In this article, I continue discussing new directions in investment research.

See: Crowdsourcing investment research: opportunities in OSINT and Free information and the Efficient Market Hypothesis and Crowdsourcing investment research: Capital Market Taxonomy

There are two basic tasks in security analysis.

The first is gathering facts about the security.

With the facts in hand, the next job is to analyze the data critically and answer the question being asked.

Current information technology and the costs of data determine how the research should proceeds.

Note: This is a rather long article, describing changes in investment information over 100 years.

Facts: Complete and accurate

The most common way to judge the effectiveness of an analyst depends upon ex-post results relative to the analyst’s recommendation, compared to results of competing analysts.

Reasoning skills being equal, the analyst with the most complete and accurate information should have the best relative performance.

Analysts can not compensate for incomplete or inaccurate information — and it costs time and/or money to obtain quality input. As Bernard Baruch, the famous Wall Street speculator of the early 20th century put it:

“Every man has a right to his opinion, but no man has a right to be wrong in his facts.”
“If you get all the facts, your judgment can be right; if you don’t get all the facts, it can’t be right.”

The larger and more varied the investment market, relative to the analyst population, and to the level of technology applied to securities research, the higher the cost of gathering facts and the greater the chances that superior opportunities may be overlooked.

In other words, complex markets that have more information than can effectively be handled by the average analyst are inefficient. Inefficient markets have more opportunities waiting to be discovered than efficient markets.

Investment research in Ben Graham’s time

In 1934, when Benjamin Graham, the legendary security analyst and teacher of Warren Buffet, published the investment classic, Security Analysis, the US SEC had recently been created and the first rules on disclosure were being issued.

Most information about securities was not free and, compared to today, was difficult to obtain.

Registration statements and offering documents submitted to the SEC were available to the public only in SEC reading rooms in selected cities. To get this information, the analyst had to go to the SEC, find the document, request a photocopy, and pay copying costs per page.

Once you had the document in hand, financial analysis consisted of copying numbers to a paper spreadsheet, summing columns by hand or with a non-electric adding machine, and calculating ratios, also by hand, or perhaps by using a mechanical calculator driven by a little crank.

Stock prices were delivered on a telegraphic printer, called a ticker, on a long band of paper. To create a graph of stock prices it was necessary to plot prices and volumes, one by one, on graph paper. Calculations of present value, bond interest rates, or annuities were done by using little volumes of financial tables.

Everything was paper-based. Newspapers clippings were annotated, pasted on backing paper, and filed. Some subscribed to clipping services.

Not only was financial information hard to get at and work with, there was far, far less of it than today. Corporations were much simpler. There were no organized exchanges for financial derivatives. Products like asset-backed securities, index funds, and swaps were not available. Most markets were local, only 5% of the population invested in equities, and far fewer companies were listed than today.

Early providers of standard statistics

In 1860, Henry Varnum Poor published a book on the “History of Railroads and Canals in the United States”. In 1906, Standard Statistics Bureau was formed to publish previously unavailable data on U.S. corporations.

In 1910, the Standard Statistic Bureau service consisted of a 600-page bound 2-volume quarterly covering U.S. and Canadian stocks and bonds. The service was also provided in the form of index cards, with updates delivered periodically.

Standard Statistic Bureau advertised the dual convenience of having hard to get information gathered into one place, but also having this information indexed and filed — what served the purpose of a database in those days.

This early service proudly proclaimed that only facts, not opinions, would be delivered:

No advice on the market, no “tips” but every bit of authentic information that can be obtained from any source about any one or any number of securities in which you may be interested.

The data was not only hard to get (requiring trips to libraries, archives, and corporate headquarters and tedious hand-copying of data and payment of copying fees), but, in terms of the market at that time, quite complete.

Of course, there just wasn’t that much information available.

A typical Standard Statistics report (on Western Union, a leading stock at the time), shown on the right, had:

• Unaudited four year income statements with five items: Gross revenues; Operating expenses; Other income; Interest on bonds; and Dividends.
• Unaudited two year balance sheets with 22 items, divided between assets and liabilities.
• High and low stock prices of the last nine years.
• Dividends paid with dates.
• Directors’ names, addresses of the company, transfer agent, etc.

The entire 1911 Western Union “Standard Statistics” report could fit into a single item on a footnote to an auditor’s report on public companies in today’s market.

More »