Individual Investor Challenge: Can You Succeed by Going It Alone?

Includes: DIA, QQQ, SPY
by: Jeff Miller

We have a special concern for individual investors. This was at the heart of my reason for starting this blog. Many evenings I imagine and write directly to such a person.

With this in mind, let us imagine an investor who is unhappy with his/her manager, and is ready to embark on a self-directed program. The TV ads from brokerage companies all suggest that you can beat the performance of your former advisor, needing only a few consultations from their helpful telephone staff. They will help you make a lot of trades to implement your strategy.

The first step is to set up your trading account. Then you go online to get a lot of information about stock ideas and the economy, starting with the leading websites.

You need to have four attributes:

  1. Intelligence
  2. Knowledge
  3. Training
  4. Discipline

I assume that all readers are intelligent. I also assume that you are informed, but please keep in mind that most people are not. My favorite knowledge test is that more people can name the Three Stooges than can name the three branches of government. I cite many similar examples in this humorous post. But try this one.

James Fallows writes that 44% of Americans are crazy, since that is the percentage who believe that China is the top global economic power. I know that you would like to play poker against these people -- or take the other side of their stock trades -- but the least knowledgeable people do not constitute the market.

Training is different. Everyone understands that you need knowledge. When it comes to training, something really strange happens. People do not see the relevance.

Discipline is also a test. Most individual investors fail here as well. In a recent article, Dr. Brett Steenbarger writes about "weather(ing) setbacks without losing either self-control or self-confidence." In an article that everyone should read, he wisely observes the following:

You don't gain resilience by winning. Rather, you become resilient by losing--and by seeing that you can learn from (and overcome) those losses.

Two Examples of the Need for Training

Here are two recent examples (more to come in future articles) of analytical errors related to the lack of the analytical skill that comes from training.

Case one comes from a pundit critical of the recent retail sales report. Much of the analysis compares the official government data to other sources and notes the margin of error from the survey. So far, so good. But the author, seeking something more, writes as follows, first citing the report:

Special Notice – The advance estimates in this report are the first estimates from a new sample. The new sample for the Advance Monthly Retail Trade Survey is selected about once every two and a half years. For further information on the sample revision, see our website.

Now comes the analysis:

Did any pundit or guru actually read the Advance Retail Sales report? This first paragraph in the report warns that a ‘new sample’ technique has been employed.

Ergo, comparisons are futile at this point.

Next, the analysis is picked up by a widely-respected voice, that of Doug Kass.

Perhaps this "new sampling" methodology (and wide sampling error) and message by the U.S. Census Bureau helps to explain the difference between the strength reported late last week and the weakness in private and state tax data. (Hat tip to Bill King.)

Our take? I am perfectly willing to analyze and embrace data wherever it leads. I recently questioned the BLS employment data. But the criticism should be informed and accurate. This is not. The Census analysis does not change the "sampling methodology" or the "technique." It draws a new sample based upon new information. Many surveys, including the most popular private sources, use a different sample for each poll. It is not a change in method.

It is an analytical error to cite this as a change in method. It happens when analysts reach too far in trying to make a point. The original error is lost in the later citations and media coverage. There is a cottage industry based upon substituting anecdotal information for data.

How many readers have the skill to spot the mistake? Those who cannot get a distorted view of economic reality.

Case Two comes from the academic world, which can also be a source of misinformation. A news report tells us that companies reporting earnings on days when the sun is shining in New York have stronger reactions in their stock prices.

The study, which analyzed earnings announcements for all publicly traded companies from 1982 to 2004, found that companies that announce earnings during sunny days saw their stocks perform better than expected. This held true for companies that announced earnings that were below expectations, said John J. Shon, an assistant professor of accounting and taxation at Fordham University, and one of the authors of the study. “Those companies didn't do as badly as expected.”

While the difference in performance wasn't huge — around 50 basis points — it's enough for investors and companies to pay attention to, said Mr. Shon, who worked with Ping Zhou, vice president in the quantitative investment group of Neuberger Berman LLC, on the study.

Does anyone really believe this? I am going out on a limb here, but only because I have seen this so many times. I am willing to make a bet with the accounting profs. This will not work over the next year.

Why not? It is an obvious case of data mining, taking many variables and all of the old data and searching for a relationship. When you slice and dice the data and then look for your hypothesis, you can always invent an explanation. In this case, even the post-facto explanation does not seem credible. Nonetheless, it seems to have been accepted at face value. My Google search showed no criticism. And the article was published in an accounting journal. Wow!

Strong research starts with the hypothesis and then tests. I am out on a limb since I am speculating -- with some confidence -- that the researchers did not begin with the idea that sunshine in NY led to better stock returns for companies announcing earnings. This looks like one of many possible "causal" variables used in the research. Readers of this site might recall this great illustration of the problem.

Anyone who has experience in research methods spots this immediately. Those without the right training accept the findings at face value. There are many other similar conclusions to befuddle the average investor, most recently the September effect, which I discussed in timely fashion on September 1st.

Briefly put, I would not be buying stocks about to announce earnings based upon the NY weather report.


In the investment world, there is little respect for the training and education of others. There is a lot of attention to who has had the hot hand. This is a serious mistake.

If you needed to get your car fixed, you would look to an expert in car repair. Not so in research and economic analysis, where everyone is invited to substitute his own opinion.

Here we take a distinctly different perspective. We respect all sources of information, but we carefully assess what can be learned from each. That is the challenge for an investor seeking to "go it alone."