I have not been very good at posting book reviews recently — so much to do and so little time, as they say. I have just read The Heretics of Finance by Andrew W. Lo and Jasmina Hasanhodzic, a book of interviews with practitioners of technical analysis along the lines of Market Wizards. The comparison with the latter work should not be taken too far, however — Market Wizards is a book of conversations between a less-successful and various more-successful traders, while The Heretics of Finance is the result of academics putting a standardised set of questions to a number of people on the basis of their prominence in the field of technical analysis, not their track record.

The book contains a lot of the opinions of people who do not run money and who have a vested interest in supporting (and, indeed, believing in) technical analysis — strategists, analysts, newsletter-writers and the like. I read some of the interviews with them, but after a while started skimming through for the opinions of people who do run money. Happily, the introductions to the first thirteen chapters, of which each is a record of an interview with one person, make it reasonably clear who actually takes market risk and who does not. Of those who do, the ones most worth reading were Laszlo Birinyi, who was pretty sceptical about most technical analysis, and Linda Bradford Rashke.

Rashke was the most interesting to me, because some of what she said confirmed the way I am currently trying to do things (she does not use any fundamental analysis, but does not oppose it as a way of organising information). I trade on simple entry signals that work in trends, when I think there is a trend. Rashke says:

You can reduce everything down to the fact that you’re trying to enter into a trending market, in the direction of the trend.

Page 234 gives the most tantalising glimpse of the kinds of things she investigates:

For me, an indicator is anything I can quantify. I have to be able to run an indicator on multiple markets and multiple time frames and still see a similar profile… If you’ve been up for two days, what are the odds that the next day is going to be down… those are the types of things I look at.

This has made me think that I have more work to do on improving my entry signals. My system ought to work, but if Raschke investigates things like this, then perhaps they can add to what I am doing. It sounds like they are not so much indicators as ways of adding to the basic edge — which is entering into the trend. I have neglected things like the odds of one kind of day right after another kind because I have thought that what is generally true of a market may not be true around my trading signals, so it is not obvious that this kind of analysis would help me — but the book has made me think it is time to test that hypothesis by back-testing.

Finally, Rashke says this:

You can make a lot of money on trends, but most people don’t have the patience to stay on that type of time frame

I have not had the patience, and it is something I have been thinking about a lot recently. I did make a positive return in the past three months (Oct-Dec), but I would have made more if I had been more inclined to stay with my trades. I am not sure it is so much a matter of patience as of understanding that there is a tension between cutting losses quickly and avoiding exiting into weakness. I think I may have the balance between those things wrong when it comes to closing out winning trades.

Incidentally, I have recently read (actually, skimmed) The Evolution of Technical Analysis by the same authors. I thought it was of purely academic, historical interest, and of little use to a market practitioner. It starts with the ancient Babylonians — quite unnecessarily, considering that the bits of history that are likely to be of interest to the modern technical analyst, like early Japanese candlestick charts and Dow Theory, came along rather later, and most of the intellectual history of technical analysis has occurred in the past 100 years. I have always been frustrated by books that start in the ancient world and end in the present day, like several histories of economics, where again the useful bit is the past century or so (or Gombrich’s The Story of Art, which would be much more interesting if it were the same length but began around 1500, or any number of other similar examples). The final section, on academic approaches to technical analysis, contained a short description of academic research by Lo and others that suggests there may be limited predictive power to some of the traditional technical-analysis patterns — but so limited that it tended to reinforce my scepticism. It does report that Lo has found strong empirical evidence against the random-walk hypothesis, but then, that was not a hypothesis that had ever given me much concern.

Advertisements