.com/2007/12/

-

.....

-

Mark Sellers at the 3rd Annual New York Value Investing Congress by Marcelo Lima

-
Mark Sellers of Sellers Capital spoke at length about the Kelly Criterion (you can read all about it in the book Fortune’s Formula). It’s essentially a betting system that guarantees that (1) you never go broke and (2) you compound your bankroll at the fastest, most optimal rate possible.

-

Of course, because it’s a formula that requires you to handicap odds and payoffs, it’s subject to the to the garbage in / garbage out pitfall of any model, and can give seemingly ridiculous results – like “bet 500% of your bankroll on this stock” – if you make incorrect assumptions about the likelihood and magnitude of winning or losing on a stock investment. It could also mean that betting 500% of your bankroll is the optimal thing to do. The strength of your assumptions really matter.

-

To mitigate the risk of making mistakes in handicapping the odds and payoffs, Mark limits his purchases to situations where the expected return of a stock exceeds 20%, the upside is 3x greater than the downside, and the Kelly Formula tells him that he should bet.

-

Mark still likes Contango (MCF), quite a bit actually, since he’s put 50% of his fund’s NAV in the stock. He considers Ken Peak, the CEO of MCF, the “Warren Buffett of the natural gas industry.” My favorite part of the story is that Ken owns 25% of the company, has never sold a share, and his ex-wife gets his salary. He’s about $10m in debt – he borrows money to live – but since his stake is worth close to $200m, he’s not so worried.

-

[Mark’s] made quite some money on Contango, but with the stock at $47, he thinks that $42 is the worst case, $57.50 the base case, and $70 the best case. The company has recently hired Merrill Lynch to explore alternatives.

-

Mark’s second idea, which is “only” 25% of his fund’s NAV, is Lowe’s (LOW). Mark believes that the odds of LOW trading forever at a forward P/E of 11 are essentially nil. Over time, one gets earnings growth plus multiple expansion. In his estimation, the base case for LOW is a value of $30 per share, with $20 being the worst case and $42 the best case. The probabilities for these scenarios are 35%, 50%, and 15% respectively.