Right out of graduate school, I landed a job as a commodity research analyst. I was pleasantly
surprised to find that my economic and statistical analysis correctly predicted a number of major
commodity price moves. It was not long thereafter that the thought of trading came to mind. The
only problem was that my department generally did not permit analysts to trade. I discussed my
frustration over this situation with Michael Marcus (first interview), with whom I became friends
while interviewing for the research position he was vacating. Michael said, "You know, I had the
same problem when I worked there. You should do what I did—open an account at another firm."
He introduced me to a broker at his new firm, who was willing to open the account.
At the time, I was earning less than the department secretary, so I didn't exactly have much risk
capital. I had my brother open a $2,000 account for which I acted as an advisor. Since the account
had to be kept secret, I could not call in any orders from my desk. Every time I wanted to initiate or
liquidate a position, I had to take the elevator to the building's basement to use the public phone.
(Marcus' solution to the same problem is discussed in his interview.) The worst part of the situation
was not merely the delays in order entry, which were often nerve-wracking, but the fact that I had
to be very circumspect about the number of times I left my desk. Sometimes, I would decide to
delay an order until the following morning in order to avoid «rearing any suspicion.
I don't remember any specifics about my first few trades. All I recall is that, on balance, I did only a
little better than break even after paying commissions. Then came the first trade that made a lasting
impression. I had done a very detailed analysis of the cotton market throughout the entire post-
World War П period. I discovered that because of a variety of government support programs, only
two seasons since 1953 could truly be termed free markets [markets in which prices were
determined by supply and demand rather than the prevailing government program]. I correctly
concluded that only these two seasons could be used in forecasting prices. Unfortunately, I failed to
reach the more significant conclusion that existing data were insufficient to permit a meaningful
market analysis. Based on a comparison with these two seasons, I inferred that cotton prices, which
were then trading at 25 cents per pound, would move higher, but peak around 32-33 cents.
The initial part of the forecast proved correct as cotton prices edged higher over a period of months.
Then the advance accelerated and cotton jumped from 28 to 31 cents in a single week. This latest
rally was attributed to some news I considered rather unimportant. "Close enough to my projected
top," I thought, and I decided to go short. Thereafter, the market moved slightly higher and then
quickly broke back to the 29-cent level. This seemed perfectly natural to me, as I expected markets
to conform to my analysis. My profits and elation were short-lived, however, as cotton prices soon
rebounded to new highs and then moved unrelentingly higher: 32 cents, 33 cents, 34 cents, 35
cents. Finally, with my account equity wiped out, I was forced to liquidate the position. Not having
much money in those days may have been one of my luckiest breaks, since cotton eventually
soared to an incredible level of 99 cents—more than double the century's previous high price!
That trade knocked me out of the box for a while. Over the next few years, I again tried my hand at
trading a couple of times. In each instance, I started with not much more than $2,000 and
eventually wiped out because of a single large loss. My only consolation was that the amounts I lost
were relatively small.
Two things finally broke this pattern of failure. First, I met Steve Chronowitz. At the time, I was
the commodity research director at Hornblower & Weeks, and I hired Steve to fill a slot as the
department's precious metals analyst. Steve and I shared the same office, and we quickly became
good friends. In contrast to myself, a pure fundamental analyst, Steve' s approach to the markets
was strictly technical. (The fundamental analyst uses economic data to forecast prices, while the
technical analyst employs internal market data—such as price, volume, and sentiment—to project
prices.)
Until that time, I had viewed technical analysis with great skepticism. I tended to doubt that
anything as simple as chart reading could be of any value. Working closely with Steve, however, I
began to notice that his market calls were often right. Eventually, I became convinced that my
initial assessment of technical analysis was wrong. I realized that, at least for myself, fundamental
analysis alone was insufficient for successful trading;
