Wednesday, April 25, 2007

Recap: Gasoline futures seasonal trade

The gasoline futures seasonal trade that I mentioned in a previous post and discussed in details in my premium content area reached its exit today. It has been profitable for at least 11th consecutive years: the profit this year is $4,321.80 per contract of RT.


Anonymous said...

Nice blog!

Have you explored the seasonability of agricultural commodities? Prices show very consistent seasonal patterns, expecially corn.

On another note, what brokerage firm do you recommend for someone wanting to trade stocks, futures and options? Unfortunetely, it seems that many firms such as Interactive Brokers, do not give access to the LME.

Ernie Chan said...

Dear anonymous,
Thank you for your interest! I personally only use Interactive Brokers for both stocks, options and futures. I know there is another broker
that has a good futures backtest platform, but I haven't tried them personally. Incidentally, this futures broker publishes a lot of the ideas that I did my research on.

Anonymous said...

What do you mean by IB PUBLISHes? Are the articles on site?

Ernie Chan said...

I didn't mean IB publishes -- I meant does.


Jock said...

Ernie, I very much appreciate the effort that you have put into your seasonal spreads. Well done.

Another energy spread that shows potential based on historical results is buying Sep Crude oil and selling Sep Natural Gas about now, with a view to exiting mid-July.

My question though is … when looking at the historical results of seasonal spreads and when initiating the trade do you feel that it’s more appropriate to trade a fixed volume or a fixed dollar value? For example should we test and trade by buying one CL contract and selling one NG contract or is it more appropriate to buy a fixed dollar value of CL & sell the same dollar value of NG each year.

I ask because even though this seasonal seems valid, these two markets certainly don’t always move in tandem. This can mean that buying/selling say $1,000,000 of CL and $1,000,000 of NG each year will result in quite a variance in the number of contracts from year to year. On the other hand just using 1 contract for each leg of the spread means that in some years we are giving a lot more dollar weighting to one of the two markets.

While a back-test will likely favor one method over another, I wonder what the logical rationale should be.

Ernie Chan said...

Dear Jock,
Thank you for suggesting this new spread for us to consider.

In general, I would recommend scaling the two sides of the spreads by their volatilities (or ATR in technical analysis parlance). This ensures that a move of 1 standard deviation in one instrument results in the same P&L change as an 1-standard-deviation move in the other instrument.

Hope this makes sense.

Anonymous said...


Your Crude/Nat Gas trade spread was a big winner. My prices were...

Nat Gas... short May 9 $8.01 cover mid July at $6.60 May 9 $64.88, sell mid July at 73.50

Nice profit. Congrats.