Compared to stocks, futures are specific not only for what they do represent (see a previous post about the 14 Differences Between Stocks & Futures) but also for the way they are built. In fact futures trade in a serie of contracts that are only active for a few months.
Futures Contract Rollover
Because of the limited contract life, a position in an expiring contract, if it has to be continued, must be rolled to the next nearby futures contract. For a long position, this means selling the expiring contract and buying the next one.
Most futures exchanges nowaday allow market participants to roll their futures positions from one quarterly futures contract month to the next at any time they choose. However, the trading floor convention is to roll the expiring quarterly futures contract month eight calendar days before the contract expires.This is known as the roll date.
There is usually a difference in price between the two contracts of a roll, referred to as the forward premium or discount (or as a spread) and this difference needs to be taken into account when calculating the unrealized profit or loss on a position or the setting of a protective stop order after the roll.
Each expiration month is represented by a letter.
Futures spread between nearby contracts
The spread between nearby contracts it is itself a whole world, both for analysts and traders. For those interested to dive into this fascinating topic, I recommend Joe Ross Trading Spreads and Seasonals.
For starters, the difference between two contract months of the same commodity ( ie., corn, sugar) represents the carrying charges or the cost of holding the commodity for a period of time. Carrying charges are determined by the cost of interest and storage when physical commodities are held in store.
Obviously, the spread between nearby and the distant months in the same commodity tells us about the relative strength or weakness of the market itself. For example; at the time I’m writing this post, July 2016 Sugar contract is trading at a substantial discount to July 2015 contract. This tells that traders (and hedgers) expect oversupply head.
Futures Continuation Contracts
As we have seen above, differently from stocks and FOREX – that trade in a continuous stream of price – futures are actively traded only for few months.
In order to backtest long-term trading systems and to perform technical analysis that requires more than few days of data, traders use a method of stringing futures contracts together to make a continuous series. They call such a series a continuous contract.
There are several methods with many additional variables to create continuous futures data series 1)B.Fulks, “Back-Adjusting Futures Contracts”, Trading Recipes DB, 20002)CSI, “Computed Contracts Get Real”, Online Journal, July 2001, http://www.csidata.com/techjournal/csinews/200107/page01.htm#1 3)Logical Information Machines (LIM), “Rollovers”, Digital publication, 2008. 4)J.Hyerczyk, “Continuous Data: Not so easy”, Futures magazine, July, 2008.. What differentiate these methods is how they stitch contracts in order to generate a continuous stream of prices.
Not adjusted continuous contracts
In a not adjusted continuous contract, when one contract ends, the next one starts without any attempt to adjust prices to make a smooth transition. So at every join, or every 3 months, there is a gap in the continuous contract.
Backward adjusted continuous contracts
Backward adjusted data uses the actual prices of the most recent contract with a backward correction of price discontinuities for earlier active delivery months. This method requires to recalculate all previous contracts anytime we add to the continuous series a new one contract. This is the continuous adjusted contract most used by active traders.
Forward adjusted continuous contracts
The Forward adjusted contract is similar to the backward adjusted one, with the difference that, since we correct the latest contract to stitch with the previous one, it eliminates the need to recalculate old historical price data. The disadvantage is that forward-adjusted contracts do not represent actual values for today’s markets.
This method is more a theoretical than a practical alternative for traders. It tries to overcome the fact that in an inflationary environment the first method (backward adjusted series) can produce negative past prices.
Proportionally adjusted continuous contracts
With this method, the gap between nearby contracts is offset by dividing the price of the new contract by the price of the old contract. As you can image, big spreads between nearby contracts produced wide price fluctuations in the historical data.
Gann adjusted continuous contracts
The Gann series method combines same delivery months contracts into one.
The Gann approach may be better than the nearest future variety because there are fewer discontinuities. On the other hand, the one-year segments of a “Gann contract” may be too long for practical use and trading.
I personally never used Gann contracts.
The nearest contract is the purest (short term) data serie available. Where possible we always like to have a clean plot of the nearest contract at hand.
For statistical researchers and cross market analysis, the Perpetual continuous contract is the most suitable 5)Continuous futures data series for back testing and technical analysis, http://www.ipedr.com/vol29/48-CEBMM2012-R00003.pdf
For real trading activities where the real values of the market prices are necessary, the back-adjusted one is the wildest used. In few cases (i.e. financial markets) also the continuous not adjusted contract can be used.
References [ + ]
|1.||↑||B.Fulks, “Back-Adjusting Futures Contracts”, Trading Recipes DB, 2000|
|2.||↑||CSI, “Computed Contracts Get Real”, Online Journal, July 2001, http://www.csidata.com/techjournal/csinews/200107/page01.htm#1|
|3.||↑||Logical Information Machines (LIM), “Rollovers”, Digital publication, 2008.|
|4.||↑||J.Hyerczyk, “Continuous Data: Not so easy”, Futures magazine, July, 2008.|
|5.||↑||Continuous futures data series for back testing and technical analysis, http://www.ipedr.com/vol29/48-CEBMM2012-R00003.pdf|