My personal search for alpha

Thursday, October 30, 2008

My earliest experience investing didn’t go so well, but that didn’t completely crush my belief that there’s alpha out there somewhere. I know, for instance, that top venture capitalists have consistently delivered higher (risk-adjusted) performance than other securities (The full argument can be found in Paul Gompers and Josh Lerner’s The Venture Capital Cycle). But that doesn’t seem to be true of mutual funds. What’s the difference? It looks like venture capitalists get to take advantage of network effects (the best VCs have lots of connections, making them better, giving them more connections, adding up to a substantial advantage over median VCs), and market illiquidity (there are very few VCs (and fewer “excellent VCs) providing funding for an enormous number of potential entrepreneurs). These two factors add up to a unique (non-public) informational advantage and strong bargaining power. Mutual funds, on the other hand, rely on mostly public information and trade primarily in liquid markets.

“Illiquid markets and unique information” sounds like a scam, but if venture capitalists can do it, maybe there are opportunities to be found in other areas as well. After reading Amarillo Slim's amusing biography, I realized I’d already participated in a number of illiquid markets, including some with unique information whenever I engaged in proposition betting -- e.g. wagering on who would win a basketball game or which member of our dinner party would arrive next. The reason why these are so lucrative is not only because people don’t always judge probability well and are risk averse, but also because entry into these bets is restricted, making the market illiquid (ultimately allowing 'incorrect' prices and, therefore, high expected returns/high expected losses for the participants).

For instance, in a class I had a few years ago, a professor was explaining expected value and decision trees with the following game: "You pay $2 for a chance to flip a coin. If it comes up heads, you get nothing; tails you get the opportunity to pay $2 to flip a second coin. If the second coin comes up heads you get nothing; tails you get paid $11." The expected value is positive (50%*[-$2] + 25%*[-$4] + 25%*[-$4+$11] = +$0.75), but three quarters of the time you lose money. It was meant to introduce the concept of expected value, then illustrate one of its limitations. After working through the example, the professor asked if anyone would like to play and about half of our class of 90 raised their hands. Then he asked who would like to play if the stakes went up by a factor of 10. Only one person was willing to risk $20-40, even though the expected value was positive. There would certainly have been other willing participants in the world, but they weren't given access to the game -- the market was illiquid because it was restricted.

This left me thinking about poker, proposition bets, and illiquid markets with imperfect information. Eventually, I found one in baseball.

Read more...

Good argument, questionable chart

Tuesday, October 14, 2008

The New York Times had a graphic today showing how the economy fared under Democratic and Republican presidents since Herbert Hoover. They seem to be making a shortened and simplified version of the argument presented by Larry Bartels in his book Unequal Democracy (For those who don’t have a book-length period of free time, Steve Benen summarizes).

While I think Bartels’ argument is fascinating, I’ve really got to take issue with the Times' graphic. They obscure the progression of time and the length of term that each president had. Yes, the S&P 500 grew 10.8% under Gerald Ford, but he was only president for about 2 years. And yes, the S&P 500 grew 15.2% under Clinton, but was that his economic genius, or the continuation of a trend started 20 years earlier and lasting through three administrations under both parties?

I think the time series clarifies things greatly. Here’s a recreation of the original NYTimes graph, which sorts by party and orders by growth rate:

Here's a new one that orders by year, and shows term length:

I think the new chart does a much better job of showing the sequence of presidents, which helps the reader understand larger trends. Maybe the economy is more likely to grow under a Democratic President than a Republican one, but if that's the case, I'd like to see it shown more robustly.

Read more...

Why I'm in love

Sunday, October 12, 2008

The other day I was reading xkcd and came across this fabulous comic.


It really spoke to me, since I think one of the greatest parts of a relationship happens when one partner recognizes and indulges a socially unpopular part of the other. It might be embracing something geeky like in this comic, or providing a hidden pleasure like Peeps, but it’s touching when it happens. Me – I like graphs.

For my anniversary with Sarah, one thing I got her was a pack of “fruit”-flavored Mentos, which we always have enjoyed on our vacations (consumption-based enjoyment, not science experiment enjoyment). The next day she sent me this graph of the contents of the Mentos pack.


I'm in love.

Read more...

  © Blogger template Writer's Blog by Ourblogtemplates.com 2008

Back to TOP