Some thoughts on May 25

After reading Mr. Cho's article, I realized that the people in Shanghai have used their profit in the stock market to buy the flats in Shanghai. Originally, the housing market in Shanghai was suppressed seriously, so that the housing price was hardly found to have unanimous growth rate. Recently, the turnover and the price for the houses in Shanghai have started rising since April, and the rate became more conspicuous in May. He also mentioned that the agency also accepted the foreign currency such that some restrictions have been lifted.

On the other side, in Beijing, the people are somewhat different. They prefer selling their houses in order to instill their capital into the hottest stock market. Some of them may even borrow money by using the houses as the bilateral. The people still prefer adopting the valuable goods as the bilateral in a bid to grasp profit in the bullish Chinese stock market.

Here I postulate a serious point. The Shanghai people are traditionally smarter in the financial activities since Shanghai is traditionally the financial center in China. People in Shanghai have been exposed to the fluctuating stock market, commodity markets for so many years, whereas people in Beijing are more relaxed and concern in the politics, culture and academic activities. My wild guess is, the stronger side who controls the capital has passed their stock certificates to the weaker side, which does not have much information and reasoning in the investment activities.

Can this become an important and directly related contrarian point to sell the stocks before the bearish trend comes? Excluding the above news, I found that there have already been two news regarded as the contrarian points. One is, the CEO of the Chinese spirit Maotai has been caught to forge the financial statements. Another is, Yang Baiman, who is a legendary and respectable investor in China, alleged that he has sold all the stocks for fear that the bubble is now unsustainable.

* * *

One more hint...

It is somewhat hard to guess. I stayed in this city for nearly ten months two years ago. It's old but nice!

Implication of time series II

The data I use is the MSCI Hong Kong index, Standard & Poor 500 index and the Shanghai A-share stock index. The period I use is from Jan 1992 to April 2007, where I adopt the weekly return data since the daily return data consists of much noise.

The models for Hong Kong and US are fitted by the maximum likelihood estimators and for China, it is fitted by the least square estimators, since later I found that there is no any GARCH effect. For the maximum likelihood estimation, the distribution I used is the t distribution (standarized) instead of the normal distribution since the t distribution is suitable to describe the shape of the return and error terms. The models are illustrated as follows*.

From this, rhk indicates the return of the Hong Kong index; rus indicates the return of the US index; rsh indicates the return of the China index. epsilon is the error terms. sigma is the standard deviation (volatility) of the errors. One may think that the stock index returns are generated from the process above with the errors specified by the t distribution.

There are some important implications from the above models. Here I would like to describe the characteristics of different return series.

- Hong Kong's one does not depend on any past information of either the returns or the error terms, while US depends on the past returns. It is somehow surprising that it is much harder to predict the future return for Hong Kong stock index, as one may think that US stock market should be more efficient. As for China, it is as expected that China stock market is easy to be predicted since it is significantly dependent of the past returns and the error terms.

- Both Hong Kong and US stock markets do have the GARCH effect such that another equation to estimate the dependency of the volatility can be well-fitted. However, for Chinese one, it was found that fitting the GARCH effect inside Chinese model is not effective such that we had better ignore the GARCH effect for the Shanghai A-share stock market.

The GARCH effect indicates the volatility clustering, where the stock price or index fluctuates a lot in a "grouped" period. This cannot be found in the Shanghai A-share index. The possible reason is

- China stock index does not have any short selling mechanism and stock index future, which may feedback the stock index through the dynamic hedging, stop-loss to create large fluctuation.

- China stock index has the characteristic that the surge and dip are limited in the intra-day trade, i.e. 10% upper bound and lower bound.

From the time series, this scientifically illustrates how the Chinese stock index is different from the other stock indices. I just wonder the above reasons are the cause of this crucial difference. Will there be any structural change if the Chinese stock market allows the stock futures trading?

--------

* The above models are estimated through the EView, the econometric software.

Errata: epsilon_t/sigma_t^2 ~t(.) should be replaced by epsilon_t / sigma_t~t(.).

Implication of time series I

(In this topic, the concepts are rather complicated and abstract and it is better that you have acquired some basic concept in statistics and mathematics. I will not use much mathematics, though.)

I would like to discuss a well-known tool in econometrics and finance, which is rarely mentioned and noticed by the ordinary investors in the public. This is called the time series model. These models were developed by Box and Jenkins. The objective is to model the data such as stock price, GDP or price level measured in different time in order to describe the characteristics of the data and then forecast the future trend. I will discuss about the time domain approach which is frequently applied to the finance.

In the time domain approach, firstly the stock price has to be transformed into return. Afterward, the correlations between the return today and the return in n days ago are calculated. Box and Jenkins found that there is a set of model called ARMA model used to fit the return series. AR means that the stock return series depends on the current and lags of the return, while MA means the stock return depends on the current and lags of the errors.

If the stock return is fitted (by either least square regression or maximum likelihood) and does not consist of AR coefficients and the MA coefficients, then it is possibly running in the random walk, which is the idealistic case satisfied the weak form of efficient market hypothesis. Otherwise, the stock return can be forecasted by using the past returns.

The above ARMA model assumes that the volatility of the return is constant throughout the time, which is not reasonable enough. As we can see, sharp drop usually accompanies sharp rise or further sharp drop, where future direction may be inconclusive. Engle (1982) and Bollerslev (1986) developed another important model called GARCH model. The objective of this model is used to model the volatility of the financial time series such that G indicates that the volatility depends on the volatility in the previous periods, while ARCH indicates the volatility depends on the previous square terms of errors.

The ARMA model and the GARCH model can be joined together to model the stock return series. After fitting the stock price series, it comes to some intuition about the characteristics of different markets. In next part, I will post some examples in Hong Kong, US and China stock indices.

----------

Reference:

Box, G.E.P.; Jenkins, G.M.; and Reinsel, G.C. (1994) Time Series Analysis: Forecasting and Control. N.J.: Prentice-Hall International, Inc., 3rd ed.

Enders, Walter (2004). Applied Econometric Time Series. N.Y.: John Wiley & Sons, Inc., 2nd edition.

Wikipedia: Time Series, ARMA model, GARCH model

Relax man~

Recently I usually see one characteristic, where most people would regard that 10% rate of increase in profit is not attractive at all. All are aiming at 20%, 50% or even 100% return so that these should be the most probable return. In China, amid the frenzy bullish moment, the people dare not to talk about their return when they only grasp 100% return. They can only be content if their return is 400% or 500% of their principal.

I myself do not fully aim at the marvelous return. Instead, I maintain myself to get the happiness through the hardship in seeking the value or growth stocks in different periods. Unlike speculation on some surprising stocks, the imminent return is not the most critical matter. I just train myself as a lubricant to improve the market efficiency. I do not need to be worried and nervous about the possibly high volatility generated by the ST stocks. Having the experience in January and February, I hope that my sentiment can be further improved in the next turnaround period.

Yesterday when I had the jogging, I could find that people talked about stocks. Everywhere are about stocks, where people are skeptical about the economy, but frenzy about the small cap or penny stocks.

* * *

To relax, I just share some photographs. This is the one I love.

Do you know where it is?

Anecdotes on May 22, 2007

Finally I finished most of the tasks in this term. In this tough period, I can hardly execute trades and notice the stock market trend. I need sometime to revise the materials and news and until now, I haven't finished yet. In the early period of May, there are mixed but volatile sentiments in the market. Therefore, the market consists of much uncertainty.

- Revision

Firstly, PetroChina discovered the greatest oil field in which they have ever found since 1960s. In turn, it dramatically changes the destiny of PetroChina and the oil reserve in China. PetroChina surged for approximately 13% within a day.

Secondly, the QDII, which has been discussed for certain weeks ago, was clearly announced to allow more capital from China to invest in the stock markets in foreign countries, thus Hong Kong is most probably benefited by this news. The result is that the market seems to over-react as the degree of surging is conspicuously slowed down after the nearly 600 points increase in HSI.

Thirdly, on Friday after the markets closed, the new style of monetary policy is carried out where three policies including the exchange rate policy enable the government to reduce the liquidity flooding to the overheating economy and the current account surplus between US and China. Today China just announced another fiscal policy to impose the export tax in order to dampen the current account surplus further. It resulted in a sharp drop in A-share and sharp rebound yesterday, and the disparity in A-share (still increased) and B-share (decreased sharply with high volatility) today. The Hong Kong investors felt it is worrisome, thus the HSI and HSCE dropped gradually.

Fourthly, yesterday Kuwait determined to revalue its currency by converting its fixed exchange rate regime to the mixed rate (a basket of currency) regime. This creates the potential risk that US dollar would be further depreciated since other Arab nations will follow Kuwait's decision as they feel that the inflation is costly to their countries. The instability of US dollar is crucial for many locations including Hong Kong.

- My views

If I solely use the simple technical analysis for HSI and HSCE, I can merely see the signal which indicates that the market is overbought by using the MACD. There is only slight divergence for the RSI. Although the index has broken through the Bollinger channel, it does not create the conspicuous sharp drop afterward. From my point of view, the volatility level created by the QDII news has been dissipated and is not strong enough to create the sharp drop.

Hong Kong is at the embarrassing situation where PE is appropriate but the neighbor is over-excited. People in China are not sensitive in the interest rate policy and continue to speculate. The interaction has gone to the inflation and now it is the war of negative real interest rate we Hong Kong people have faced in early 1990s. In view of the attitude of the Chinese government and its characteristic hierarchy and expectation, the surging rate of the stock market can be slowed down only if the lagged effect is shown later by the monetary and various policies.

Having read Peter Tsang's advice in his commentary, I realized that I nearly forget the effect of the Japanese Yen in the financial market. The position of the Japanese Yen becomes an unanticipated shock to the current conflict on the destiny of Chinese and Hong Kong stock markets through the worldwide liquidity fluctuation.

Most importantly, the current situation of the world economy has slowly affected our daily life, as we can see that the imported inflation from China is so salient. Under this situation, I have to be allured to buy some assets. As I feel uncomfortable on the inflation and the depreciation of the Hong Kong dollar, I can only carry out the portfolio handling action to buy more assets (currently, stocks) to hedge my overall portfolio. However, the law of tub enlightens me to buy the (hopefully) non-index value stocks as a means to contribute to the market efficiency. Moreover, I would like to escape from the index shocks and their relatively high value status. Now this is also a kind of training on my stock picking technique.

* * *

I also get one lot of Belle. I hope I can write my point on the short-run position of this stock. Currently I prefer selling this soon.

Recent anecdotes (May 6, 2007)

- My stuffs

I don't know how many times I have said that I will be busy in April and May. In April, I miraculously wrote such amount of essays. It seems that it would be a bit hard to write more in May, since there will be lots of deadline coming soon.

Recently, with regard to the works related to the stock market research, I have done the following things.

a Interrelationship of stock markets between Hong Kong, China and US;

a Revise my technical analysis techniques;

a Devise some efficient simulation and bootstrapping techniques. I just finish the most critical step, but it is still at the starting point.

In the coming future, possibly I have to follow up the above things. Moreover, I will need to focus on trading rule evaluation in Japanese stock market. As for the stock-picking technology, I still get difficulties to revise the old techniques well. I am still thinking if it is the case that simple is the best.

By the way, they are not the main stuffs in my works. I hope that there will be conspicuous progress.

- The market

Since I am a bit busy and hard to have spare time to do further analyses, I had better write some market sentiment for the evaluation.

Everyone in the market turns to be bearish in May, but my mum thinks that May seems not to be bad. She is proud of her victory in several trades in April. My father once met a friend who earned five times of the investment amount in the A-share market and intended to persuade my father to enter the A-share market. Well, my father is always risk-averse such that he can hardly be enticed. My friend from mainland started trading the A-share equity since January this year. Most importantly, he gradually focuses on the monitor to keep track of the stock market trend and the internet radio channel. Recently he feels bearish on the A-share market, but he said that mostly the people in China are absolutely bullish.

So, can you figure out what will happen next?

- Notice: This is just an advice to risk-loving investors

Actually many have mentioned this recently. If one really wants to find the stocks having more potential to surge now, one has to try the "cigarbutt" with low PE, with stock price much lower than the NAV. Of course, at least you should know that they have stable track records in earning and a reasonable business model. Moreover, they are just between the annual result announcement and interim result announcement so as to get rid of the excited moment. Also remember to look at the trend to see if it is an "easy-going friend". On the other hand, remember that there is law of tub (in Chinese).