Early in my career I made investment decisions on the basis of intensive fundamental research. I switched to quantitative techniques after about eight years.
By intensive fundamental research, I mean talking to a company’s management, talking to analysts of a company, reading all of the financial reports, researching the industry, etc. I found that such activities, on average, did not pay off as nicely as a quantitative approach.
Here are some of the problems I encountered:
Quality of management
The quality of management is critical to the future of a company. However, I found it difficult to make money in attempting to evaluate management quality.
First, if management is good then one should see evidence of their skill reflected in superior corporate financial performance. So, in one sense, one can rely on quantitative techniques alone to discern management quality.
Second, one should be wary of the idea that high quality management is worth something beyond the valuation justified by the company’s financial performance. Surely the quality of management will be reflected in superior corporate profitability. If one is not careful, one could “pay up” for superior profitability and then “pay up” again for quality management, when in reality the former is a function of the latter.
Third, it is difficult to determine whether credit for superior financial performance is due to management or to the business. A great business can show superior profitability in spite of mediocre management. A lousy business may show terrible profitability in spite of outstanding management.
In theory, fundamental investors should have an advantage over quantitative investors. Fundamentalists have access to everything that the quants have, but they can also attempt something that quants tend to avoid: forecasting what will happen.
In practice, the value of forecasting is at best uncertain. Its value compared to the cost of forecasting is poor for the following reasons.
First, forecasting in a way that contributes to investment returns is very difficult. You have to guess the future correctly, which is difficult. But you also have to get the timing right. It is not useful to say that something will happen at some point over the next ten years.
Second, forecasting is not enough. You have to come up with a forecast that is meaningfully different than what most investors expect. If your forecast is the same as everyone else’s then there is no reason to expect an opportunity.
Third, you have to anticipate the market’s reaction to your forecast. If you have a different forecast and you are right, all of your work will amount to nothing if other investors do not react to whatever happens.
So, earning a return by forecasting strikes us as very, very unlikely to succeed: You have to come up with a forecast that is different than everyone else; you have to get the timing right; and you have to predict how other investors will react.
The unstructured nature of fundamental stock research tends to allow greater bias in thinking. For example, this style of research allows one to find patterns in information that have little practical value in selecting stocks. Or, one may gravitate towards fundamental data that tends to confirm pre-existing beliefs instead of fostering new thinking.
WHERE FUNDAMENTAL RESEARCH MAY HELP
That’s not to say that all research beyond the quantitative model has no value. We think fundamental research can help in limited ways, such as:
- Verifying that the quantitative models are working with the right data; or
- Attempting to identify what, if anything, the quantitative models are overlooking. Such insights could form the basis for new quantitative research projects in the future.