Does Historical Performance Have Predictive Power for Chinese Banks?

On February 6, I wrote an article titled “Chinese Banks: A Non-Traditional Approach to Predicting Their Performance”, in which I described how I rate three Chinese banks' future performance based on their historical non-performing loan [NPL] problems. My argument was simply “bad habits die hard”, especially in China, where management appointment is more based on relationship (though it's changing).

At a time in which all analysts were promoting ICBC (saying ICBC is the largest bank in China, economies of scales, with the most advanced IT system, etc) and Bank of China (saying BOC is the only Chinese bank with international presence), I analyzed them using a totally non-traditional approach and rated China Construction Bank (939.HK) as the best performer versus ICBC (1398.HK) and Bank of China (3988.HK)

Today, I’ve revisited their relative performance both from the date of my article and also their recovery capability after the February market correction.

The following charts will speak for themselves.

From the date of my article on February 6:

click to enlarge

From the early March market correction:

click to enlarge

In my last article I said:

...We also noticed that CCB was the first state owned bank listed overseas. This is by no means a coincidence. The Chinese leaders have to ensure the success of the first listed so as to set a stage for the others; as such, it is reasonable to assume that the leaders should have selected the best bank to be the pioneer. Who else would have better insider information than the Chinese leaders? It would be your decision to bet with the Chinese leaders’ choice or analysts’ DCF models.…

In general I do not believe in DCF models; I think even the CEOs and CFOs are themselves unable to tell you their firms' grow rates for initial years/terminal growth rates, nor could anyone predict the interest rates for next 10 years (or whatever horizon in the model). While DCF analysis is very sensitive to those assumptions, how valuable is this time-value based modeling of a firm's fair price? I would rather use P/E comparables if I want to have a gut feeling of a firm's value.

In my analysis of the Chinese banks, my approach is totally unscientific. While it's a common belief that historical data cannot predict future performance, that may not be completely true for China and I would rather give my vote of confidence to our great Chinese leaders than forecasts based on DCF models!

Comments

[JW:]

why letsthinkchina.com can not be visited?

2007 Jun 12 05:59 AM