History Fun | Commodity Time Spreads, Inventory and Cancelled Warrants
This piece was penned years ago, highlighting the then-prevailing market structure in commodities. In the aftermath of crisis, big banks got into commodity warehousing and became a force of influence on commodity price and curve behaviour. Fundamentals matter, but a keen sense of market structure and incentive mechanisms is necessary to sidestep losses stemming from known unknowns, or plain old ignorance.
Contango (Futures price higher than Spot price) is a bank/warehouse owner’s best friend.
Market participants seem to believe that a spurt in LME Cancelled Warrants and a fall in Inventory is a sign of pickup in commodity demand. This fall in inventory is expected to lead to narrowing Time Spreads, as Spot prices jump. They expect the opposite when there is a fall in Cancelled Warrants and a rise in Inventory. This view may well represent a distorted interpretation of an illusory reality.
The line of causation may well be running in the other direction. To understand this linkage, we should see the world through the eyes of a commodity producer/physical trader that owns physical commodity.
Producer’s Incentive System
(1) When Contango is Steep | Deplete LME Inventory, Rise in Cancelled Warrants
A physical trader/producer that holds metal in storage could indulge in Cash-n-Carry arbitrage. Cost would comprise of: (1) Spot; already owns the metal, (2) Inventory financing (Financing cost) and, (3) Storage Cost. If prevailing futures price is higher than this net cost, one could short Futures and deliver the metal on expiry, locking in a gain for a known future date.
Profits could be realised if Time Spreads (difference between Futures and Spot price) narrowed. This could happen if Spot moved up; as this is the primary determinant of Curve structure. In the short-run, financing and storage costs are relatively inelastic. This outcome could be realised through a picture (real, or illusory) of depleting inventory.
One could move metal, which is currently residing in an LME registered warehouse (also called ‘on warrant’), ‘off warrant’. The metal could be moved to a non-LME registered warehouse, where storage costs were possibly lower. This would create a depletion in LME registered visible inventory. Concomitantly, this would cause a spike in Cancelled Warrants. A spurt in Cancelled Warrants would lead market participants to assume an increase in demand. Spot would be bid up, and Time Spreads would narrow.
(2) When Contango is Flat | Glut LME Inventory, Fall in Cancelled Warrants
It is now in the metal holder’s interest to have a steepening Futures Curve. This could happen if Spot prices moved lower, relative to Futures prices. This outcome could be realised through a picture (real, or illusory) of rising inventory. ‘Off warrant’ inventory could be moved back ‘on warrant’. There would be a spurt in LME registered visible Inventory. Concomitantly, Cancelled Warrants would fall. This would cause Cancelled Warrants to evolve on a falling trend, and Inventory on a rising trend. Market participants would be led to assume a ‘demand slowdown’ and of ‘adequate stocks’, causing Spot prices to fall and Time Spreads to widen.
In a happily alternating cycle, as Contango widened, one could play (1). As Contango flattened, one could play (2).
What about actual demand/supply?
It is pertinent to note that no reference has been made to the actual Demand/Supply equation. Independent of actual demand, the movement of metal in and out of LME inventories alone could cause Time Spreads to gyrate. Even though metal has been taken off warrant, they are sitting somewhere. They are just not being counted.
Visible + invisible inventory may be constant and yet Spot prices could move higher or lower, under the illusion of a ‘deficit’/’surplus’, or a ‘pick-up’/’slowdown in demand’!
Bank / warehouse owner’s incentive System | Contango is their best friend
Any movement into its warehouses would represent income for the bank/warehouse owner. Since banks may also responsible for financing inventory purchases of producers/physical traders, it is in the lender/warehouse owner’s monetary interest to initiate actions that causes metal to move into its own warehouses, and stay there.
The easiest way to achieve this is to expect outcomes that make (1) above more likely than (2).
Following the path of least resistance, a bank/warehouse owner is less likely to tolerate Backwardation (Futures price lower than Spot price) than Contango.
