StrategyTrading1

I started creating trading systems in the “good old days”,  before personal computers were common. The first Apple computer was sold as a kit in 1976, and the IBM PC was introduced in 1981. In those days if you wanted to test a trading idea, or even calculate a simple moving average, you had to do it by hand. So, trading systems were not easy to test, at least for the average investor. Large companies had been crunching payroll numbers in back rooms for years, and I have no doubt that some of the programmers used these computers to crunch stock market numbers during their off-hours. 

However, most traders then, and  now, used both  fundamental analysis and what I call “personal analysis” to decide what and when to trade stocks. Before the PC era you could subscribe to newsletters and charting services that mailed you weekly or monthly books of stock charts. The idea was that once a week, or month, you would look through the charts and decide what you were going to trade. I subscribed to a number of these services and had piles of their soft cover booklets stored for reference in my home office. Any buy or sell ideas had to be tested by writing down the trade prices then computing the numbers by hand. Starting in the late 70’s a  few brave soles used the Apple II, and eventually the first IBM PC, to do this, using the primitive spreadsheet programs that became available over time. 

The bottom line was that “system traders” were out there, but coming up with any meaningful backtesting was not easy to accomplish with the tools available. 

I though it might be of interest to look at how far we have come in this roaring number-crunching era, by looking at how I started my quest for calculating help.

As a child my favorite Sci-Fi movie was The Day The Earth Stood Still (1951). The inside of the alien flying saucer contained a computer with twinkling lights and no paper. In the 1960’s my favorite TV program became Star Trek - again, with a computer that Captain Kirk spoke to, and of course, the Enterprise had no paper, and the ship’s computers had twinkling lights and beeping sounds. 

Now, that was a vision of the future!  However, the reality of the mid-1960’s was that the electronic calculator hadn’t been invented yet. Computers were huge and extremely expensive machines that were locked in back rooms of large companies. And, if you wanted to calculate anything you had to do it by hand, or use a slide rule, or a calculating machine of some sort. 

I had a number of slide rules - long ones, and circular ones. One day I decided to search for a calculating machine. I loved crunching numbers, and at the time my hobby was weather forecasting, and I wanted to calculate averages and moving averages. So, I found a used mechanical calculating machine, that looked like, and felt like a small cash register, with a pull handle to activate the gears. This was a heavy, clunky and fantastic addition to my life. I believe I paid about $100 for it, and it really worked. I later found an electronic version and retired the pull-handle clunker. 

Those machines used mechanical parts inside to spin number wheels and come up with answers. Huge progress. 

One day, in the late 60’s,  I heard about a company that was selling an electronic calculator, with lights to display the numbers, and no moving parts! I bought one for $50. It was a Bowmar, about the size of a hardbound book, plugged into the wall, and could add, subtract, divide and multiply, and was the beginning of something incredible. 

By the middle 70’s I was back at the University of California, and one day picked up a magazine that had an article about something new - a personal computer kit. I wanted one, but couldn’t figure out what I would do with it if I actually built the thing. Also, the electronic calculator era was in full swing, and Hewlett Packard had a nice collection of them for sale, so I decided to pass on the PC and I spent $300 for a calculator from HP. I put my slide rules away, and moved into a new age of calculating bliss. 

Then, around 1980, Casio, also a maker of calculators, came out with one that included the BASIC programming language. It stored programs on a cassette tape recorder with a special attachment. While in college I had learned FORTRAN and had run programs at the university computer center, typing in the data on punch cards. After graduation, I worked for a company in Berkeley in its computer department. They used a PDP-11/45 computer to calculate numbers that were fed into it with punch cards, so the leap to a hand-held programmable calculator was something I couldn’t pass up. I bought two Casios, and still have them, and I believe one still works. 

I wrote a number of programs to calculated gains and losses from my trading systems, and the Casio carried me forward until my first PC purchase. 

When the first IBM PC came out, it was very expensive. A simple setup could run from $2500-$5000. Fortunately, IBM later offered a PC Junior. It turned out that the Junior was, like the Edsel, a bad business idea. IBM put one CPU chip into it’s flagship PC and another one into the Junior. So, programs that would run on the PC wouldn’t run on the Junior. At least Junior was a lot cheaper than the PC, so I bought one for $1250. 

I converted my Casio programs to the Junior and was able to test my systems, until I eventually sold the Junior and bought a used IBM PC with the “real” Intel chip. The Junior used an Intel 8088 chip running at 4.77 MHz while the “real” PC used the 8086 chip. Today’s CPU chips run at 2-3Ghz - as in Giga, rather than Mega. 

Other personal computers were available at the time, including offerings from Atari, Tandy (Radio Shack) and Commodore. I just felt better going with the IBM as the name seemed to have a more serious business intent than the other offerings. 

And what about Apple? It was the name. How could a computer named Apple be a serious tool for business? For artsy purposes, an Apple seemed fine, but for stock market number crunching, I, and most other technical traders, had to have the IBM PC. 

Of course, there were still no programs that could backtest stock market trading systems. In order to do any testing or charting in the 1980’s you still had to get data into the PC. Various data services appeared and I bought a program from Metastock called The Downloader, that allowed me to log into  a stock data service, and download, via dialup phone line, stock prices at the end of the day. This was a huge leap forward from writing down the prices by hand on 4×4 to the inch graph paper, or entering the data by hand into a spreadsheet. I still miss the little tweets and log-in tones from those dialup years. 

Then, in 1989, I read about a new software product called System Writer. The advertising said that it used a new language called Easy Language, that was similar to BASIC, and that anyone could easily write and back test trading systems. Or, the program came with pre-canned systems. 

At first I stubbornly avoided the program. I felt I was doing fine writing programs on my own, and System Writer was quite expensive - over $1000. 

Eventually, Omega Research cut the price in half, and I purchased a copy of SW. It evolved into TradeStation. TradeStation later purchased an existing brokerage, thus making it possible for the software platform to place trades automatically, with no human intervention. 

I once remarked that I could become rich if only I could plug in one of my trading systems into a computer, and it would trade the system for me automatically, and I didn’t have to make the buy and sell decisions myself, and then have to call them in every day. Well, that era finally arrived and is with us today and is being offered by an expanding list of brokerages. Of course, I learned that getting rich isn’t easy, no matter what trading system or platform is used. 

Is it possible that market trading will eventually be dominated by computers making automated trade decisions - essentially a battle of the algorithms. To what extent is trading being computer driven today?

After the huge stock market crash of October 19, 1987, an article appeared in Barron’s saying that in 1929 brokers jumped out of windows, but during the 1987 crash it was the computers that jumped out of the windows. 

And, as I look back on my decades of using technology to make trading decisions, that image never leaves my mind.

And while today’s computers make it easier than even to follow, chart and analyze stocks price patters, they don’t guarantee trading success. A bad system, traded by a fast computer, will still lead to a bad result. And a great trading idea can still lead to a great result. And, all it takes is one great idea.