Analysts are overrated

Meteorologists have the best job in the world - they can predict heavy rains tomorrow, but when tomorrow passes by with nothing but great sunshine, the meteorologist keeps his job and try again.

For those of you who are into the theory of forecasting, Rule #1 of forecasting states this.

Forecast are never accurate.

Nothing is 100%. Predicting weather is based on what is likely to happen. Similarly, choosing a boyfriend that seems nice enough doesn't guarantee that he doesn't stray, he's just less likely to do so.

Analyst predicts the future outcome based on past events (Markov Rule). Methods like moving averages, stochastic processes, regression, extrapolation, neural networks, weird simulation models and all the fancy forecasting methods are used to create a black box model, which the layman doesn't understand.

And presto, an analyst report is produced.

In almost every analyst report released to the market, they are normally accompanied by lots of footnotes stating assumptions. Like for example, an insurance brochure saying how fast you can grow your money in the best likely situation of 15% annual growth (and a year later after you buy the policy, the recession sets in).

Facts do not lie. Biased opinions and the way that selected data is used however do try to skew the reader's judgement towards signing up for a financial plan.

Though analysts are supposedly objective, there is normally a motive for the report. Even a reporter who wants to report objectively may tone down his criticism because of his editor, or intervention from government-related agencies.

My first job upon graduation was an analyst at some government-linked company. I had complex mathematical models to do some forecast, but the numbers that I came up with were often deal with sceptism by the experienced people. Eventually I realised this.

The correct answers are what the client wants to listen.

And also to all aspiring analysts, here's a tip from Uncle Shingo.

It's better to have a simple forecasting model that the client understands, than a more accurate model that acts like an unknown black box.

Moving on, investment reports have always been accompanied with words like caveat emptor, which means buyers beware. Always take an analyst report with a pinch of salt, digest the facts, but don't take their recommendation blindly. No one is responsible for your own losses (and wins) but yourself.

A stock market newbie often make this mistake by following friend's advices or overweight calls from reports. I know, because I have been through that phase, and lost money because of it. And I have got no one to blame but my own laziness to spend time reading up.

Some time back, a research team pit a bunch of experienced traders against a bunch of kids in picking stocks. And as you might have guessed correctly, the noobs won the experienced traders. Sometimes you are better off making a wild guess than to take the advice of an "expert".

On an interesting note, JP Morgan, a leading financial service firm, recently published a paper on the World Cup. It predicted the winners of each round, and interetingly, if you followed them, you would have lost money.

What JP Morgan predicted
Netherland beat Denmark
Cameroon beat Japan
Italy beat Paraguay

And the result were as follows
Netherland beat Denmark (2-0)
Japan beat Cameroon (1-0)
Italy drew Paraguay (1-1)

Any football savvy guy could have easily make bets that will give JP Morgan a run for its money.

To the guys at JP Morgan, just stick to what you do "best".