Futures Journal

Thought I'd give a little insight on what a normal trading is like for me.

Just to give a little background, I'm trading a relatively well capitalized personal account. I use two systems -- a long-term system that follows 19 different futures and forward metal markets and a short-term system that trades US stock indices. These systems are entirely mechanical, with no discretionary decisionmaking involved.

The long-term system trades London Sugar, Coffee, Corn, Bund, 10 YR T-Note, Long Gilt, Dollar Index, EuroFX, Swiss Franc, Yen, NY Crude, Natural Gas, Heating Oil, London Aluminum Alloy, London Copper, London Nickel, KC Wheat, Nikkei 225 and Cotton. Average trade duration is 6 months.

The short-term system trades the big stock indec contracts -- S&P 500, S&P 400 Midcap, Nasdaq 100 and Russell 2000. Average trade duration is 3 days.

I developed the combination of these two systems using commercial software. Each has quite a nice positive expectation by itself -- approximately 2-3 MAR for the long term system and 5 MAR for the short term system. The two systems are completely noncorrelated so that when combined they produce a MAR of about 8. I run them relatively hot (1.75% risk for each) and backtesting reveals an average ROI of over 300% over the last 6 1/2 years.

The systems use end-of-day data. Therefore, my decision-making is done in the evening before the relevant trading day.

Last night I downloaded my data. After downloading, I ran it on my software to determine orders for my longterm system. There were no orders to be placed for today, which is not unusual at all -- most days there are no orders.

Then I ran the index system. It was short 3 markets from the day before after intraday reversals on stop orders -- long the SP but short the MD, and short the RL, and short the ND. The only order for today was to reverse the SP and go short if the SP moved more than a small amount down from the open.

After seeing what the order was to be, the next step was to run my position sizing algorithm to determine how many contracts to make the order for. My algorithm is based on market volatility, which is used as a proxy for actual risk. Last night the algorithm indicated that I should place the stop order to sell 4 Dec contracts -- if filled, this order would reverse my current position of Long 4 SP to Short 4 SP. I would also be quite nervous if the order filled, because it would have me short all 4 indices.

I sent that order off to my broker via e-mail. This morning, upon waking up, I saw an e-mail from my broker indicating that he had received my order and would place it.

Shortly after the markets opened at 9:30, I received another e-mail from him that the order had filled -- we actually had zero slippage, which is quite unusual in general, although not surprising for the SP.

The rest of the day was spent watching the markets with no working orders in place.

To begin the day, I was in a drawdown of about 1.5%. It appears that I made about a 2% equity increase for the day, perhaps returning to the old equity high (achieved on Friday) or perhaps just making a new equity high. Gains came in the indices for the most part, with the long-term markets relatively stable (small gains in wheat).

Doctor Nevo

 

Doctor Nevo when you speak of "MAR" what exactly is it that you are referring to?

TIA,
Commisso

 

It's a ratio of (Cumulative Average Rate of Increase in %) / (Maximum Drawdown % Experienced Over Testing Period).

Testing was done over 19960101 - present, which is a period that each contract I trade spans entirely, and also represents the period after which the SP began trading on the Globex.

Sample size is well over 2000 trades during that period.

$75 / trade in commissions and slippage for the long-term system, $300 for the short-term system.

Doctor Nevo

 

... I was wondering whether your index system is always in the market or is it out at times? Also, with this system do you enter with stops, MOO orders, or something else? Also, do you use Trading Recipes for testing?

Thanks much

 

Hi Bozwood,

Generally it is always in the market, although it does have infrequently triggered profittaking exits that are MOO. It reverses on stop orders.

I use Trading Recipes for testing and for trading.

Doctor Nevo

 

Doctor, I am working on developing a short-term index trading system. Could you put forth some advice, ideas or direction for me? Probably like yourself, i have ES tick data since contract inception.

 

What does *actual trading* reveal? How long have you been using this system and are the results so far in line w/ your backtesting?


thanks,
kosmo

 

Hi Doc,

You have just started to trade these systems?

Regarding your long term system (there's something you rarely hear about on ET), given the ready availability of long term EOD data, why did you choose to only test from 1996? If your average holding time for these long term trades is six months, are you sure you produced a large enough sample to draw valid conclusions from?

And just out of curiosity, does the long term system take positions in "far out" contracts, or do you take the nearest month and roll positions over?

 

Thanks for explaining Doc!

 

Hi Trader Kay,

The details of this short term index system are proprietary. However, I would point out that it's chief advantage is that it has zero correlation with the long term system. To achieve that it is countertrend.

If you do a good amount of research online, you should be able to learn interesting ideas about opening range breakout systems. Several vendors also sell systems that are reputed to be quite profitable.

Doctor Nevo