BarroMetrics Views: Area of Competence
The other day I was on Bloomberg’s ‘Asia Confidential with Bernie Low’. As is the normal pattern with this type of shows, viewers are invited to ask questions of the guest commentator. Here you take pot luck with the questions: the viewers will tend to ask questions from their own mindset and experience – a context that may well be outside one’s area of expertise.
Well I had one of those shows – all viewers who had questions based on technical analysis must have switched off for the day. I kept getting questions like: “I can’t buy your call that the S&P will rise to 1500 in 18 months to 2 years. We are de-leveraging; where is the leverage going to come from?”
My answer: “I wouldn’t have a clue. I don’t think in that manner”.
If it weren’t for the other guest, it would have been a long 15-minutes for me. At least he was able to give what sounded like well thought out answers.
The point I am making is the markets are like the shows. There are times when conditions suit one’s knowledge and area of competence; and at other times behave in a manner that leaves one flummoxed. Take the two attached charts, one of the Shanghai Index and the other of the S&P. Both had similar underlying conditions:
- The 13-week line up (quarterly trend) was statistically stretched and
- The 5-d had possible topping patterns.
But the Shanghai Index (with the less reliable pattern on my stats) gave a 600 point profit (ATR of mean + 19 stdev) before a rally warning; the S&P gave a maximum of 20 points (ATR mean +1 stdev); and to make the 20 points, you would have had to sell the high of the signal day and buy the low of the day before the rally warning. Clearly a massive difference between profit potential of the two signals.
In my early days as a trader, I’d spend much time looking for distinctions that would identify when a pattern would and when it would not work. While that is still a worthwhile endeavour (new robust distinctions are always useful), I spend more time looking for ways to manage my risk.
The fact is by being aware of when one should increase position size and when one should reduce will make a huge difference to the bottom line. And the way to identify those situations is by being aware of the metric patterns relevant to one’s situation. The best guides for newbies are the patterns derived from the Expectancy Reward formula:
(Avg$Win x WinRate) – (Avg$Loss x Loss Rate).
I found that generally the components will provide patterns to guide and advise us.