Baltic Dry Index and freight indices

The Baltic Dry Index is a weighted composite of three dry bulk timecharter averages published daily by Baltic Exchange Information Services Ltd, computed as RoundedSum(C5TC x 0.40, P5TC x 0.30, S11TC x 0.30) x 0.10 under the Guide to Market Benchmarks v8.6 of June 2026. It carries no unit and no dollar value: the money is in its three constituents, each of which is assessed in USD per day.

Around it sits a family of freight indices covering dirty and clean tankers, LPG and LNG carriers, and container trades, all built the same way. A route is defined down to the ship, the cargo, the laycan and the turn time; a panel of brokers assesses that route each day; and the assessments are averaged. A fixture recap records what a real ship actually agreed; an index records what the defined ship would. Everything a practitioner needs to know about reading a freight index follows from how tightly the route is defined and how the average is taken.

What a freight index measures, and what it does not

A Baltic route assessment is the panel’s judgement of the prevailing open market level for a precisely defined voyage or period employment, on a precisely defined ship. It is not the average of last week’s fixtures, and it is not a price at which anything can be bought.

That distinction decides how the index should be used. A real ship differs from the benchmark vessel in age, gear, consumption, scrubber fitting and position, and a real cargo differs in quantity, laycan and terms, so a fixture that prints away from the index is normal rather than anomalous. Nor does an index carry the operational terms that decide when a voyage starts earning, such as the notice of readiness regime or a deadfreight exposure. The index measures the market; a voyage estimate measures the ship.

The index is also the settlement reference for the freight derivative market, which is what gives its methodology legal weight. A change to a basket, a benchmark vessel or a derived constant changes the underlying of every open forward freight agreement on that route, which is why the Guide carries a formal benchmark change and cessation regime rather than a change log.

The Baltic Dry Index formula, its weights and its multiplier

The BDI is 40 percent Capesize, 30 percent Panamax and 30 percent Supramax, applied to the three timecharter averages and scaled by a flat 0.10. The Supramax input is S11TC, the eleven-route average, not S10TC: S10TC exists but is derived by subtracting a published constant.

Two figures in wide circulation are wrong. The multiplier is not 0.10809, a string that does not appear in the Baltic’s methodology document. And Handysize is not in the basket, having been removed in 2018.

The Baltic’s own account of its history is narrower than the one usually repeated. Its benchmarks were initially produced in 1985, and the BDI itself came into operation on 1 November 1999 as successor to the Baltic Freight Index. A base date and base value for the BDI do not appear in any Baltic methodology document or benchmark statement, so this article states none.

The 1 March 2018 re-weighting, and what came before

The current weights took effect on 1 March 2018 under Baltic Exchange Circular 04/18, announced in January 2018 after member consultation. The stated rationale was research measuring each vessel type’s contribution to the dry bulk market at 40 percent Capesize, 25 percent Panamax, 25 percent Supramax and 10 percent Handysize, derived from fleet composition, utilisation including ballast legs, and cargo moved from import and export reporting and AIS data. Handysize was dropped rather than carried at 10 percent, and the residual weight went to Panamax and Supramax.

From 1 July 2009 until that change the index ran on four classes at equal 25 percent weights. Comparing a 2015 BDI level with a 2025 level therefore compares two different baskets.

The mechanism that keeps the series continuous is the multiplier reset. On a composition change the new multiplier is the rounded average of the daily ratio of the old headline index to the new timecharter average across a dual reporting period, which preserves the level and the direction of the series while the basket underneath it changes. A BDI history is comparable as a direction over decades and comparable as a basket only within a methodology period.

Publication times and reporting windows

Every index has a fixed reporting window and a fixed publication time, and they are not all the same, which matters to anyone timing a settlement or a hedge.

Source: Guide to Market Benchmarks v8.6, Appendix 1.

Publication more than 45 minutes past the stated time is a deemed late publication. A panellist is not barred from contributing outside the window where BEISL judges the contribution the most accurate indicator available, which is the Guide accepting that a thin market does not respect a clock.

Baltic Capesize Index: five routes, and the weight most people get wrong

The Capesize timecharter average is C5TC_182 = C8 x 0.15, C9 x 0.125, C10 x 0.35, C14 x 0.25, C16 x 0.125, on the 182,000 dwt benchmark vessel. The headline BCI applies the same weights and a multiplier of 0.11026.

C10, the Qingdao delivery Pacific round voyage of 35 to 45 days, carries 35 percent, the largest single weight in any Baltic dry basket. A description of the Capesize basket as five routes at roughly equal weights, or one that uses the older 2014-suffixed route codes, is describing a superseded index.

The benchmark vessel is BCI182: non-scrubber-fitted, 182,000 tonnes deadweight on 18.2 m summer salt water draft, maximum age 10 years, 292 m long, 45 m beam, 123 tonnes per centimetre immersion, 199,500 cubic metres grain. Consumption is 52 tonnes laden and 44 ballast at 14 knots, 44 and 36 at 13 knots, 36 and 29 at 12 knots, 29 and 23 at 11 knots, with no diesel at sea. Every panellist assesses as though the ship were this ship, which is why an owner comparing a real Capesize against the index has to adjust for age, gear and fuel curve before the comparison means anything.

The four timecharter routes each cover a distinct pattern: C8 a transatlantic round voyage of 30 to 45 days from the Gibraltar to Hamburg range; C9 a fronthaul from the Amsterdam, Rotterdam and Antwerp range or passing Passero to the Far East over about 65 days; C10 the Pacific round; C14 a Qingdao delivery round voyage via Brazil or West Africa over 80 to 90 days; and C16 a backhaul from North China and South Japan to the Continent and Mediterranean, adjusted to 65 days.

The Capesize voyage routes

The voyage routes are quoted in USD per tonne and are what a cargo owner actually fixes against.

Source: Guide to Market Benchmarks v8.6, Appendix 2.

Two definitional points are worth stating because they are commonly misreported. C5 is defined as West Australia to Qingdao and names no load port, so any account that specifies Port Hedland or Dampier is adding a term the route does not carry. C14 is a round voyage via Brazil or West Africa, not Brazil alone.

The C3 against C5 spread is the most watched relationship in the dry market, because it prices the ton-mile difference between Brazilian and Australian iron ore into the same fleet. C4 and C15 have been retired, and the C4 forward contract was deleted in April 2024.

Baltic Panamax Index: five routes, with 30 percent on P6

P5TC = P1A x 0.25, P2A x 0.10, P3A x 0.25, P4 x 0.10, P6 x 0.30, on the 82,500 dwt BPI82 vessel, with the headline BPI applying a multiplier of 0.111111. The heaviest weight sits on P6, a Singapore delivery trip of 90 to 105 days via the Atlantic with redelivery in the Hong Kong to South Korea range including Taiwan.

The benchmark vessel is 82,500 tonnes deadweight on 14.43 m draft, maximum 12 years, 229 m long, 32.25 m beam, 97,000 cubic metres grain, burning 33 tonnes laden at 13.5 knots or 31 tonnes ballast at 14 knots plus 0.1 tonnes of gas oil at sea, dropping to 22 and 23 tonnes at the slower pair. The Panamax beam figure is a legacy of lock dimensions rather than a current constraint, since the Panama Canal Neopanamax locks opened in 2016.

Two grain voyage routes sit alongside the timecharter basket: P7, Mississippi river to Qingdao, and P8, Santos to Qingdao, both 66,000 tonnes of heavy grain, soya beans and sorghum with a minimum 13 m arrival draft. P5_82, the South China trip via Indonesia known as BEP Asia, is published from Singapore.

The Panamax 74 series is derived rather than assessed: P4TC = P5TC minus 1,336, with matching constants of 1,284, 1,489 and 1,302 on the individual P1A, P2A and P3A routes. P7’s forward contract was deleted in March 2024 and the P4TC forward assessment was removed in May 2026.

Baltic Supramax Index: eleven routes on a geared ship

S11TC covers S1B, S1C, S2, S3, S4A, S4B, S5, S8, S9, S10 and S15, weighted 0.05, 0.05, 0.15, 0.15, 0.075, 0.10, 0.05, 0.10, 0.075, 0.10 and 0.10, with the headline BSI applying 0.079112625. S2, the north China Australian or transpacific round voyage, and S3, north China to West Africa, carry the joint-largest weights at 15 percent each.

The benchmark vessel BSI63 is the one that best explains why a Supramax index exists at all: 63,500 tonnes deadweight on 13.418 m draft, maximum 15 years, 199.98 m long, five holds and five hatches, and four 30 tonne cranes with four 12 cubic metre grabs. The gear is the point. A Supramax serves berths without shore equipment, which is a different trade from the Capesize and Panamax ore and grain lanes even where the commodity is the same. A geared ship is a different asset from a gearless bulk carrier of similar deadweight, and the two indices price that difference directly.

The Baltic Supramax Asia Index, S3TC_63, is a three-route regional basket of S2 at 42 percent, S8 at 29 percent and S10 at 29 percent, published from Singapore. S8 and S10 are both defined on a coal cargo basis. S10TC, the Supramax 58 series, is S11TC minus 2,034. S6TC was removed from the index in June 2026 as discontinued.

Baltic Handysize Index and the 2026 regional split

HS7TC covers seven routes on the 38,200 dwt BHSI38 vessel, weighted 0.125 on HS1 to HS4 and 0.20, 0.20 and 0.10 on HS5, HS6 and HS7, with the headline BHSI applying 0.055556. The benchmark ship is a self-trimming geared bulk carrier on 10.538 m draft with four 30 tonne cranes, maximum age 15 years.

April 2026 added two regional averages, with the Pacific weighting corrected in June 2026. HSATC is the Atlantic basket, HS1 to HS4 at 25 percent each, covering the Continent to South America, the Continent to the United States Gulf, and both return legs. HSPTC is the Pacific basket, HS5 and HS6 at 40 percent each and HS7 at 20 percent.

The Handysize 28 series is derived as HS7TC minus 1,966. Handysize left the BDI in 2018, so the index is a standalone read on the smallest geared segment rather than a component of the dry composite.

Derived series: why some quotes are subtractions

Four dry series are published as a constant subtracted from an assessed basket rather than assessed in their own right: C5TC(180) = C5TC(182) minus 3,503; P4TC = P5TC minus 1,336; S10TC = S11TC minus 2,034; HS6TC = HS7TC minus 1,966.

This matters commercially rather than academically. A trader quoting the 180 series and a trader quoting the 182 series are looking at one panel assessment offset by a fixed number, so the spread between them cannot move and carries no information. And when a derived series is retired, as P4TC’s forward assessment was in May 2026, nothing about the underlying market has changed.

Baltic Dirty Tanker Index: eleven routes at one eleventh

BDTI is the equal-weighted sum of TD2, TD3C, TD6, TD7, TD8, TD9, TD14, TD15, TD19, TD20 and TD32, each at 0.0909090909, multiplied by 8.415737054. Every dirty route carries 3.75 percent total commission, and since January 2026 each Worldscale route has a twin reported in USD per tonne.

TD18 was removed and TD32 added in June 2026. Sources still listing TD18 in the basket are out of date, including some of the Baltic’s own service pages, and TD1, TD12 and TD17 were retired earlier.

Source: Guide to Market Benchmarks v8.6, Appendix 2.

The Baltic also publishes dirty routes outside the index, including TD21, TD22, TD23, TD25, TD26, TD27, TD28, TD29, TD33 and TD34, several added during 2026. A route outside the basket is assessed to the same standard; it simply does not move the composite.

Baltic Clean Tanker Index: five routes since April 2026

BCTI is TC1, TC2_37, TC5, TC6 and TC23, each at 20 percent, multiplied by 4.32344602027436. TC16 was removed in April 2026, which is what took the basket from six routes to five and changed the multiplier from 4.540991088.

TC1 and TC5 are the Ras Tanura to Yokohama long-haul pair at 75,000 and 55,000 tonnes; TC2_37 is Rotterdam to New York at 37,000 tonnes of unleaded gasoline; TC6 is the 30,000 tonne Skikda to Lavera cross-Mediterranean route; TC23 is 30,000 tonnes of clean product or ultra low sulphur diesel from Amsterdam, Rotterdam and Antwerp to the UK Continent.

Two published routes are routinely misdescribed as constituents and are not. TC17, 35,000 tonnes Jubail to Dar es Salaam, and TC7, 35,000 tonnes Singapore to Sydney, are both assessed and both outside the basket. The wider clean route list runs to TC24, added in October 2025, and includes lumpsum-basis routes such as TC15 and TC20 where a percentage would not be meaningful. The dirty and clean families are treated in more depth at the Baltic dirty and clean tanker indices .

One assessment, three published numbers

Every index tanker route is assessed by panellists in Worldscale points, and the Baltic then publishes three things from that single assessment: the Worldscale percentage, a USD per tonne figure obtained by applying the applicable flat rate, and a TCE in USD per day derived from its own published formula.

This is what makes the tanker indices usable across a year boundary. A Worldscale percentage is a percentage of a flat rate that is recalculated annually and published each November for the following calendar year, so WS 60 in one year and WS 60 in the next are percentages of two different dollar figures. The flat rate and differential structure is what carries the cost inputs, and the USD per tonne and TCE figures are what survive the annual reset.

The Baltic also publishes derived TCE baskets: VLTCE is the average of the TD3C, TD15 and TD22 TCEs; SZTCE the average of TD6 and TD20; ATCE the average of TD7, TD8, TD14, TD19, TD25 and TD26. Two MR baskets, MA2TCE and MP2TCE, are built on a triangulation basis rather than a round voyage, which is the index acknowledging how those ships actually trade.

The Baltic standard tankers

Each tanker TCE runs on a named reference ship, all non-scrubber-fitted, and the descriptions are what make the daily figures comparable across routes.

Source: Guide to Market Benchmarks v8.6, Appendix 2. Eco speed and consumption pairs are published alongside these.

Two features of the descriptions are invisible in a Worldscale percentage and both move a TCE materially. The standard ship steams a knot slower in ballast than laden, so the ballast leg is longer in days on an equal-distance round voyage. And discharge burns far more fuel than loading, 110 tonnes against 20 on the VLCC, because of cargo pumping. Anyone reconciling a house estimate against a Baltic TCE has to match those before comparing. The full tanker size class ladder maps onto these reference ships, and the parallel product trades are priced on the same vessels through a chemical tanker or clean fleet rather than a crude one.

Gas: the LPG and LNG indices

BLPG is the rounded average of the BLPG1, BLPG2 and BLPG3 TCEs multiplied by 0.1. All three routes are 44,000 tonnes of fully refrigerated LPG, one to two grades, 96 hours total laytime, maximum age 20 years and 1.25 percent commission: Ras Tanura to Chiba, Houston to Flushing, and Houston to Chiba. The third is defined as routing through the Panama Canal with two days of total waiting built into the definition, which is a route definition carrying a canal queue as a fixed assumption. The TCEs run on VLGC84, an 84,000 cubic metre carrier.

BLNG is the rounded average of BLNG1-174, BLNG2-174 and BLNG3-174 multiplied by 0.1, all in USD per day on the LNGC174: 174,000 cubic metres, two-stroke propulsion, 0.085 percent boil-off, delivered cold and ready to load. BLNG1 is a Gladstone round voyage via Tokyo of 22 days, BLNG2 a Sabine round voyage via the Isle of Grain of 28 days, BLNG3 a Sabine round voyage via Tokyo of 53 days routed through the Panama Canal.

The LNG method differs from every other Baltic index. Panellists submit a headline daily rate and a lumpsum, the lumpsum being compensation for positioning or repositioning, and the Baltic computes a round voyage daily hire net of the repositioning cost, using a 6 percent weather allowance rather than the 5 percent used on tankers. Three further assessments, BFLNG1 to BFLNG3, are published in USD per cubic metre and are delivered LNG price assessments rather than freight. The reference ship is a modern two-stroke LNG carrier rather than a steam turbine one, which is itself a statement about which fleet sets the market.

The 160,000 cubic metre LNG series ceased in October 2025, its routes were deleted from the guide in January 2026 and the residual index entry was removed in February 2026. Only the 174 series survives.

Container indices: four publishers, four methods

Container freight is indexed by lane and by box rather than by ship, so the container indices are constructed differently from the bulk ones and are not interchangeable.

The outlier treatments differ and the difference shows up in volatile weeks. FBX uses a two-sided median absolute deviation test with coefficients set per index between 2 and 5, falling back to the median alone where more than half the data share one value. Drewry publishes the median of submitted rates, discards any rate more than 30 percent above or below it, a threshold revised from 20 percent in week 20 of 2021, then averages within each monitored carrier before averaging the carriers.

The SCFI is the only one of the five with a published basis: 16 October 2009 at 1,000 points, assessed from a panel of 20 liner companies and 17 forwarders and NVOCCs. The FBX is administered by BEISL under the United Kingdom regime, not by an EU authority, and describing it as EU-regulated is wrong. The relationship between the indices and the surcharge structures they include is treated at ocean freight cost and surcharges and ocean freight rates , and the fleet context at container ship size classes and the container ship itself.

How an assessment is made

Input data is the professional judgement of the panellist, at the time of assessment, of the prevailing open market level for the defined route. Where an active market exists the Guide expects that judgement to be informed by and anchored in fixtures and current negotiations, adjusted to the Baltic’s route definition rather than reported raw.

Expert judgement is an explicit governed mechanism rather than a gap in the method. A route can be routinely fixed when adopted and go quiet later, and panellists then assess from the wider market, evaluating discretion-based input internally at an appropriate level of seniority before submission. That is the honest answer to the question of what a freight index is on a day when nothing fixes.

The index is a simple arithmetical average of all input data received. There is no trimmed mean on the physical routes. The single exception is the wet bulk forward curve, where inputs are weighted by each panellist’s clearing house market share in the preceding month, which ties the forward assessment to who is actually clearing volume.

Panel size and route selection are governed. BEISL will not normally create a physical route index without at least five panellists and an average trade pattern of at least two vessel voyages a week, and will not normally publish where fewer than four can contribute; a forward curve may run on two, with a published statement of the shortcomings. Route criteria are trade volume, avoiding seasonally closed trades such as the Great Lakes, transparency guided at an average of two accurately reported fixtures a week over twelve months, and standard terms. Panellists must be members of the Baltic Exchange Limited, and a shipbroker house contributes as a firm rather than an individual.

Outliers, validation and the code of conduct

With no trimming, data quality is managed by contact rather than by arithmetic. BEISL’s system flags input falling outside a predetermined per-route tolerance range, and the assessor contacts the panellist. The Guide is explicit that BEISL will never require a panellist to change input data or impose such a change; the panellist may ask for its own submission to be corrected, and the intervention is system-recorded.

Assessor priority runs missing data first, then flagged tolerance breaches, then illiquid or difficult routes, then panellists who may not be attentive. All contact and all individual panellist inputs are retained for five years. Publication is authorised by the Senior Assessor only after satisfactory validation.

The Guide itself constitutes the code of conduct for the determination of BEISL benchmarks, and panellists must additionally comply with The Baltic Code, whose Code of Ethics and Market Practice holds that impeccable standards of honesty and integrity are critical to the role. A methodology change that would impose significant additional cost on panellists, or significantly affect their underlying business, requires consultation and the written consent of at least 75 percent by number of respondents, except where the change is required by law. Governance sits across three bodies: the BEISL Board, the Baltic Index Council , and the BEISL Oversight Function.

The regulatory position

The regulated entity is Baltic Exchange Information Services Limited, not the Baltic Exchange itself. Baltic Exchange Ltd is a wholly owned subsidiary of Singapore Exchange Ltd, and the Guide records that it is not directly involved in the production, management or distribution of the data.

BEISL was authorised by the Financial Conduct Authority as a benchmark administrator in an announcement of 16 March 2020, made while the FCA was still an EU national competent authority under Regulation (EU) 2016/1011, and that authorisation carried into the onshored UK Benchmarks Regulation after the exit transition ended on 31 December 2020. The Capesize, Dirty Tanker and Clean Tanker families are classified as significant benchmarks under Title III Article 24, on the ground that there are no or very few appropriate market-led substitutes.

Twelve benchmark statements exist, covering BCI, BPI, BSI, BHSI, BDTI, BCTI, BLNG, BLPG, BITRA, BOPEX, BSPA and BSRA, published on 16 March 2020 under Article 27 and Commission Delegated Regulation (EU) 2018/1643. They carry no date of last update and they are materially stale on composition: the Capesize statement still lists C4 and the older route suffixes, and the dirty tanker statement still lists TD1, TD12, TD17 and TD18. Where the statement and the Guide disagree, the Guide is the current document.

Assurance is external and dated. PricewaterhouseCoopers LLP issued a reasonable assurance opinion on the design and operating effectiveness of BEISL’s controls for container box and ocean bulk index administration for the twelve months to 31 December 2023, published 8 May 2024 in the fourth annual cycle, tested against the UK Benchmarks Regulation, the IOSCO Principles for Financial Benchmarks and the IOSCO Principles for Oil Price Reporting Agencies. All control objectives were reported as achieved.

The Baltic indices are not EU-regulated, and the distinction is not pedantic. A live query of the ESMA benchmarks register on 4 September 2026 returned six United Kingdom administrators, none of them a Baltic entity. Regulation (EU) 2025/914, in force 8 June 2025 and applying from 1 January 2026, narrows the Benchmarks Regulation broadly to critical and significant benchmarks, EU climate benchmarks and certain commodity benchmarks and makes ESMA the single supervisor for non-EU benchmarks, so the absence of a register entry is not by itself evidence that an EU firm cannot reference the indices. The FCA states that third-country administrators must be approved through recognition or endorsement, absent an equivalence decision, by 31 December 2030.

From a route assessment to a TCE: the Baltic’s own formula

The Guide gives the tanker TCE as (gross freight minus fuel cost minus voyage costs) divided by total voyage duration, published net of commissions in USD per day.

Gross freight is cargo quantity multiplied by the USD per tonne rate, itself the Worldscale assessment at the applicable flat rate or a lumpsum divided by the loaded quantity. Fuel cost is laden, ballast and port days each multiplied by the matching consumption and the fuel price. Voyage costs are port costs plus commissions plus miscellaneous items including additional insurance, cleaning, security guards and emission costs. Total voyage duration is laden plus ballast plus port days, with sea days carrying a 5 percent weather allowance.

Three points follow that a practitioner should hold onto. Commissions sit inside voyage costs, so the published TCE is already net and adding address and brokerage commission on top double-counts. Port fuel sits inside fuel cost, using the in-port consumption from the standard vessel description. And the Baltic derives tanker TCEs but not dry bulk TCEs: the dry timecharter routes are assessed directly by panellists in USD per day, so C5TC is not a computed TCE at all, and a daily equivalent quoted off C5 or C3 is somebody’s own voyage calculation.

The priced inputs are named rather than assumed. Fuel is taken from the Baltic Bunker Report of the day before the assessment, at the main bunkering port nearest the load port with the grade matched to the activity, using the last published value on a UK public holiday. Port and canal costs come from Cory Brothers Shipping and distances from the AtoBviaC tables. A route with a canal transit, such as the several routes crossing the Suez Canal , carries the transit days and cost in the route description rather than as an adjustment.

Worked example: a Capesize C5 voyage

Structural inputs are from the Guide: cargo 170,000 tonnes, 5 percent total commission on the timecharter routes, and the Baltic Capesize vessel at 14 knots on 52 tonnes laden and 44 ballast with no diesel at sea. The freight rate, distances, port costs and bunker price below are illustrative and are not market levels.

Now change one input. Double the ballast leg to 7,200 nautical miles, leave the freight rate untouched, and the round voyage becomes 39.75 days with 495 tonnes more ballast fuel: the TCE falls to USD 17,182 per day, a drop of 47 percent. That is the whole argument for reading a daily equivalent instead of a dollar per tonne quote, and it is why ballast bonus negotiation exists.

Worked example: a TD3C voyage on the Baltic method

Structural inputs are again from the Guide: cargo 270,000 tonnes, 3.75 percent total commission, VLCC300 at 13 knots laden on 57.3 tonnes and 12 knots ballast on 39.5, in port 20 tonnes at load, 110 at discharge and 10 waiting, with 2 days load, 2 days discharge and 2 days waiting, delivery Ningbo for a round voyage loading Ras Tanura, 5 percent weather margin. Worldscale points, flat rate, distance and bunker price are illustrative.

The two asymmetries called out earlier are visible in the numbers. The ballast leg takes 21.875 days against 20.192 laden over the same distance, and the discharge port burns 220 tonnes against 40 at load. Neither is anywhere in the Worldscale percentage.

Reading the indices without misreading them

Four habits separate a useful reading from a misleading one, and each of them follows from something in the methodology rather than from market opinion.

Read the constituents, not the composite. The BDI is a direction. C5TC, P5TC and S11TC are the numbers with dollars attached, and they routinely move in opposite directions because the Capesize and Supramax fleets serve different trades. A composite that is flat can hide a Capesize collapse against a Supramax rally.

Match the basis before comparing. A Baltic TCE uses a defined ship, a defined round voyage, a 5 percent weather margin and a named bunker price. A speed assumption is part of that basis, so an estimate built on slow steaming economics or on an optimum speed calculation is not measuring the Baltic’s ship. A house estimate that differs on any of those should differ in its answer, and the gap is information about the ship rather than error in the index. That reconciliation is the same discipline a time charter equivalent comparison always needs.

Do not compare a Worldscale percentage across a year boundary, because the flat rate underneath it is reset annually. The liner equivalent of that reset problem is the surcharge layer, which is why a container index states whether it is all-in and a dry cargo fixture states its bunker adjustment factor separately.

Check the version. The composition of these baskets changed materially during 2026: TC16 out in April, TD18 out and TD32 in during June, S6TC discontinued in June, HSATC and HSPTC added in April and re-weighted in June. A description sourced from a benchmark statement of 2020, or from a secondary write-up, is describing an index that no longer exists. Where an index moves in a way the physical market did not, the first question is whether the basket changed.

Fixed-price exposure to any of these levels is managed in the derivative market rather than the physical one, through forward freight agreements settling against the same published series, and period cover through a time charter or a trip time charter rather than a series of voyage charters . A contract of affreightment or a GENCOA can carry an index escalation clause, in which case the index definition becomes a contractual term and its cessation regime becomes a commercial risk. The charter party framework around all of these, including the laytime and demurrage terms embedded in each route definition, is what the index is abstracting from. Older benchmarks such as the AFRA average freight rate assessment survive in parallel for different purposes, and emissions-linked reporting frameworks such as the Sea Cargo Charter and the Poseidon Principles for Marine Insurance increasingly reference the same trade data.

Limitations

An index is an assessment, not a transaction record. No money changes hands at the published level, and on an illiquid day the number rests on governed expert judgement rather than on reported fixtures. That is disclosed in the methodology rather than hidden, but it bounds how much weight a single day’s print will carry.

Composition changes break comparability. The BDI basket changed in 2009 and again in 2018, the tanker baskets changed twice in 2026, and derived series are retired and added. A long chart is comparable in direction because of the multiplier reset; it is not comparable as a basket.

The benchmark statements are behind the methodology. The twelve statements of 16 March 2020 carry no update date and list routes that have since been retired. Anyone taking a route list from a statement rather than from the current Guide will be wrong about the composition.

No index level appears in this article, deliberately. A published assessment is stale the day after it is printed, and a level quoted here would be a worse source than the Baltic’s own daily publication. The same applies to the container indices.

Some numbers commonly attached to these indices could not be traced to a Baltic document and are therefore absent. The BDI base date and base value, the pre-2018 multiplier, the BDTI and BCTI launch base, the number of panellists on any index, and the XSI base are all in wide circulation and none of them appears in the Guide or the benchmark statements. Where a figure matters commercially, take it from the administrator rather than from a secondary description, including this one.

Related Articles