There is some need to explain how and why we have chosen to do an assessment of Manuel Blay's work on Dow Theory published on SeekingAlpha.com. Below is a brief summary of how we went from never commenting about Manuel Blay's work to "suddenly" bringing to light data that questions his work.
How It All Began
On January 7, 2013, Manuel Blay wrote an article titled "Dow Theory Special Issue: Assessing The Current Primary Bull Market Signal". In that article, Mr. Blay gave a breakdown of a recent call for a "primary trend bull market signal" according to the dictates of Dow Theory. In summary, Mr. Blay said the following:
"As with any timing device, the Dow Theory 'detects' the existence of a new bull or bear market with some lag. No timing system is able in 'real time' to spot the emergence of a new trend. However, as I have previously written in this Dow Theory blog, the Dow Theory does a good job at signaling new bull and bear markets in a timely fashion. As I wrote in my post 'Revisiting the 1987 crash', which you can find here, 'the Dow Theory tends to do a remarkable job at getting investors out of investments on a timely manner'. By the same token, the Dow Theory also excels at signaling new bull markets close enough to the bottom."
In the comment section of the same January 7, 2013 article, we left a response that addressed several issues on the Dow Theory analysis provided by Mr. Blay. We highlighted the following problems:
- the failure of short duration of primary trend changes
- conflicting calls of when signals were given
- Jack Schannep's performance using short duration Dow Theory analysis
Mr. Blay's response was a detailed article titled "Dow Theory Special Issue: An Answer To The New Low Observer (NLO)" dated January 12, 2013. From that point in time, we could not offer an acceptable rebuttal without appropriate data to support our initial claim that such frequent short-term changes in the primary trend require additional study as it is not consistent with Dow Theory and should result in negative performance results.
The Collection of Data and Findings
This brings us to our analysis in a July 13, 2016 article titled "This Can't Possibly Be Dow Theory" which took data from January 24, 2013 to the present. We felt it was only fair to take all market calls from the date that we questioned the "style" and duration of Dow Theory primary trend calls and examined the actual results based on the work of Manuel Blay.
The following is a reposting of the assessment of Mr. Blay's work (in chronological order) from January 7, 2013 to the present. The font that is in bold is a profitable transaction and the unbolded font is an unprofitable transaction, if a transaction were entered into based on the published date of the change in the Dow Theory Primary Trend.
- Bear Market: -31.54% for GDX
- Bear Market: -26.08% for SIL
- Bull Market: -24.91% for GDX
- Bull Market: -26.95% for SIL
- Bull Market: -5.24% for HAO
- Bear Market: +7.46% for DJT
- Bear Market : +6.21% for DIA
- Bear Market: +7.00 for SPY
- Bull Market: -9% for GDX
- Bull Market: -23% for SIL
- Bull Market: -11.30% for GLD
- Bull Market: -16.01% for SLV
- Bear Market: +8% for DIA
- Bear Market: +7% for SPY
- Bear Market: +8% for DJT
- Bear Market: +4% for HAO
- Bull Market: +2% for DIA
- Bull Market: -9% for GLD
- Bull Market: -11% for SLV
- Bear Market: +17% for GLD
- Bear Market:+10% for SLV
- Bear Market: +2.60% for DIA
- Bear Market: +52% for GDX
- Bear Market: +58% for SIL
- Bull Market: -4.39% for DIA
Thus far, the data (source citations found at end of July 13, 2016) matched our January 2013 thesis that short duration primary trend calls based on Dow Theory would result in greater than necessary negative performance results. Not only has there been overwhelming negative results from the Dow Theory analysis the frequency of the transactions alone result in unnecessary transaction costs that add to the negative performance.
It could be said that our assessment is selective, whereby we have chosen to ignore calls made prior to January 2013. However, we only used data from the date that we questioned the work. Another criticism is that the exact date of the change in the "primary trend" did not occur at the exact date of the published work or commentary. We stand corrected on the performance numbers if the published start and published end dates are incorrect. It could be said that we have an axe to grind with Mr. Blay. We don't.
So what is our point? Why bother someone else about their work on a topic that is a theory, at best? Isn't it possible to be wrong and still enjoy the process? Isn't is a low blow to tarnish the hard work of another?
The point, as we see it, is to advance the topic of Dow Theory. As we've said in the past, Dow Theory is often right about the market it is only the interpretation that is incorrect. We are not excluded from this dilemma as we have been wrong on many occasions in the past. However, when there is a clear learning opportunity, why should we stand on formality when the data is staring us in the face? At least, that is how we see it
Disclosure: I am/we are long GDX, GDXJ.