@sdelsad/commodity-desk-daily 1.0.31 → 1.0.32

package/email.txt

@@ -0,0 +1,717 @@
+SOFT COMMODITY TRADING
+Episode 10 · Friday 21 August 2026 · 12 min 35
+
+Freight: Dry Bulk and Chartering
+Why the smaller ship can be the expensive one, and why freight is the one
+leg of a hedged trade that stays open.
+
+Listen:      https://cdn.jsdelivr.net/npm/@sdelsad/commodity-desk-daily@1.0.31/ep10.mp3
+Read online: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep10.html
+
+MARKET PULSE
+============
+
+Corn took the session, and the most interesting price on the screen was not
+a grain at all.
+
+Commodity         Contract    Price       Change
+------------------------------------------------
+Corn              Sep (CBOT)  478¾ c/bu     +5¾¢
+Corn              Dec (CBOT)  503½ c/bu     +5½¢
+Soybeans          Nov (CBOT)  1236½ c/bu     −¾¢
+Soybeans          Jan (CBOT)  1251½ c/bu     +¼¢
+Soymeal           Sep (CBOT)  —            −1.0%
+Soyoil            Sep (CBOT)  —            +2.0%
+Wheat SRW         Sep (CBOT)  682¾ c/bu     +2½¢
+Wheat HRW         Sep (KC)    762¼ c/bu      +¼¢
+Baltic Dry Index  —           2,791          +15
+
+Corn was the clear winner and December closed above five dollars. The buying
+is coming from the Pro Farmer scouts, who keep walking out of fields with
+less than last year. Illinois was pegged at 184.2 bu/ac against 199.6 a year
+ago and a three-year average of 199.2. Indiana, Nebraska, western Iowa,
+South Dakota and Ohio all came in below average. Illinois soybean pod counts
+at 1,430 per three-by-three square were under last year's 1,479 but above
+the three-year average of 1,390, which is the split that keeps beans flat
+while corn rallies.
+
+Beans were narrowly mixed and the complex went the other way from Wednesday:
+meal off about 1%, oil up about 2%. Weekly export sales were unremarkable at
+41 m bu of corn, 66 m bu of beans and 14.5 m bu of wheat. New-crop Chinese
+commitments now stand at 5.69 mmt with a further 3.82 mmt to unknown
+destinations — still a promise on a balance sheet rather than a fact on a
+vessel.
+
+The geopolitical read: the missing thing is a ship, not a tonne. Ukrainian
+strikes have left Russia unable to move grain through the Sea of Azov. The
+transmission most people reach for is export capacity, and that is right,
+but the mechanism underneath it is vessel class. Azov is a shallow river-sea
+trade worked by 3,000–5,000 t coasters drawing under five metres. No Panamax
+can substitute into that water. The cargo has to be railed or trucked
+several hundred kilometres to deepwater at Novorossiysk, which is carrying
+its own constraints. Capacity does not shrink because tonnage vanished. It
+shrinks because the only ships that fit are the ones that cannot sail.
+
+[chart] The smaller ship costs more — A Supramax is roughly a third smaller
+        than a Panamax and earned 1,687 dollars a day more. The classes are
+        not one ladder — they serve different trades, and US Gulf Supramax
+        routes have been firming while Asia-Pacific has gone quiet. — Baltic
+        Exchange sub-index average earnings, Wednesday 19 August 2026. —
+        https://storage.googleapis.com/podcast-audio-2647223968/commodity-
+        desk-daily/ep10_chart1.png
+
+  * Vessel class is chosen by draft, gear and lot size, and only then by
+    price. Read the berth before you read the index.
+
+  * Deadweight is not cargo. An 82,000 dwt Kamsarmax carries roughly 66,000
+    t of beans once fuel, water and stores are aboard.
+
+  * Baltic index points are not a price. They are a broker panel's route
+    assessments converted into a time charter equivalent in dollars per day,
+    and nobody can pay an index.
+
+  * Dollars per tonne and dollars per day are the same freight under two
+    different risk allocations. A voyage charter leaves the queue with the
+    owner; a time charter buys it back onto your book.
+
+  * A smaller vessel can be more expensive twice over — a higher day rate
+    spread across fewer tonnes. On today's rates that is 30% more per tonne
+    on the same cargo.
+
+  * The moment a CFR sale is made and no vessel is fixed, the seller is
+    short freight. The grain hedge does nothing about it, and no position
+    sheet carries freight at the resolution it carries corn.
+
+  * Twenty dollars a tonne of freight on a Panamax cargo is about 22,000
+    dollars a day of hire. That is an ordinary quarter in a market that has
+    better than doubled inside twelve months.
+
+  * FFAs hedge an index built from a basket of named routes, not your
+    voyage. It is the same cross-hedge problem as pricing Black Sea wheat
+    off Matif.
+
+  * Bunkers sit underneath the freight. A Panamax burns about 30 t a day at
+    sea, so a hundred-dollar move in fuel is seventeen percent of a sixteen-
+    dollar margin, decided in the oil market.
+
+  * Merchants charter rather than own because a ship is a twenty-five-year
+    position taken to solve a sixty-day problem, and because knowing where
+    the cargo is and knowing where the ship is are different businesses.
+
+Term                             Meaning
+----------------------------------------------------------------------------
+Deadweight (dwt)                 The total weight a vessel can carry — cargo
+                                 plus fuel, water, stores and crew — so
+                                 always more than the cargo it can load
+Draft                            The depth of hull below the waterline,
+                                 which rises as the ship loads and is the
+                                 hard physical limit on which berths and
+                                 rivers a vessel can enter
+Handysize                        The smallest mainstream dry bulk class,
+                                 roughly 10,000–40,000 dwt, geared and able
+                                 to work berths larger ships cannot reach
+Supramax                         A dry bulk vessel of roughly 50,000–60,000
+                                 dwt, normally carrying its own cranes,
+                                 which works minor bulks and shorter legs
+Panamax / Kamsarmax              The 75,000–82,000 dwt workhorse of the
+                                 grain and coal trades, usually gearless and
+                                 drawing about fourteen metres fully loaded
+Capesize                         A bulk carrier of about 180,000 dwt and up,
+                                 too large for the Panama Canal, used mainly
+                                 for iron ore and coal
+Geared vessel                    A ship carrying its own cranes, which can
+                                 therefore discharge at a berth with no
+                                 shore equipment
+Part cargo                       Loading a vessel below capacity because the
+                                 berth, river or canal cannot take her full
+                                 draft
+Baltic Dry Index (BDI)           The Baltic Exchange's headline dry bulk
+                                 freight index, a weighted composite of the
+                                 Capesize, Panamax, Supramax and Handysize
+                                 route assessments
+Time charter equivalent (TCE)    A voyage's economics restated as dollars
+                                 per day, which is how a shipowner compares
+                                 one employment against another
+Voyage charter                   Hiring a vessel to move a stated cargo
+                                 between named ports for a price in dollars
+                                 per tonne, with the owner carrying the
+                                 voyage and delay risk
+Time charter                     Hiring the vessel itself for a period at a
+                                 price in dollars per day, with the
+                                 charterer taking speed, weather, port delay
+                                 and usually fuel
+Fixing                           Agreeing the charter of a specific vessel,
+                                 the moment a freight exposure stops being
+                                 open
+Forward freight agreement (FFA)  A cash-settled swap on a Baltic index route
+                                 or basket over a calendar month, the only
+                                 liquid way to hedge freight
+Bunkers                          The vessel's fuel, priced separately from
+                                 the hire and carried by the owner on a
+                                 voyage charter and by the charterer on a
+                                 time charter
+P7 / P8                          Baltic Panamax route codes for US Gulf to
+                                 Qingdao and Santos to Qingdao, the two
+                                 assessments that set the soybean origin arb
+River-sea vessel                 A small shallow-draft ship built to work
+                                 both inland waterways and short sea legs,
+                                 the only class able to load in the Sea of
+                                 Azov
+Demand-to-supply ratio           The Baltic's measure of tonne-mile demand
+                                 growth against fleet growth, above 1.0 when
+                                 cargo is outrunning ships
+
+
+CONVERSION DRILL 10 OF 12 — DOLLARS ↔ EUROS PER TONNE
+=====================================================
+
+Rule: divide by the EUR/USD rate to go from $/t to €/t
+
+Fast method: at 1.10, $ → €: take off 9% (÷1.1 ≈ ×0.91). € → $: add 10%. At
+1.05, it is 5% each way; at 1.20, take off 17% / add 20%.
+
+  * $239/t wheat at 1.10 → 239 − 21 = €218/t
+
+  * €205/t Matif at 1.10 → 205 + 20 = $225/t
+
+  * $350/t at 1.05 → €333/t
+
+Why it matters: Matif quotes euros per tonne, Chicago dollars per bushel.
+Any Matif–CBOT arb crosses both a unit and a currency.
+
+
+QUIZ
+====
+
+Q1. A trader has 66,000 t of soybeans to move from Santos to Qingdao and can
+use either a Kamsarmax at 18,964 $/day or two Supramaxes at 20,651 $/day
+carrying 33,000 t each. Assume a 60-day round voyage for either class.
+Compute the freight cost per tonne both ways and state the difference in
+total dollars. Then give the two physical conditions under which the more
+expensive option is the only option, and explain why the day-rate comparison
+alone is the wrong first question.
+
+Q2. A desk buys 66,000 t FOB Santos and sells the same cargo CFR Qingdao at
+a gross spread of $74.00/t. Freight is expected at $52.00/t and other costs
+run $6.00/t. The board is fully hedged and both differentials are fixed in
+writing. The vessel is not yet fixed. Compute the expected margin in dollars
+per tonne and in total. Then recompute after freight rallies $20/t. Express
+that $20 as an equivalent change in daily hire, say which of the desk's
+risks was actually open, and name the instrument that would have covered it
+and the residual risk that instrument leaves behind.
+
+Q3. Baltic assessments this week put US Gulf–Qingdao at $73.81/t and
+Santos–Qingdao at $51.98/t. A Chinese crusher is indifferent between origins
+on delivered cost. US Gulf beans are offered FOB at a differential 15 c/bu
+below Santos. Convert that differential to dollars per tonne, combine it
+with the freight spread, and say which origin wins and by how much. Then
+explain what has to happen to the freight spread — not to the differentials
+— before the US Gulf becomes competitive, and why a Brazilian exporter
+watches the Panama Canal draft restrictions more closely than he watches
+CBOT.
+
+Q4. (Ep 9) Ep 9 established that a mandate creates demand that does not
+respond to price. Indonesia funds its B50 subsidy from a palm export levy it
+raised from 10% to 12.5%. Explain why a rising levy rate is a bearish signal
+for the programme's durability even though it is a bullish signal for palm
+prices today. Then say what a freight desk would want to know about
+Indonesian export volumes before pricing a Panamax to Rotterdam six months
+forward.
+
+Q5. (Ep 9) On Thursday soymeal fell about 1% while soyoil rose about 2%.
+Using ep 9's point that oil demand and meal demand are joined at the bushel,
+explain what that split does to the board crush and to the oil share, and
+say which of the two products a crusher would rather see lead a rally if the
+plant is running at capacity.
+
+Q6. (Ep 7) Ep 7 established that ending stocks is a residual roughly a tenth
+the size of production, so a 1% crop error is a 10% carryout error. Pro
+Farmer has Illinois corn at 184.2 bu/ac against USDA's national 180.7.
+Explain why a state number above the national number is not evidence that
+USDA is too low, and set out the two things you would need before letting a
+tour result move your own balance sheet.
+
+Q7. (Ep 7) A desk's own corn sheet uses 88.6 m harvested acres and its own
+yield. It wants to test how much of a yield miss the export line could
+absorb before the carryout falls below 1.400 bn bu. Starting from a carryout
+of 1.653 bn, compute how many bushels of cushion exist, convert that into
+bushels per acre of yield, and then explain why using exports as the
+offsetting line is a different kind of assumption from using feed and
+residual.
+
+Q8 — Conversion drill. A Rouen exporter is offered Matif November milling
+wheat at €218/t. A competing cargo is quoted CFR Rotterdam at $268/t. Assume
+EUR/USD at 1.16. Put both on the same currency and say which is cheaper, and
+by how much in euros per tonne. Then, if freight from the Black Sea to
+Rotterdam is $18/t, compute the implied FOB Black Sea value of that cargo in
+both dollars and euros per tonne.
+
+
+============================================================================
+                       SOLUTIONS BELOW — ANSWER FIRST                       
+============================================================================
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+SOLUTIONS
+=========
+
+A1. Take the hire first, then divide by what actually loads.
+
+                    Kamsarmax  Two Supramaxes
+---------------------------------------------
+Cargo                66,000 t    2 × 33,000 t
+Day rate              $18,964    $20,651 each
+Days                       60              60
+Total hire         $1,137,840      $2,478,120
+Freight per tonne      $17.24          $37.55
+
+The two-ship answer is roughly $20.31/t worse, or $1,340,280 on the cargo.
+That is the compounding penalty: a higher rate spread over half the tonnes
+each, twice.
+
+The realistic version is a single 55,000 t Supramax at $22.53/t, $5.29/t
+worse than the Kamsarmax — still about 31% more freight for the same job.
+
+The two conditions that force the smaller ship are draft and gear. If the
+discharge berth carries nine metres of water, a loaded Kamsarmax at fourteen
+metres cannot enter, and part-cargoing her destroys the per-tonne advantage
+anyway. If the berth has no shore cranes, only a geared vessel can
+discharge, and Panamaxes are usually gearless. A third, softer condition is
+lot size: a buyer who wants 55,000 t is not going to take 66,000 t to help
+your freight.
+
+The trap is the day-rate comparison. Freight per tonne has a numerator and a
+denominator, and the denominator is set by physics — hydrography and crane
+fitting — not by the market. The first question is what the berth can take.
+Only then does the rate matter.
+
+A2. The margin, twice.
+
+                                   As expected  After the rally
+---------------------------------------------------------------
+Gross spread (CFR sale − FOB buy)     $74.00/t         $74.00/t
+Freight                              −$52.00/t        −$72.00/t
+Finance, insurance, port              −$6.00/t         −$6.00/t
+Margin                                $16.00/t         −$4.00/t
+On 66,000 t                         $1,056,000        −$264,000
+
+A swing of $1,320,000 with the board never moving.
+
+As daily hire: $20/t × 66,000 t = $1,320,000, spread over 60 days, is
+$22,000/day. Hire would have to go from about $18,964 to roughly $41,000 —
+near today's Capesize level. The Baltic Dry Index has ranged from 1,261 to
+2,845 over the past 52 weeks, better than a double, so that is not an exotic
+scenario. It is a bad quarter.
+
+The open risk was freight, and it was open because the desk sold CFR.
+Selling CFR means promising delivered cargo at a fixed price while still
+having to buy the ocean leg. That is structurally short freight. It hid
+because a position sheet has columns for bushels and lots and none for
+tonne-miles.
+
+The instrument is an FFA — a cash-settled swap on the Baltic Panamax index,
+sized to the freight exposure, bought (long) to cover a short-freight
+position. The residual is basis risk in a new suit. The FFA settles against
+the P5TC basket average over a calendar month. The exposure is one voyage,
+on one route, fixed on one day. The Santos route can move against the
+basket, the fixing date can miss the settlement window, and the tonnage
+mismatch is never exact. It is the Black Sea wheat problem from ep 5,
+wearing shipping clothes.
+
+A3. Convert the differential first. Soybeans run 36.7439 bu/t, so 15 c/bu ×
+36.7439 ÷ 100 = $5.51/t.
+
+                       US Gulf     Santos
+-----------------------------------------
+FOB differential      −$5.51/t  reference
+Freight to Qingdao    $73.81/t   $51.98/t
+Delivered, relative  +$68.30/t  +$51.98/t
+
+Santos wins by $16.32/t. The freight spread is $21.83/t and the FOB discount
+recovers only a quarter of it.
+
+For the US Gulf to compete, the freight spread has to close by more than
+$16.32/t on its own — the differential is not going to do it, because 15
+c/bu is already a serious concession and $16.32/t is another 44 c/bu on top.
+Freight spreads move for structural reasons: Panama Canal transit
+availability and draft restrictions, Capesize demand pulling Panamax tonnage
+out of the Atlantic, US Gulf river levels forcing part cargoes, and the
+seasonal ballast pattern that leaves ships positioned in the wrong ocean.
+
+The Brazilian exporter watches canal draft because it is the cheapest lever
+on his competitor's cost. A draft restriction at Gatún adds days and dollars
+to every US Gulf–Asia voyage, or pushes it around the Cape of Good Hope, and
+every dollar of that lands on the delivered price of the beans he is
+competing with. CBOT moves both origins together and nets out of the arb.
+The canal moves only one of them.
+
+A4. The two signals point in opposite directions because they are about
+different horizons.
+
+Today the levy is bullish palm: a higher export tax raises the FOB price a
+foreign buyer must pay and withdraws exportable supply, which is a supply
+shock decided in a ministry rather than in a plantation.
+
+For the programme it is bearish, because the levy is the funding mechanism
+and the mandate's purpose is to shrink the export base the levy is collected
+on. A subsidy funded by a tax on the thing it is designed to reduce has to
+raise the rate as it succeeds. The rate is therefore the honest gauge of
+affordability. A rising rate says the per-tonne burden on a shrinking export
+volume is climbing, which is exactly the path toward a mandate that gets
+quietly deferred.
+
+A freight desk pricing a Panamax to Rotterdam six months forward wants
+export volume, not price. Freight is paid on tonne-miles. If B50 plus the
+levy takes several million tonnes a year out of the Indonesian export
+programme, that is Panamax and Supramax cargoes that stop existing on the
+Indonesia–Europe and Indonesia–India legs, and tonnage that has to
+reposition or accept lower rates. The palm price can rise while the freight
+demand it generates falls. Those are not the same trade.
+
+A5. Oil up 2% and meal down 1% moves the oil share up and does very little
+for the board crush, which is a weighted difference rather than a level.
+
+Meal is the larger revenue line for a US crusher — a bushel yields roughly
+44 lb of meal and 11 lb of oil — so a 1% loss on meal is close to offsetting
+a 2% gain on oil in absolute dollars. Beans were roughly unchanged, so the
+crush was roughly unchanged: a flat day disguised as a two-way move.
+
+If the plant is running at capacity, the crusher would rather see meal lead.
+Meal is where the tonnage is, and meal has to be sold locally into a feed
+market that cannot be arbitraged across an ocean cheaply. An oil-led rally
+is the one the crusher captures least of, because meeting the oil demand
+forces out four pounds of meal for every extra pound of oil, and that meal
+has to find a home at whatever price clears it. This is the joined-at-the-
+bushel constraint: at capacity there is no volume response available, so the
+only thing that helps is the price of the product you are already making the
+most of.
+
+A6. A state yield above a national yield is not evidence of anything,
+because they are different populations. The national number is an area-
+weighted average across all producing states, and it includes the ones the
+tour has just called below average — Nebraska, South Dakota, Ohio, Indiana,
+western Iowa. Illinois is usually one of the strongest corn states in the
+country. Its yield sits above the national average in almost every year, so
+184.2 against a national 180.7 says nothing until it is compared with
+Illinois's own history, which is 199.6 last year and 199.2 on a three-year
+average. On that basis it is a very large miss, not a beat.
+
+Two things are needed before a tour result moves a sheet. First, the
+methodology gap: Pro Farmer counts ears and measures grain length in a
+sample of fields on a fixed route in mid-August, then applies a fixed
+kernel-weight factor. USDA's August number comes from a farmer survey plus
+objective plot counts, and it will be revised as kernel weight is actually
+measured in September and October. The tour is a good early read on ear
+counts and a poor read on final kernel weight, and kernel weight is where
+the last ten bushels live. Second, the weighting: a yield is worth nothing
+without harvested area behind it. Moving Illinois down fifteen bushels
+changes the national number by roughly the ratio of Illinois harvested acres
+to the US total, so the arithmetic has to be done state by state and
+weighted, not intuited from a headline.
+
+A7. The cushion first.
+
+                      Bushels
+-----------------------------
+Carryout, USDA  1,653,000,000
+Target floor    1,400,000,000
+Cushion           253,000,000
+
+Across 88.6 m harvested acres that is 253 ÷ 88.6 = 2.86 bu/ac. A yield of
+180.7 falling to about 177.8 takes the carryout to 1.400 bn, all else equal.
+
+Using exports as the offsetting line is a behavioural assumption. Exports
+are a competitive outcome: they depend on what Brazil has, on freight, on
+FX, on Chinese buying policy. If the crop shrinks and the price rises,
+exports should fall, so the two lines are genuinely linked and the offset
+has an economic story behind it. But it is a forecast about other people's
+decisions, and it can be wrong in either direction for a whole season.
+
+Using feed and residual is a different animal, because that line is not
+measured at all. It is backed out of the quarterly Grain Stocks survey, so
+it absorbs both real livestock feeding and every measurement error in
+production and in stocks. Flexing it is not a demand forecast — it is an
+admission that the sheet does not balance and a decision about where to put
+the difference. Both are legitimate. But one is a view on the world and the
+other is a view on your own arithmetic, and confusing them is how a desk
+convinces itself it has an edge when it only has a rounding error.
+
+A8 — Conversion drill. Cross the currency first.
+
+                          Quote  Conversion  Common currency
+------------------------------------------------------------
+Matif Nov milling wheat  €218/t  × 1.16            $252.88/t
+CFR Rotterdam cargo      $268/t  ÷ 1.16            €231.03/t
+
+The Matif value is the cheaper of the two, by $15.12/t, or €13.03/t.
+
+Now back the freight out of the imported cargo. CFR Rotterdam $268/t less
+$18/t of freight leaves an implied FOB Black Sea of $250.00/t, which at 1.16
+is €215.52/t.
+
+Which is the useful part of the exercise: the imported cargo looked $15
+expensive on a delivered basis, but its FOB value sits €2.48/t below the
+Matif quote. The freight was carrying the entire difference and then some.
+Any Matif-versus-origin comparison that skips the freight leg is comparing
+two prices that were never quoted on the same terms.
+
+
+THE EPISODE, IN WRITING
+=======================
+
+Yesterday a Supramax bulk carrier earned $20,651 a day. A Panamax, about a
+third bigger, earned $18,964. The smaller ship was the more expensive ship,
+and it was the more expensive ship per tonne by a much wider margin than
+that gap suggests.
+
+That inversion is a good place to start, because it breaks the intuition
+that the dry bulk fleet is a ladder with cheap small ships at the bottom and
+expensive large ones at the top. It is not a ladder. It is four separate
+markets that happen to be measured on the same page.
+
+The fleet, and why the sizes are not arbitrary
+----------------------------------------------
+
+The classes, in deadweight tonnes:
+
+Class               Deadweight          Typical cargo       Gear
+----------------------------------------------------------------------------
+Handysize           10,000–40,000 dwt   Minor bulks, short  Geared
+                                        legs
+Supramax            50,000–60,000 dwt   Minor bulks, grain  Geared
+                                        part cargoes
+Panamax /           75,000–82,000 dwt   Grain, coal         Usually gearless
+Kamsarmax
+Capesize            180,000 dwt and up  Iron ore, coal      Gearless
+
+Deadweight is not cargo. It is everything the ship can carry: cargo,
+bunkers, fresh water, stores, crew and their effects. An 82,000 dwt
+Kamsarmax sailing Santos to Qingdao is carrying something like 3,000 tonnes
+of fuel before a single bean goes in the hold. She loads roughly 66,000 t.
+Treating dwt as cargo capacity overstates a cargo by a fifth, and that error
+propagates straight into a freight-per-tonne number.
+
+The classes exist because of draft — the depth of hull below the waterline,
+which increases as the ship loads. A fully laden Panamax draws about
+fourteen metres. A very large number of the world's berths, rivers and
+approach channels cannot take fourteen metres. When they cannot, there are
+exactly two options: load the big ship partly full, which is called a part
+cargo and throws away the scale advantage, or take a smaller ship.
+
+The second constraint is gear. A Supramax normally carries its own cranes. A
+Panamax normally does not. If the discharge berth has no shore equipment,
+the vessel class has already been decided, and no freight rate changes that.
+
+This is why the Sea of Azov story is a vessel-class story rather than a
+tonnage story. Azov is shallow water worked by 3,000–5,000 t river-sea
+vessels drawing under five metres. When that trade is shut, the world's
+spare Panamax tonnage is entirely irrelevant to it. The ships that fit
+cannot sail, and the ships that can sail do not fit.
+
+The screens: what a Baltic index actually is
+--------------------------------------------
+
+The Baltic Exchange publishes a headline index, the BDI, and a sub-index for
+each class. On Thursday the BDI printed 2,791, up 15.
+
+Here is the unit moment that matters most on this subject: those points are
+not a price, and nobody can pay them. The Baltic surveys a panel of
+shipbrokers each day on a fixed basket of named routes, then converts the
+assessments into a time charter equivalent — TCE, in dollars per day. That
+is the number a shipowner thinks in, because it is what lets him compare a
+grain voyage against a coal voyage against a period fix.
+
+Index             Points  Average earnings
+------------------------------------------
+Capesize (BCI)     4,376       $39,684/day
+Panamax (BPI)      2,107       $18,964/day
+Supramax (BSI)     1,634       $20,651/day
+Handysize (BHSI)     867       $15,605/day
+
+Read the middle two again. The Supramax is the smaller vessel and it is
+earning more per day. They are different trades: Panamaxes live on coal and
+large grain lots, Supramaxes on minor bulks and shorter legs, and US Gulf
+Supramax routes have been firming recently while Asia-Pacific has gone
+quiet.
+
+[chart] Same cargo, two vessel classes — A 60-day round voyage at
+        Wednesday's rates. The Supramax costs 31% more per tonne — a higher
+        day rate spread across fewer tonnes, so the penalty compounds.
+        Nobody chooses this. The berth chooses it. — Worked example, episode
+        10, using Baltic average earnings of 18,964 and 20,651 USD per day.
+        — https://storage.googleapis.com/podcast-audio-2647223968/commodity-
+        desk-daily/ep10_chart2.png
+
+The two currencies of freight
+-----------------------------
+
+Freight is quoted two ways, and fluency in both is not optional.
+
+                        Voyage charter             Time charter
+----------------------------------------------------------------------------
+Unit                    $/tonne of cargo           $/day of hire
+What you buy            Carriage of a stated       The vessel itself, for a
+                        cargo between named ports  period
+Weather and speed risk  Owner                      Charterer
+Port queue and delay    Owner (subject to laytime  Charterer
+                        and demurrage)
+Bunkers                 Owner                      Charterer
+
+Same steel, two prices, two completely different risk allocations.
+Converting between them is what a chartering desk does all day, and the
+conversion is never clean, because it requires a view on how long the voyage
+will actually take.
+
+That view is the whole argument:
+
+  | TRADER: Sixty-six Santos, first half October. What's it costing me?
+
+  | CHARTERER: Fifty-two on voyage. Or I put you on a Kamsarmax, time
+  | charter, nineteen five.
+
+  | TRADER: Which one do I want?
+
+  | CHARTERER: Depends who you think eats the queue at Santos.
+
+Neither of them called freight $52 and left it there. One number is a price;
+the other is a transfer of risk. Whichever is cheaper depends on the line-up
+at Santos in October, and neither of them knows the line-up in October. What
+they are actually negotiating is who carries that ignorance.
+
+The arb, and the leg everybody leaves open
+------------------------------------------
+
+Baltic route assessments this week put P7, US Gulf to Qingdao, at $73.81/t,
+and P8, Santos to Qingdao, at $51.98/t. That $21.83/t gap decides whose
+beans China buys far more often than the FOB differential does — 15 c/bu of
+origin discount is only $5.51/t, barely a quarter of the freight spread.
+
+Now the trade that actually kills people. Buy 66,000 t FOB Santos, sell the
+same cargo CFR Qingdao.
+
+Line                               $/tonne
+------------------------------------------
+Gross spread (CFR sale − FOB buy)    74.00
+Freight                             −52.00
+Finance, insurance, port             −6.00
+Margin                               16.00
+
+Sixteen dollars a tonne is $1,056,000 on the cargo. Sell the board against
+the physical and flat price is gone. Both differentials are agreed in
+writing. The trade looks locked.
+
+Except the vessel is not fixed. Freight rallies $20/t before it is.
+
+[chart] Santos to Qingdao, one Panamax — Freight is two thirds of the gross
+        spread before a single dollar of margin is counted. A twenty-dollar
+        rally on the unfixed leg turns 1.06 million dollars of profit into a
+        264,000 dollar loss, and the board never moved. — Worked example,
+        episode 10, freight from Baltic P8 Santos–Qingdao at 51.98 USD per
+        tonne. — https://storage.googleapis.com/podcast-
+        audio-2647223968/commodity-desk-daily/ep10_chart3.png
+
+The margin goes to −$4.00/t, a loss of $264,000. A swing of $1.32 m on a
+fully hedged trade.
+
+Is $20/t an outrageous move? Spread over the 66,000 t cargo and the 60-day
+round voyage, it is about $22,000 a day of extra hire — taking a Panamax
+from $18,964 to roughly $41,000, which is near today's Capesize level. The
+BDI has ranged between 1,261 and 2,845 over the past 52 weeks, better than a
+double. Twenty dollars is not a tail event. It is a bad quarter.
+
+The position nobody writes down
+-------------------------------
+
+Here is the sentence worth keeping: the moment a CFR sale is made and no
+vessel is fixed, the seller is short freight.
+
+It runs both ways. Buy CFR and sell FOB and the position is long freight.
+Two trades can look identically flat on the grain and point in opposite
+directions on the ocean.
+
+The reason this hides is structural rather than careless. A position sheet
+has columns for bushels, lots, months and locations. It very rarely carries
+freight at the same resolution, and when it does, the exposure is often
+booked at the expected rate rather than marked to the index. A book that is
+flat in every grain column can be carrying seven figures of directional
+freight risk that nothing on the page names.
+
+FFAs are the answer, as far as there is one: cash-settled swaps on a Baltic
+index route or basket, over a calendar month. A short-freight position is
+covered by buying FFAs. But the hedge is an index built from a basket of
+named routes, and the exposure is one voyage on one route fixed on one day.
+The route can move against the basket, the fixing can miss the settlement
+window, and the tonnage never matches exactly. This is the cross-hedge
+problem from ep 5 — a Black Sea cargo hedged on Matif — wearing shipping
+clothes.
+
+Underneath the freight sits fuel. A Panamax burns roughly 30 t a day at sea,
+so a 60-day voyage is about 1,800 t of bunkers. A hundred-dollar move in
+fuel is $180,000, or $2.73/t on the cargo — seventeen percent of a sixteen-
+dollar margin, decided in the oil market. On a voyage charter the owner
+wears it. On a time charter, the charterer does, which is a large part of
+what the two quotes in that dialogue were really about.
+
+Why merchants charter rather than own
+-------------------------------------
+
+If freight matters this much, why does almost every merchant rent its
+tonnage?
+
+Three reasons, and only the third is about shipping.
+
+A ship is a twenty-five-year asset and a cargo is a sixty-day problem.
+Owning one to solve the other means holding a two-decade position to cover a
+two-month exposure, which is not a hedge — it is a second business.
+
+The merchant's edge is knowing where the cargo is, not knowing where the
+ship is. Those are genuinely different information games, played by
+different people, on different cycles.
+
+And shipping is more violently cyclical than grain. A trading margin
+financed by a shipping balance sheet stops behaving like a trading margin:
+the freight cycle's drawdowns are deep enough to constrain the working
+capital that the grain business runs on, at exactly the moments when grain
+opportunities appear.
+
+The houses that do own tonnage mostly own it for a narrower reason: to
+guarantee access on the days when access, not price, is the binding
+constraint. That is a real reason. It is just not a trading reason, and the
+distinction is worth holding onto, because it is the same distinction that
+separates owning an elevator from having a view on basis — which is Monday's
+subject.
+
+
+----------------------------------------------------------------------------
+Soft Commodity Trading — a daily briefing on physical commodity trading.
+
+GLOSSARY
+Every unit and expression the show has introduced lives on the episode page:
+https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/ep10.html#glossary
+
+All episodes: https://storage.googleapis.com/podcast-audio-2647223968/index.html
+RSS: https://storage.googleapis.com/podcast-audio-2647223968/commodity-desk-daily/feed.xml