Trading System - Burke's New Lows System

I found this post somewhere on the web. It caught my attention because I remembered that MIRAT was one of the most successful mechanical timing systems ever developed (it failed in 1999). In this post, the creator of MIRAT Michael Burke, discusses his New Low's market theory. You can be sure, that although MIRAT failed (which why we must always trade with stops) the concepts discussed here are valid and will work again.

My name is Michael Burk (not to be confused with Michael Burke of Investors Intelligence), in a previous life I was a photographer. Typically, I made some investments based on tips from friends and brokers and nearly always lost money.

As a photographer I spent a lot of time in the darkroom listening to cassette tapes of all kinds. One of the tape sets that I listened to was a 1980 investment conference at which Joseph Granville was a speaker. Granville was an exciting speaker and I subscribed to his newsletter, purchased his recommendations and lost money. I also studied his 1976 book Granville's New Strategy of Daily Stockmarket Timing for Maximum Profit. The book does a good job of pointing out the irrelevance of news and the value of technical indicators. The Granville book made a technician out of me and I began to develop technical trading systems.

After developing many systems that backtested well over relatively short periods, I decided that not to trade again until I had a system that always worked. That limitation slowed me down until I looked at new lows. New lows usually draw an accurate picture of the markets condition. When the market is falling, new lows build to a peak that occurs within a day or two of the price low. After the peak in new lows they (new lows) diminish rapidly. Prices may not rise immediately, but they will not fall significantly below the low established around the peak in new lows. Once you have identified a peak in new lows there is very little risk in the market. The prices will rise, new lows will decline and everything is good. Near a top, prices will continue to rise, but new lows will begin to increase. After usually 2-6 weeks of rising prices and increasing new lows prices begin to fall, new lows increase and the cycle plays itself out all over again. With new lows to identify the trend nearly every other indicator works much better.

I wrote several articles for Technical Analysis of Stocks and Commodities in 1989 detailing these concepts and offered simple programs demonstrating them. I then wrote a graphical program incorporating all of concepts. Several purchesers of that program told me if the system was really any good I should be able to write an objective implementation of it, saving them the trouble of learning all that stuff. In January of 1990 I wrote an objective implementation of the system and called it MORON to show my contempt for objective systems. To my dismay, most of the users liked MORON better than my graphical programs. MORON was incorporated into the graphical programs and renamed MIRAT for Mechanical Intermediate RAnge Timer and I made a living selling those programs as well as a market commentary based on that system until 2000. After 21 years of success (12 years backtested and 9 years real time) MIRAT failed in 1999.

In late 1999 I began writing programs that calculated the technical information (advances, declines, advancing volume, declining volume, new highs, new lows, momentum etc.) on the component issues of a segment, sector, index or fund. This system was intended as a sector selection method and I began trading it, with spectacular success, in January 2000. That euphoria lasted until the NASDAQ peak in March of 2000 with an encouraging period from June to mid July. Looking back on it, the system was bullishly skewed because of the bullish nature of the backtesting period. This underscored what I already knew, the condition of the broad market is more important than the condition of individual issues or sectors. Since closing my business (Tools for Timing) at the end of 2001 I have continued to work developing indicators and trading systems. The results of some of this work, you can see in my weekly e-letter.


Thank you Michael… I believe Michael still posts his excellent work on the Safehaven website. You can subscribe to his free newsletter there.