1. Picked the pair of expiries that looked most extreme on a historical basis, perhaps not worth the extra uncertainty 2. I find warehouse stocks difficult in figuring out what's priced in and what's not - but implications of higher stocks being that HRW potentially less susceptible to weather scares? In any case, I suppose like any asset it boils down to figuring out what's priced in and what's not. Seems to me that very little weather premium is being priced in at the moment for this spread. 3. Would you be able to share your process using seasonals? How do you for example reconcile a strong seasonal pattern with 'stretched valuations' that should mean-revert, with the risk of offsetting the seasonality trend? What's the best way to screen for seasonality opportunities? How much of HRW discount to SRW do you think is driven by trade war (I'm assuming that China imports more HRW than SRW?) and how much is increased production in Ukraine/Russia of higher protein wheat?
1. Picked the pair of expiries that looked most extreme on a historical basis, perhaps not worth the extra uncertainty It was just rhetorical. In this case, I doubt it will change the outcome. But beware of "choosing the historical/backtesting best case with no fundamental justification. 2. I find warehouse stocks difficult in figuring out what's priced in and what's not - but implications of higher stocks being that HRW potentially less susceptible to weather scares? In any case, I suppose like any asset it boils down to figuring out what's priced in and what's not. Seems to me that very little weather premium is being priced in at the moment for this spread. You can check the crop progress by state if you want a clear picture of HRW( Kansas, Oklahoma...) vs SRW conditions right now ( Illinois, Indiana ). Check also if I am still right since the wheat states have changed recently. South Dakota is now the largest wheat producer in the US. 3. Would you be able to share your process using seasonals? How do you for example reconcile a strong seasonal pattern with 'stretched valuations' that should mean-revert, with the risk of offsetting the seasonality trend? What's the best way to screen for seasonality opportunities? If you have a good seasonal trend, but you are close to extreme valuations, just don't take the trade. It depends on the product in fact. Sometimes you can almost bet on an extreme valuation to get more extreme( hogs ), but on the other side, when you are short wheat spreads and getting close to full carry, you'd better liquidate it... It's just experience of the products. Try to have a fundamental reasoning before doing the yearly math stats. That's the way you get reliable trades. How much of HRW discount to SRW do you think is driven by trade war (I'm assuming that China imports more HRW than SRW?) and how much is increased production in Ukraine/Russia of higher protein wheat? Not much since I think China doesn't import much wheat from the US. Not sure. But Russia is becoming the price setter in wheat and they have different wheat qualities. By the way, french milling wheat is also close to HRW specifications.
https://www.reuters.com/article/us-...mily-farmers-to-leave-the-field-idUSKCN1S613U No Futures Contracts on Auction Houses
=> HRW Fob was above SRW Fob on my last check. I think most of the spread is explained by the very high spread between KW & HRW Fob. I have no idea why it's so costly to transport KW on a FOB Gulf. I think this trade is less good that it seems because it has a negative carry (if nothing change, back month expiry become front month expiry and your spread decrease).
Nice, I have never tried to use these datas. I would like to test if high spread OI increase or decrease the quality of my trades. HRW term structure is exceptional/dislocated so I am not surprised to have a lot of people in these trades.
05/17/19 : Long ZC K20/N20 @ -4.5. - The weather forecast is terrible for those who haven't been able to plant their corn( almost everybody in the eastern corn belt ) - K/N normally shows a smaller carry. - K/N didn't move with Z/K or very little to date.