Iniciar sesión Solicitar una demo
Solutions · Advisors

Map every move the documents permit

Liability management is a question of what the documents allow, asked under time pressure and answered in front of a committee. ExactCov maps the capacity across the structure and cites every figure, so the analysis stands up on the other side of the table.

Solicitar una demo How much can they actually raise?
IndenturesCredit agreementsIntercreditor agreementsSecurity and guarantee documentsConsents, amendments and exchange documentsCompliance certificates
The problem

Capacity is not a number in the document. It is a table with a history.

The analysis is mapped by hand across a dozen documents, usually over a weekend, from one side of the table. The committee deck carries figures nobody can trace back to a clause, and the whole exercise is repeated when the structure changes mid-process. The documents hold the answers; the work is finding them and showing where they came from.

MappedDebt, lien and RP capacity across every entity and document, with each component and the definition it rests on.
Both sidesPriming, drop-down and uptiering exposure assessed from the creditor and the company side.
CitedEvery figure in the Excel and PDF export traceable to its clause and page.
Re-runThe amended documents uploaded, every table recomputed within nine seconds.
The structure map

What the register holds for a liability-management analysis

Built from the executed documents, kept with its history through every consent and exchange.

Capacity by entity

Where in the structure the room is.

  • Ratio debt and general baskets, used and remaining
  • Non-guarantor and unrestricted subsidiary capacity
  • Grower legs, re-valued by the latest certificate

Moves and mechanics

What each move requires and reaches.

  • Unrestricted subsidiary designation and what leaves with it
  • Priming capacity and the vote it needs
  • Cross-default thresholds and the facilities they reach

Consents and exchanges

The thresholds that decide the process.

  • Amendment and waiver thresholds by provision
  • Sacred rights and what cannot be amended without each holder
  • Exchange mechanics read against the original indenture
Capacity

How much can they actually raise?

Capacity is a stacked bar of baskets, each with a used and a remaining portion, re-valued by the latest certificate, consumed by every incurrence notice and moved by every amendment. It is a table computed from the documents and kept with its history, which is what makes it a credit signal months ahead of the maintenance tests.

Read the piece
Compliance certificatere-values every grower Incurrence noticeconsumes a basket Amendment 3moves a cap CAPACITY TODAY, ONE ROW PER BASKET BasketCap todayused / remaining Ratio: unlimited ≤ 4.0xpro forma 3.6x open, 0.4x headroom General: gtr of 100 / 25% EBITDA25% × 920 = 230 80 left Acquired debt75 60 left Capital leasesgtr of 40 / 10% = 92 10 left Incremental free-and-clear150 fully used Debt they can still raise150 + ratio room every cell cites its clause and its certificate line; every rebuild keeps the previous version the same structure again for liens and for restricted payments, sharing the definitions and diverging in the baskets
Capacity is not a number in the document. It is a table computed from the baskets, re-valued by every certificate, consumed by every notice, and kept with its history.
Standing up in negotiation

Numbers the other side can follow to the page

Every figure on every screen traces through a test result and an observation to a page in a stored document. Every change is a new row with an actor. An analysis built this way survives the other side's counsel, the committee and the court, because every number can be followed to the clause it rests on.

Read the piece
EVERY FIGURE TRACES TO A PAGE Screen headroom 8.2% Test result value, basis, comparator Observation NAV 440m, official Stored notice page 3, passage Stored schedule clause, page 11 citescitescitescitesthreshold cites the clause EVERY WRITE IS AUDITED, APPEND-ONLY limit 25m → 22.5mrule: 5% of NAV · obs #418engine · 06:10 · applied limit 22.5m → 25msupersedes row aboveA. Rao · 09:42 · approved next row appends here; no row is ever edited or deleted
Two rules: every figure on a screen can be followed to a page in a stored document, and every change is a new row with an actor. Corrections supersede; they never overwrite.
From our research

Deterministic first, generative second

A parser builds the section tree, the definitions index and the cross-reference graph before any model sees the document. The model chooses from spans the parser found and must cite one, and a validator rejects any answer whose cited span does not contain the value. Capacity arithmetic is code over cited inputs, not a model's opinion.

An architecture for cited extraction
Stage 1: deterministic parser, no model, fully testable section tree definitions index cross-references numbers, dates, % candidate spans per field each with an id and a page box Stage 2: model choose and read answer = value + span id no id, no answer span holdsthe value? to the register, cited data lands → every affected test recomputed: inside 9 seconds first pass 97.4% citation-exact · 99.7% after the second read no: retry once with the mismatch, then review What model-first gets wrong reads 3.50x from the recital, cites the covenant, sounds certain nothing downstream can tell
The model never searches the document. It picks from spans the parser found, and its answer is rejected unless the cited span actually contains the value. The order is the design.
Go deeper

Reading for advisors

Short pieces from the blog and the measurements behind them from Research.

Blog · 3 min

How much can they actually raise?

Incurrence capacity is the question that keeps returning and never has a stored answer. What a capacity table looks like, why it needs a history, and why it is a credit signal months ahead of the maintenance tests.

Read Blog · 3 min

Designing for the auditor, not the dashboard

Every figure traces to a page in a stored document and every write is audited. Second-line functions cannot adopt a tool that produces numbers they cannot defend to internal audit or a regulator, however good the numbers are.

Read Blog · 3 min

What document extraction gets wrong

Threshold schedules, basket stacking, defined-term chains, grower baskets, restated sections and cross-references that leave the document. The six failure modes we test against, and what to ask a vendor to show.

Read Blog · 3 min

The covenant package nobody reads twice

Documents get read hard once at close, then filed. Every question afterwards, capacity, headroom, permitted payments, sends someone back to a PDF. The problem, framed without naming a product.

Read Research · 4 min

Deterministic first, generative second: an architecture for cited extraction

A parser builds the section tree, the definitions index and the cross-reference graph before any model sees the document. The model then answers from candidates and must cite a span the validator can check. Why the order matters, the failure it prevents, and how it recomputes every affected test inside nine seconds of data landing.

Read Research · 3 min

How deep do defined terms go? Definition chains in 300 agreements

A leverage covenant is one line. Resolving it means following Consolidated EBITDA into Consolidated Net Income and onward, a median of four hops and up to nine. Where the chains go, how often they leave the document, and what that does to extraction.

Read
On the platform

The parts advisors use most

Everything on this page runs on the same register as the rest of ExactCov. These are the parts this team lives in.

Platform

Agreements Intelligence

Debt and lien capacity, restricted payments, portability, MFN and value-leakage flags on the day the documents arrive, every figure cited to its clause and every definition resolved to its leaf.

Agreements Intelligence Platform

Vault

Every agreement, amendment, certificate and notice, every version, with the page each number came from. Dependencies between documents tracked, regional residency and zero data retention.

Vault Platform

Command Center

The whole book on one screen: what has arrived, what is stale, what is in breach and what is drifting, with headroom trends, scenario shocks and committee packs built from the register.

Command Center
Getting started

Run it in parallel. Don't run a pilot.

The documents are loaded once. What already arrives by email is copied to a mailbox we watch. Nothing changes for the team, and after one cycle the two outputs are compared, disagreement by disagreement.

Week 0The structure's documents loaded. Capacity tables built and checked against the team's own model.
Weeks 1 to 4Live mandates run on both sides. Structure maps, capacity tables and exports produced from the register.
Week 4Compare: capacity the model missed, mechanics the documents allow that the deck did not show, hours per mandate.
AfterPer mandate. Every figure in the committee deck opens to its page.

Map a live structure with us

Bring the indenture, the credit agreement and the intercreditor from a mandate. We show the capacity by entity, the moves the documents permit and where each figure came from.

Solicitar una demo Contact us