Sentiment Is Slow to Change: A Basketball Lesson, Part I

We realize how difficult it can be to understand investment principles. It is important to have good examples. It is even more powerful to step outside the normal investment world, where the reader already has a strong opinion.

In our classroom days we would frequently set up a different and unfamiliar problem. Students would examine and discuss the case with enthusiasm. Then we would reveal the analogy that was the point of the lesson. Some in the class would object, of course, but that was fine. Discussing whether or not an analogy is apt is part of the process of education.

Getting people to think in an open-minded way is crucial. Undaunted, we present an analysis of sentiment that uses a subject of current interest.

A Basketball Example

One way to analyze cognitive biases is to take an example completely outside one's own experience. We have found TradeSports.com to be a useful laboratory for this analysis.

Background. TradeSports, the sister site of (Intrade.com where there are various political, economic, and event markets) has real-time trading in futures contracts on the outcome of sporting events. There is a lot of volume and those participating in the market are watching the game in question. This adds a new dimension to the concept of interactive television. There is even trading pit chatter, featuring the viewpoints of a number of arrogant, opinionated participants. Their language is even more colorful than that found on the sites of stock market bloggers.

We are obviously not advocating participation in this market, deemed illegal for US residents. It is just another interesting piece of information to follow when watching the game.

The Case. Saturday's semi-final contest between North Carolina and Kansas is a great example of market sentiment. We have observed trading on many sporting events, so we urge you to take our word for the general interpretation. Readers should note that we are not offering an opinion about the respective teams. If the game were to be played 100 times, there would be many different outcomes. The general result would be much closer than what actually occurred.

The market began with a bias about which team was better. That opinion persisted, in defiance of the apparent results. Sentiment changes very reluctantly.

The initial condition was that North Carolina, the overall #1 seed in the tournament, was favored by 3 1/2 points, 60% to win on a straight-up basis. The Tradesports futures contract for a team settles at 100 for a win and zero for a loss. The chart can therefore be interpreted as the percentage chance of victory at any point in time.

Let us start by looking at the chart.

(click to enlarge)

Kansas had a very hot start to the game. Since we are analyzing sentiment, not basketball, we shall just note that with ten minutes played, Kansas had a twenty-point lead.
At this point, the futures contract still showed a 20% chance for a North Carolina victory. Let us think about this. At the start of a game, a team that is a twenty-point favorite in the point spread is over 97% to win the game. There was a clear disconnect between trading and reality, with North Carolina over-valued by the market.

A few minutes later, the lead grew to 28 points. North Carolina futures still held a minimum bid of 5, reflecting 19-1 odds. Market sentiment still held that UNC had a chance.

Now one might expect a team with such a lead to change strategy, but it is difficult to do so in practice. Bill Self later stated that his team went "brain dead" for about thirteen minutes, allowing the margin to be narrowed at the end of the first half and for the first ten minutes of the second half. Anyone who watched Gene Hackman in Hoosiers understands the difficulty in getting players to follow the coach's strategy!

Did this mean that the UNC investors were correct? The television analysts opined that the game was over. Despite this, with ten minutes to go, the lead was narrowed to five points. The futures contract got as high as 50, reflecting a continuation of UNC momentum.

The best trade of the contest was to buy North Carolina at 5.5 and sell at 48 or so, a nine-bagger in less than fifteen minutes of playing time. This profit could have been accomplished by investing in a team that ultimately lost by a wide margin, and never got the lead! One does not need to predict the result to make money, only to predict the market reaction.

In fact, this was a highly improbable result. That it actually happened seems to confirm the sentiment.

When Kansas slowed things down a bit and worked the ball inside, their lead grew once again.

As the clock ran down, the inevitability of the result became clear. It was only at this point that sentiment switched.

Conclusion

The lesson here is that market sentiment is very "sticky." Those investing in UNC futures were not just fans. Most were regular players of many sporting events, seeking a profit on this one. They began with an opinion, and the opinion was slow to shift.

The value of looking at an example like this is that the entire picture of sentiment and reality can be captured in the space of a few hours.

In the stock market, a similar process may take many months or even years. We shall explore this further.

Comments

[Paul E.Meisel]

Excellent.