Hi. I'm very new to futures(never made a trade, only stocks, forex and crypto), but after watching some videos with Mike Maloney I have become very interested in trading ratios between different commodities. Like for example sugar/wheat. Sugar has fallen vs wheat and if it falls a bit more it might be time to buy sugar with the hope of selling that when it has increased vs. wheat and then buying wheat. But this is a multi year strategy. Does anyone know if the roll cost and trading costs a few times per year would eliminate any profit if I have to hold a position for say 5 years? If I look at just the continuous sugar/wheat contracts it seems to be very easy to trade. Just buy when cheap and sell when expensive.

You have to backtest it to know for sure.

But realize if you are using a ratio of continuous backadjusted contracts, every time there is a roll all those past ratios will change. You would want to use non-adjusted contracts in calculating the ratio, which means there will likely be jumps in the ratio.

I think when you really look at, it is not going to be as easy (or profitable) to trade as you might think.

So why doesn't everybody trade them? Because it's not easy to do or to test in most retail algo software.

I guess no free lunch.

ICE Sugar and CME Wheat both have Jul-2022 contracts and if we calculate the ratio of those contracts from Friday night we get 0.0222 (12.03/543) while the spot May-2020 contract is .0194 (11.1/571.25). So if nothing changes in the next 2 years, the roll yield alone will raise your price to 0.0222.

As @kevinkdog said dealing with ratio's (or percentages, or returns, or any division or multiplication of the prices) doesn't work very well for continuous backadjusted contracts. Backadjusted contracts can also have prices go negative which also messes up a lot of trading testing software. Thankfully spreads are a little easier but still difficult.

Exchanges list many native spreads. So there is an actual May20/Jul20 spread for both ICE Sugar and for CME Wheat. There are even cross commodity spreads like Crude-Heating Oil or 10yr Treasuries vs 30yr Treasuries. The beauty of these spreads is that their bid-ask is normally tighter than trading an outright month, and that there is no legging risk in execution. The problem is some software will not allow you to trade this exchange spreads (for example Tradestation doesn't, but Interactive Brokers does) so in some cases you will have to execute orders in both the individual contracts which increases the cost from crossing additional bid/ask spreads and gives you leg risk. This becomes even more difficult when looking at something like SB vs ZW as they are on two different exchanges so you will have to leg the spread yourself. There is 'Autospreader' software that makes this very easy (Trading Technologies and CQG being the most popular) but they tend to be expensive for the retail trader. Also be careful when looking at charts of spreads or ratio's. These are normally created using the last prints for each contract and may imply prices that were never actually achievable.

Finally let me say that despite my negativity I think this is actually an interesting thought process to follow. Spreads behave very differently than outright contracts and can actually be easier to trade as they have less noise, they are also a lot cheaper from a margin perspective. So why doesn't everybody trade them? Because it's not easy to do or to test in most retail algo software.