When Markets Overreact
The student of behavioral finance may profit

Admit it. You'd never dream of buying a stock whose price has recently soared by half even if the company's fundamentals — cash-flow-to-price or book-value-to-price — point toward more gains in the future. "Nine out of 10 people will not buy because they are too focused on regretting the missed opportunity of the 50% gain," says Patrick Ho, a senior portfolio manager at ABN-AMRO Asset Management in Hong Kong. Most investors also overreact to recent news, such as disappointing quarterly earnings results, and extrapolate trends based on insufficient data. All these are grist for Ho and Solange Rouschop, head of structured asset management, in picking stocks for the newly launched Behavioral Finance Japan Fund. The unit trust aims to maximize returns by exploiting irrational investor behavior, as quantified by the nascent science of behavioral finance. Ho and Rouschop spoke with Asiaweek's Cesar Bacani. 

What is behavioral finance?
Rouschop: It is an investing style that recognizes investors do not behave completely rationally when making decisions. They are influenced by psychological factors. We have identified four {significant} factors: image effect, overreaction to news, underweighting of fundamental information and overconfidence. We try to quantify these so we can gauge how they will affect the price of a stock.

Ho: The econometric model was developed by our offices in Amsterdam and Hong Kong in the last three years. After completing the model, we launched a behavior-finance based fund in Europe in 1999. That fund has been performing very well — it is about 30% up compared with the 20% rise registered by the benchmark. 

But how do you quantify emotions? 
Ho: We use measurable variables that we found are directly linked to behavior. To measure image effect, we examine a stock's fundamental ratios like cash-flow-to-price, book-value-to-price and price-per-earnings. We know that people pay a premium {above what you can expect based on fundamental ratios alone} to own famous names like Cheung Kong and HSBC. They have a certain level of comfort with these companies because they know they will not drop 50% in value in one day. But the image effect can temporarily disappear, as we have seen with the technology, media and telecom sector. A glamorous stock becomes undervalued {and would be picked out by the behavioral-finance model} when that premium disappears. 

To quantify overreaction, the variable is last month's stock-price change. If the model detects a sharp price movement, it picks out the stock as a potential victim of overreaction to recent news. Investors also tend to underweigh fundamental information such as a new trend. {When they realize what is happening}, the delayed reaction adds momentum to a stock's price, either upward or downward. We look at the price changes in the medium term {three months or more} to quantify underweighting.

Overconfidence is measured by the number of analysts that have revised their earnings estimates and recommendations for a stock. Investors tend to have excessive confidence in their own opinions. People seldom revise their raw estimates as fast as the analysts. So an upward revision in analyst recommendations is a good indicator of a later stock-price increase {as investors change their assessment}.

Are these four factors equally important?
Rouschop: We don't have a static model where we just plug in the numbers and the recommendation comes out. We also refine the weights to various factors based on different market circumstances. We can adjust for the fact that sometimes image is more important, but at another time, overreaction is more important. 

We look at the portfolio daily, but we try to strike a balance between capturing opportunities and avoiding high transaction costs. We have to make sure the opportunities are not cancelled by the costs. Typically, we deal on a monthly basis. We also adjust the sensitivity to the various factors every month. 

So what does behavioral finance tell you about the Japanese market?
Ho: Currently, we are overweight in pharmaceuticals and industrials. The top 10 holdings of our Japanese model portfolio are Takeda Chemical Industries, {semiconductor-maker} Rohm, {ceramic electronics company} Kyocera, Sin-Etsu Chemical, {factory-automation firm} Fanuc, Yamanouchi Pharmaceuticals, {electric-cable manufacturers} Sumitomo Electric and Furukawa Electric, Tokai Bank and Asahi Glass. Out of about 280 stocks in the Morgan Stanley Capital International Japan benchmark index, we choose about 50 to 60 stocks based on our behavioral finance formula. 

Don't you run the risk of affecting investor behavior when you yourself buy and sell?
Ho: If you have identified an opportunity and put in a big buy order, you can push up the price {in a not too liquid market}. One of our criteria {to enter a market} is liquidity because we have to get in and out of a stock via program trading.

Rouschop: You need to have a wide and deep market. We are looking at Taiwan {which has twice the turnover of Hong Kong}. I'm not sure Hong Kong is big enough to accommodate behavioral finance. We are looking at a regional fund {combining the smaller markets} for the rest of Asia.