Long Term Secrets To Short-Term Trading

Figure 13.1 Bond trading system without money management. 

---

174 
What I know about math, you could add up on your thumb and first finger, but I know "math works" 
so I began trading commodities using the Kelly formula (see Figure 13. 1). Here it is with F representing 
the amount of your account you will back every trade with: 
F ((R + l)*P - I)/ R 
P Percentage Accuracy of the System Winning 
R Ratio of Winning Trade to Losing Trade 
System Report 
9/11/98 3:06:15 PM 
System Number: 387 
Description: bonds 7198 no bail 
System Rules: 
Market: 
Test Period: 1/1/90 to 7/16/98 
Base Unit Calculation Rules 
BASE UNITS = account balance/(draw down*2) 
If Account Balance Increases by: units last trade 
INCREASE units on the next trade by: 1 
If Account Balance Decreases by: units last trade 
DECREASE units on the next trade by: 1 
 
Figure 13.2 Varied results based on risk % ofaccount. 

---

175 
Let's look at an example using a system that is 65 percent accurate with wins 1.3 times the size of 
losses. The math is done as follows using P as .65 and R and 1.3: 
1.3 + 1*.65 - 1/1.3 
2.3* .65 = 1.495 - 1 = 0.495/1.3 = 38% of account used to trade. 
In this example, you would use 38 percent of your money behind every trade; if you had a $100,000 
account you would use $38,000 and divide that by margin to arrive at the number of contracts (see Figure 
13.2). If margin was $2,000, you would be trading 19 contracts. 
The Good, the Bad, and the Ugly of Money Management 
What this formula did for my trading results was phenomenal. In a very short time, I became a real-life 
legend, as very small amounts of money skyrocketed. Using a percentage of the money in the account, based 
on Kelly divided by margin, was my approach. It was so good that I was kicked out of one trading contest 
because the promoter could not believe the results were accomplished without cheating. To this day, people 
on the Internet claim I used two accounts, one for winning trades and one for losers! They seem to forget, or 
not know, that in addition to being highly illegal, all trades must have an account number on them before the 
trade is entered, so how could the broker, or myself, know which trade should have the winning account 
number on it? 
But, what would you expect, when no one to my knowledge, had turned in that type of performance 
ever before in the history of trading. To make matters "worse," I did it more than once. If it wasn't a fluke 
or luck, the losers lament is that it must have been done by pinching some numbers or trades along the way! 
What I was doing was revolutionary. And, as with any good revolution, some blood flowed in the 
streets. The blood of disbelief was that first the National Futures Association and then the Commodity 
Futures Trading Commission commandeered all my account records, looking for fraud! 
The CFTC went through the brokerage firm's records with a finetooth comb, then took all my records 
and kept them for over a year before giving them back. About a year after getting them back, guess what, 
they wanted them back again! Success kills. 
All this was due to market performance that was unheard of. One of the accounts I managed went from 
$60,000 to close to $500,000 in about 18 months using this new form of money management. Then the 
client sued me, her lawyer saying she should have made $54,000,000 instead of half a million. 

---

176 
Now my believers were willing to put me on a pedestal, if they could collect some money. The 
revolution was more than anyone could handle. 
What a story, huh? 
But there are two sides to the edge of this money management sword. 
My extraordinary performance attracted lots of money for me to manage. Lots of money, and then it 
began to happen: the other side of the sword flashed in the sun. Amidst trying now to be a business manager 
(i.e., running a money management firm) with precious few skills at doing something I am no good at 
anyway, my market system or approach hit the skids, with a cold streak that saw equally spectacular 
erosions of equity. Whereas I had been making money hand over fist, I was now losing money, hand over 
fist! 
Brokers and clients screamed, and most took the off-ramps, they simply could not handle this type of 
volatility in their account balances. My own account, which had started the first of the year at $10,000 (yes, 
that is $10,000.00) and reached $2,100,000 ... got hit along with everyone else's ... it too was caught in the 
whirlpool, spiraling down to $700,000. 
About then, everyone jumped ship but me. Hey, I am a commodity trader, I like roller coasters, is there 
another form of life? Not that I knew, so I stayed on, trading the account back to $ 1, 100,000 by the end of 
1987. What a year! 
Watching all this over my shoulder every day was Ralph Vince, when we were working together on 
systems and money management. Long before I could see it, he saw it, saw there was a fatal flaw in the 
Kelly formula. I was too blind; I kept trading it, while Ralph, math genius that he is, began intense research 
into money management, the culmination of which are three great books. His first was The Mathematics of 
Money Management, followed by Portfolio Management Formulas, and my favorite, The New Money 
Management. These are all published by John Wiley and Sons, New York, and are mustreads for any serious 
trader or money manager. 
Ralph noticed the error of Kelly, which is that it was originally formulated to assist in implementing the 
flow of electronic bits, then used for blackjack. The rub comes from the simple fact that blackjack is not 
commodity or stock trading. In blackjack, your potential loss on each wager is limited to the chips you put 
up, whereas your potential gain will always be the same in relationship to the chips bet. 
We speculators don't have such an easy life. The size of our wins and losses bounces all over the place. 
Sometimes we get big winners, sometimes miniscule ones. Our losses reflect the same pattern: they are 
random in size. Figure 13.3 shows a trade-by-trade recap on a system I use so you can see the irregularity of 
wins and losses. 

---

177 
As soon as Ralph realized this, he could explain the wild gyrations in my equity swings; they came 
about because we were using the wrong formula! This may seem pretty basic as we are about to enter a new 
century, but back then we were in the midst of a revolution in money management and this stuff was not 
easy to see. We were tracking and trading where, to the best of my knowledge no one had gone before. What 
we saw were some phenomenal trading results, so we did not want to wander too far from whatever it was 
we were doing. 
Ralph came up with an idea he calls Optimal F; it is similar to Kelly, but unlike Kelly can adapt to 
trading markets and gives you a fixed percentage of your account balance to bankroll all your trades. Let's 
look at what can happen with this general approach. 
On the End of a Limb and Sawing It Off 
The problem with an optimal F approach or fixed fraction of your account is 'that, once you get on a 
roll, you roll too fast. Let me prove my point; if your average win/loss trade is $200 and you have 10 trades 
per month and you will increase on contract at every $10,000 of profits, it will take you 50 trades or 5 
months to add that first additional contract. Then it will take only 2.5 months to go from 2 to 3; about 7 
weeks to boost it up to 4 contracts; 5 weeks to jump to 5; one month to reach 6; 25 days, to 7; 21 days, to 8 
contracts. Eighteen days later, you are at 9, and at 16.5 days, you trade a 10 lot. 
Then disaster strikes, as it surely must. You have now scooted out on the end of a limb and are sitting 
there with lots of contracts on. Although the limb snaps when you have a large losing trade (3 times the 
average of $200 or $600 per contract times the 10 lot so you just dropped $6,000), you have not given back 
$10,000 yet. So you trade a 10 lot on the next trade and lose another $6,000. Now in two trades, you are 
down $12,000 from your equity high at $100,000. 
The next trade is also a loser, three in a row, for the average of $200 times the 9 lot you are now 
trading and you get tagged for another $1,800 (let's call it $2,000). You are now down $14,000. 
Meanwhile, a "smarter" trader decreases faster than you, cutting back two contracts for every $5,000 
lost, so on the first hit he or she is back to 8 contracts, losing only $2,400, sidestepping another $6,000 hit, 
and on it goes. 
And It Can Get Worse by Far... 

---

178 
Let's take a winning system. It wins 55 percent of the time, and you decide to trade 25 percent of your 
bank roll, starting at $25,000 on each trade. Wins are equal to losses at $1,000 each. Table 13.1 shows the 
way the trades played out. 
You made $1,000 yet had a 65 percent drawdown while a single contract trader would have dipped 
$16,000 with a 20 percent drawdown! 
Let's look at another scenario where we hit it right from the get-go winning 5 of 8 trades (Table 13.2), 
a great deal, right?