No Random Walking!

By ASIA, I am not talking about the Asian continent. It's ASIA as a stock. Among the few companies that are actually growing in this terrible market is AsiaInfo (Nasdaq:ASIA), a telecom software provider to Chinese telecom giants. It is benefiting from the Chinese telecom restructuring since last quarter. When the service providers (CHL, CHA, CHU) embrace new customers and initiate new services after the restructuring, their billing, customer-relationship system and customer-intelligence systems have to be upgraded. As a result, ASIA's revenue has increased 37%(?) last quarter over last year and the company is seeing similar growth in quarters to come.

Technically ASIA has one of the most beautiful charts in the whole universe (see chart below). It was dragged down in September by the financial crisis but has come back mightily: the price has almost doubled from the low a couple of monthly ago. With a forward P/E of about 15, it is still not very expensive for a high-growth play.

My screening program has revealed a few other companies that have done quite well last three months (for example, PETS and GB). I am sure each has a compeling story behind the strength. It is probably worthwhile to find out what the stories are. On the other hand, these companies, ASIA included, need to compete with other beaten-down 'value plays' for new money, so their near-term performace may not be as great as others which can jump 20% or more in a single session. However a disciplined investor would still prefer growth stocks like ASIA.

Winter is coming, but it still looks like fall in north Dallas. Here are some pictures that I took around or in my house ... ... Life is still good, no matter what happens in the market. Have a good weekend!

 










以下这个小耳朵可以接受近20个中文节目，包括我常看的凤凰卫视和湖南卫视。

... one day after I expressed optimism about the market going forward: Auto Bailout Collapses on Wages.

Hate to see the market go down, but I was with the republicans who insisted on wage cutting to UAW workers and voted against the bailout plan.

Over the years globalization has resulted in equalization of end-product prices,   access to information and skill differentials among nations. Most people have benefited from this equalization process. However one area that has lagged is the equalization of wages (and currencies which are really related). In what right should an American worker makes as much as ten times of wages  of someone in Shenzhen who does the same job but works twice harder? Something has to give. UAW is not getting it!

On a positive note, today may be the day I've been waiting for: blood on the street.

On other news: the minister of China's all-powerful Ministry of Industry and Information (MII) talks about the severe economic 'slowdown' in China. He also states that China’s government may buy up steel stockpiles and that the issuance of 3G licenses is imminent.

Independent economist Andy Xue thinks the recent market strength will be short-lived. [he has been right about the Chinese market over the last couple of years]. He said there is too much inventory [of almost everything from iron ores, steels to apartments on sale] in China that it will take months to absorb. The increase in the production capacity over the recent past has made the situation even worse.

Update a day after: The market essentially ignored the bad news. Add the fact that last week's terrible employment data didn't do any damage to the market, I feel more confident now for the weeks ahead. Cheers!

Just as I hoped for a couple of days ago, Nasdaq turned green today on my chart. A December rally in the order?  Highly likely! The economy is indeed bad but the degree of fear is not justifiable! The current valuation is assuming that most of small to mid-cap companies, at least for those I keep track of, would disappear in a couple of years and even cash is not safe any more!

Unless there is a shocker coming out of the Wall Street or Motown or a war breaks out, I am predicting a market rally, probably until end of January! (Don't take my words for it though!)

Machine learning(ML) and data mining(DM) are catchy words right now in computer sciences and related job market, thanks to the demand created as a result of the explosion of available unstructured data on the web as well as databases in the business world. Yet many traditional statisticians feel unfairly side-lined in the action: after-all there isn't much new in ML and DM that they don't know already, they so claim.

Below are links to two nice articles comparing the three fields.

• Statistics vs. Machine Learning, fight! - Brendan O'Connor's Blog
• Data Mining and Statistics: What's the Connection?, by SLAC's Professor Jerome Friedman.

Conclusions are: the statisticians have themselves to blame, which I agree. On the other hand I think it has been inevitable that ML and DM have emerged as new fields. Though the problems people have to solve in ML and DM are similar to traditional exploratory data analysis and model inference - which is where statisticians' strength is, but the scale and number of dimensions presented in ML problems calls for scientists and engineers of differing focus.   In that regard, ML and ML are similar to geostatistics when compared to the field of traditional statistics.   

Some other notes: ML, DM (and geostat) care little about sample design (may be they should); Confidence interval of predictions aren't used often, replaced instead by more convincing approach of cross-validations. ML/DM/Geostat pay a lot of attention to algorithms and their efficiencies and accuracies, while statistics emphasizes on assumptions, theoretical/mathematical justification.

The charts, as shown here, indicate that the markets in HK and mainland China have turned up. This is the first time in more than a year that they have become bullish. Hopefully the US market will follow.

... by Michael Lewis, who wrote a prophetic book Liar's Poker nearly two decades ago.

The End of Wall Street's Boom - National Business News - Portfolio.com