Here is the abstract of the Batechor and Ramyar (2004) from page 1.
Roy Batchelor and Richard Ramyar
Cass Business School, City of London
September 2006
Abstract
There is a widespread belief in financial markets that trends in prices are arrested at support and
resistance levels that are to some degree predictable from the past behaviour of the price series. Here
we examine whether ratios of the length and duration of successive price trends in the Dow Jones
Industrial Average cluster around round fractions or Fibonacci ratios. We identify turning points by
heuristics similar to those used in business cycle analysis, and test for clustering using a block
bootstrap procedure. A few significant ratios appear, but no more than would be expected by chance
given the large number of tests we conduct.
Keywords: Technical Analysis, Stock Market, Forecasting, Anchoring, Stationary
Bootstrap
JEL Classification: C15, C53, G10
Roy Batchelor, HSBC Professor of Banking and Finance, Sir John Cass Business School, City of
London, 106 Bunhill Row, London EC1Y 8TZ, United Kingdom. Telephone: +44 207 040 8733, fax
+44 207 040 8881 E-mail: r.a.batchelor@city.ac.uk
Richard Ramyar, PhD Researcher (2002-2006), Sir John Cass Business School, currently of Ramyar
Integrated Consulting. Telephone: +44 207 870 3185 E-mail: richard@richardramyar.com
The conclusion begins on page 31. This is the conclusion:
It is of course possible that our results are an artefact of the parameters of our testing
procedure. We have experimented with shorter (10 day) and longer (40 day) average
block lengths in our bootstrap, as against the base case of 20 days. We have also
conducted tests using narrower (.01) and broader (0.05) bands around the
hypothesized ratio values as against the base case value for å of .025. None of these
sensitivity tests undermine our basic, negative, result.
Our conclusion must be that there is no significant difference between the frequencies
with which price and time ratios occur in cycles in the Dow Jones Industrial Average,
and frequencies which we would expect to occur at random in such a time series. In
our introduction, we noted that empirical evidence from academic studies suggests
that not all of technical analysis can be dismissed prima facie. The evidence from this
paper suggests that the idea that round fractions and Fibonacci ratios occur in the Dow
can be dismissed.
On page 2 of this study is a reference to another study by Park and Irwin (2004). It is mentioned in the introduction to the study. See this text below:
1. INTRODUCTION
This paper tests a popular but previously untested proposition about the behaviour of
the stock market. The proposition is that when the market changes direction after a
period of trending prices, the magnitude and duration of the next trend is not random,
but depends on the magnitude and duration of the previous trend. Specifically we are
interested in whether the ratios of successive trends cluster around Fibonacci ratios or
âround numbersâ.
The idea that price trends may be arrested at predictable support and resistance levels
is one of many tools used by technical analysts. Technical analysis â the prediction of
turning points in financial markets by chart-based methods - has long been popular
among practitioners, but viewed with suspicion by academics. Burton Malkiel, in his
classic book writes, among many similarly cutting remarks - âTechnical strategies are
usually amusing, often comforting, but of no real valueâ (Malkiel, 1996, p161).
The root of the problem is the failure of technical analysts to specify their trading
rules and report trading results in a scientifically acceptable way. Too often, rules are
so vague or complex as to make replication impossible. Too often popular texts
contain dramatic examples of successful predictions of turning points, with no count
of misses or false alarms. Recently, however, academics have begun to look
systematically at some of the more easily replicable technical trading rules. Park and
Irwin (2004) provide a comprehensive review of these studies. Of 92 studies
published in the period 1988-2004, 58 reported positive excess profits from a
technical rule, 10 yielded mixed results, and 24 reported losses. Even allowing for a
bias towards publishing positive results, and the possibility that not all studies
3
properly accounted for transactions costs and risk, this does suggest that not all of
technical analysis can be dismissed prima facie.
to be continued......
