What Market Data to Use?

July 10, 2008

In collecting market data, a degree of discretion often arises, wherein analysts develop personal preferences.  Let me explain two of mine.

I collect currency rates several times each day, but I choose not to use closing New York rates to construct my data base as many others conventionally do.  Instead, the foreign exchange rates I select as the representative ones of the day for my data base are those gathered between 08:30 and 09:30 local time.  With both Eastern North America and Europe then in full swing, the market has considerably more depth than late in the afternoon when Europe is long closed, and chances for getting rogue rates are reduced.

In comparing different stock markets, I tend to use the Dow Jones Industrials for the U.S. market and have been questioned why that index instead of the S&P 500, which is a broader gauge and therefore viewed as a better benchmark by some people.  In my experience and for the purposes of capturing big trends, changes in the DJIA very closely approximate changes in the S&P.  Here are two examples to illustrate this.  While the crash of October 19, 1987 will always hold a place in stock market folklore — and a chapter in Chairman Greenspan’s memoirs — the long-term bull market of the late 20th century for practical purposes ran over 17 years from August 12, 1982 to early 2000.  Equities did not fall further after Black Monday, and the economy never reacted to the 1987 setback.  Bear markets in the DJIA and S&P 500 each ended on 08/12/82.  The bull market in the DJIA ended January 14, 2000 at 11,723, 16.85% per annum higher than its starting point of 777 in 12/82.  The bull market in the S&P ended on March 24, 2000 at 1,527, 16.58% per annum above its level on August 12, 1982 of 102.  A second illustration of how well the DJIA captures movement in the broader S&P can be found in their respective declines from October 2007 highs.  Through the close yesterday (July 9th), the DJIA had dropped 21.3% from 14,165 to 11,147, and the S&P had dropped 20.5% from 1,565 to 1,245.


One Response to “What Market Data to Use?”

  1. The Russell 2000 provides an interesting perspective on this issue.

    In March of 2000 the high of the Russell 2000 coincided with the S&P 500 and the Nasdaq Composite while the Dow had reached its high in January. However, by September of 2000 one could not have convincingly argued based on either the Dow, the S&P 500 that the market was trending down because at points in September of 2000 both of those indexes were within 5% of their all time highs. In contrast by September of 2000, the Nasdaq Composite had fallen by about 15% and the Russell 2000 had fallen by about 12%.

    The first index to make a new record high was the Russell 2000, in November of 2004. The Dow followed in mid 2006. The S&P 500 waited until October of this year. The Nasdaq never even remotely recovered.

    The Russell 2000 reached an all time high at the beginning of June, 2007 and touched that level again in July. When the Dow and S&P were at all time highs in October, the Russell was down about 10% from its high in June.

    Notwithstanding its large one day up moves from lowpoints, the Russell 2000 has been the short of choice in this market.