Wednesday, August 12, 2009

Lies, damned lies, and statistics

"Lies, damned lies, and statistics" is part of a phrase attributed to the 19th Century British Prime Minister Benjamin Disraeli. If you believe what the government is feeding you then good luck.

I am far from an economist so I cannot say that unemployment is what it is. But if you stopped looking for work you are not counted as "unemployed."

The Fed held rates steady today - no surprise. They also said they'd slow down their purchases of Treasury securities which is sort of the same thing as raising rates. Less liquidity is less liquidity.

The point of this post is to warn you not to believe anything but the market when you try to figure out what the market is doing. The technical tools are working a bit off but at least they are derived from the market. Economic statistics are man made, subject to revision and are no better than funnymental projections of next years earnings.


Ramu said...


When you said, technical indicators are bit off, I realized something that I should ask for your opinion? Aren't the computers that are trading following these indicators? Are they tweaked too to follow the momentum rather than technicals?

I heard from one guy the other day on CNBC that "High frequency trading makes up 50% of volume" and its "day trading with computers". Do you think something is brewing behind the scenes for a sharp sharp pullback? Like the black monday?


Quick Takes Pro said...

I cannot say what they are tweaked on. But if they are front running customers is it any wonder the slower technicals - as in hourly bar charts - are not up to snuff?

I think we need a whole new set of technical analysis based on different things than old Eddie and Mags ever envisioned.

Amalan said...

I must disagree. Statistics don't lie, people do. It's easy for people to blame statistics, but do we realize that without statistics we could have never quantified risk? Read the famous book "Against the Gods" by Bernstein and you'll know that the advent of statistics as a major tool in the 1500s made our approach to risk vastly different from prior centuries and allowed us to make significant progress in the subsequent centuries.

But I do get your point, which is why we should always question two things when it comes to statistics: how/what was the data collected, and is the interpretation correct?

On your Barron's article today, you speak of the significance of 950. Similarly, I think even idling above 1000 is significant in that if it breaks below 800 on its way down, it will be a decline in excess of 20%, a number that has some psychological significance due to the definition of a bear. So, I am thinking we have risen to a point where the lows probably won't break 800 (but if it does, watch out!).