Buying Your Own Debt Is Not the Same as Cancelling It: Why Loan Ownership Needs a Lifecycle

A debt investor sells a loan.

A company connected to the borrower buys it.

What happened to the debt?

The intuitive answer is:

The group bought back its own loan.
Therefore the debt went away.

That answer can be wrong.

A loan can change holder without changing borrower.

It can trade at a discount without changing face amount.

An affiliate can become the lender while the borrower continues to owe the same contractual debt.

The loan can remain outstanding even when its economic ownership moves closer to the borrower’s corporate group.

Late August 2026 provided a useful real-world example.

The Financial Times reported on 29 August that an affiliate of Guggenheim Partners had begun purchasing debt linked to its asset-management arm after a roughly $1.2 billion loan had fallen sharply in secondary trading. The loan had traded below 70 cents on the dollar before rebounding to around 84 cents after the affiliate began buying through Bank of America, which manages trading in the instrument. Financial Times: Guggenheim affiliate buys up debt linked to its asset management arm

Bloomberg had reported earlier in the week that the $1.18 billion GIH Borrower LLC loan, due in 2031, could be purchased by Guggenheim affiliates because they viewed it as an attractive investment opportunity. The lender disclosure said purchased term loans were expected to be held by an affiliate as an investment rather than acquired by GIH Borrower LLC for subsequent cancellation. Bloomberg Law: Guggenheim Investments Says Affiliates May Buy Its Hard-Hit Loan

Public SEC-filed fund holdings identify the instrument as a first-lien term loan maturing 26 November 2031, priced at a spread of 250 basis points over SOFR. SEC filing: GIH Borrower LLC First Lien Term Loan

The broader news context around Guggenheim is complex.

But the DaDepo lesson does not depend on resolving those wider issues.

It is much simpler:

A debt purchase is a change in ownership. Debt cancellation is a change in the existence or outstanding amount of the obligation. Those are different lifecycle events.

Private-credit infrastructure should be able to represent both.

A loan has more than one state

Most loan databases are organised around fields such as:

Borrower
Principal
Interest rate
Maturity
Status

That is enough for a simple bilateral loan.

It becomes inadequate once the loan trades.

A syndicated or private-credit loan can have several simultaneous states:

• borrower;
• issuer or financing entity;
• original lenders;
• current lenders;
• agent;
• outstanding principal;
• commitment;
• market price;
• accrued interest;
• legal maturity;
• current rating;
• amendment state;
• affiliate-holder status;
• trading restrictions; and
• cancellation status.

The central error is treating all of those as though they describe the same thing.

They do not.

A secondary sale changes the holder, not the borrower

Suppose Bank A owns a $10 million term loan to Company X.

Bank A sells the loan to Fund B for $7 million.

Before the trade:

Borrower: Company X
Holder: Bank A
Outstanding principal: $10m
Market value: approximately $7m

After settlement:

Borrower: Company X
Holder: Fund B
Outstanding principal: $10m
Purchase price: $7m

Company X still owes $10 million under the loan terms.

The fact that Fund B paid only $7 million does not automatically reduce the borrower’s contractual obligation to $7 million.

That is basic loan trading.

Now change one fact.

Fund B is affiliated with Company X.

The ownership has become more interesting.

The legal mechanics still need to be identified.

Affiliate purchase creates a related-party state

An affiliate acquiring a loan can create several new questions.

Which entity purchased the debt?

How is it related to the borrower?

Does the credit agreement permit the affiliate to hold the loan?

Does the affiliate have voting rights?

Are those voting rights restricted?

Is the loan treated differently for covenant calculations?

Can the affiliate participate in amendments?

Can the affiliate later sell the loan?

Could the borrower repurchase it later for cancellation?

How is the holding accounted for?

The answers depend on the credit agreement, transaction structure, accounting rules and law.

The infrastructure point is that the relationship itself is material data.

A loan record should be able to show:

Current holder: Affiliate A
Relationship to borrower: Affiliate
Acquisition method: Secondary market purchase
Purchase price: [if available]
Principal held: [amount]
Loan status: Outstanding
Cancellation status: Not cancelled

That is much more informative than:

Debt buyback: Yes

“Buyback” is an ambiguous word

Financial reporting often uses buyback loosely.

It can describe at least three very different events.

1. Affiliate secondary purchase

Existing lender
    ->
Affiliate purchaser

The loan remains outstanding.

2. Borrower debt repurchase

Existing lender
    ->
Borrower

Depending on the transaction and governing documents, the debt may then be cancelled, retired or otherwise extinguished.

3. Tender or negotiated debt reduction

Borrower / issuer
    ->
Lenders tender debt
    ->
Debt formally cancelled or retired

These events can have different consequences for:

• outstanding principal;
• voting;
• covenant calculations;
• accounting;
• tax;
• market float;
• creditor concentration; and
• control.

Using one status called Repurchased can destroy that distinction.

Outstanding principal and market price answer different questions

The Guggenheim-linked loan moved sharply in secondary trading.

That creates another common data error.

A loan might have:

Face principal: $1.18bn
Market price: 70

The market value of the instrument at that price is well below face principal.

But the borrower’s contractual obligation has not necessarily changed.

The record therefore needs at least:

Outstanding principal
Market price
Market value
Acquisition price
Carrying value
Recovery estimate

Those are separate values.

A loan trading at 70 cents is not automatically a loan with 30% of principal forgiven.

The price reflects market assessment.

The contract still defines what the borrower owes.

Debt extinguishment requires an actual extinguishment event

The word extinguished should describe something specific.

A debt can cease to be outstanding because of events such as:

• scheduled repayment;
• prepayment;
• borrower repurchase and cancellation;
• refinancing;
• legal set-off;
• exchange into another instrument;
• restructuring;
• conversion;
• compromise;
• discharge; or
• another transaction that legally retires the obligation.

The event should have evidence.

For example:

Event: Debt cancellation
Date: 15 September 2026
Principal cancelled: $50m
Authority: Credit agreement / repurchase documentation
Settlement evidence: Present
Outstanding principal after event: $950m

Without an extinguishment event, the system should not infer that debt disappeared merely because a related party bought it.

Holder identity can change several times

A syndicated loan can have a long ownership history.

Origination
    ->
Bank syndicate
    ->
CLO
    ->
Credit fund
    ->
Distressed-debt investor
    ->
Affiliate purchaser
    ->
New third-party purchaser

The borrower can remain the same throughout.

A lifecycle record should preserve each transfer.

That makes it possible to answer:

• who held the loan on a particular date;
• who received interest;
• who had voting rights;
• who participated in an amendment;
• who sold at a discount;
• whether an affiliate became a creditor; and
• whether the loan later returned to third-party hands.

This history is important in restructuring.

A loan position is not the whole facility

Large term loans are often held by many lenders.

Suppose the total facility is $1.18 billion.

An affiliate buys $100 million.

The correct state may be:

Total facility outstanding: $1.18bn
Affiliate-held portion: $100m
Third-party-held portion: $1.08bn

If the affiliate later buys another $200 million:

Total facility outstanding: $1.18bn
Affiliate-held portion: $300m
Third-party-held portion: $880m

Nothing about those purchases necessarily changes total principal.

A portfolio system therefore needs to distinguish:

facility

from:

individual lender position.

They are related assets.

They are not identical.

Loan identity must survive position splits

One lender may sell only part of its position.

Another may buy portions on several dates.

The loan can be represented as:

Facility ID
    ->
Position A
    ->
Position B
    ->
Position C

Each position can have:

• holder;
• principal amount;
• acquisition date;
• acquisition price;
• settlement date;
• voting status;
• affiliate status; and
• disposal history.

The facility retains:

• borrower;
• total outstanding;
• maturity;
• interest terms;
• covenants; and
• agent.

This hierarchy is essential for institutional loan markets.

Agent records and beneficial ownership can diverge

In syndicated lending, administrative records may be maintained through an agent.

The registered lender position may be reflected in one system.

Economic exposure may sit with another institution through participation or other arrangements.

That creates the same issue seen elsewhere in private markets:

Record holder
    !=
Economic beneficiary

A robust Asset Passport should therefore avoid assuming that one holder field explains every ownership layer.

Where relevant, it should distinguish:

• lender of record;
• beneficial owner;
• participant;
• affiliate;
• investment vehicle;
• investment manager; and
• ultimate economic exposure.

Not every transaction needs every field.

The model should support the distinction.

Cancellation changes the facility, not merely a position

Suppose an affiliate holds $300 million of a $1.18 billion facility.

If the affiliate simply holds it:

Facility outstanding: $1.18bn

If the borrower later acquires and cancels that $300 million:

Facility outstanding: $880m
Cancelled principal: $300m

That is a facility-level lifecycle event.

The position ceases to exist.

The debt-service obligation changes.

Interest expense changes.

Potential lender voting changes.

Covenant calculations may change.

The loan record must therefore link a position-level transfer to a later facility-level cancellation if one occurs.

Price recovery is not credit cure

The Guggenheim-linked loan reportedly recovered from distressed levels after affiliate purchases began.

That movement is interesting.

It should not be interpreted automatically as a change in underlying contractual credit.

A price can rise because:

• a buyer enters the market;
• liquidity improves;
• perceived default risk changes;
• technical selling pressure ends;
• new information emerges; or
• investors expect support.

The legal loan terms may remain exactly the same.

A useful timeline might show:

14 Aug: Borrower information released
17 Aug: Market price falls sharply
25 Aug: Affiliate-purchase disclosure
29 Aug: Affiliate purchases reported
29 Aug: Market price recovers

Those are market events.

They are not amendments to the loan agreement.

Market price should have provenance

Private loans trade in less transparent markets than listed securities.

A price may come from:

• executed trade;
• dealer run;
• bid;
• offer;
• valuation agent;
• fund mark;
• pricing service; or
• manager estimate.

Those are different sources.

The record should therefore distinguish:

Price
Source type
Source
Timestamp
Bid / offer / trade / mark
Quantity
Currency

A reported price of 84 cents may be useful.

It is stronger when the market context is known.

Carrying value can lag market value

Public fund filings can show private-loan holdings at values that differ from later market trading.

That does not necessarily mean one value is wrong.

Valuations use:

• measurement dates;
• methodologies;
• available information;
• market observations; and
• governance processes.

The loan can therefore have:

30 June carrying value
18 August dealer indication
25 August secondary-market bid
29 August executed purchase price

All may be relevant.

They answer different questions at different times.

A reliable loan record should never overwrite historical valuation with the newest price.

Related-party status should be time-dependent

An asset manager may have many entities.

Ownership structures change.

An investor can be unaffiliated at origination and become affiliated later.

An affiliate can later be sold.

The correct representation is therefore not:

Affiliate: Yes

as a permanent property.

It is:

Relationship
Effective from
Effective to
Source

For example:

Holder A
Relationship to borrower: Unaffiliated
Effective: 2024-11 to 2026-08

Holder B
Relationship to borrower: Affiliate
Effective: 2026-08 onward

That history can matter for governance and voting.

Voting rights are a separate state

Affiliate-held debt can be subject to restrictions in credit agreements.

Some facilities may exclude or limit affiliate votes for certain purposes.

Others may permit different classes of affiliated lenders.

The exact treatment varies.

That means the record should not infer:

Principal held = voting power

A lender position may need:

• principal;
• commitment;
• voting eligibility;
• voting amount;
• amendment rights;
• sacred-right protections; and
• restrictions.

The legal documents determine those fields.

A debt transfer may require eligibility and consent checks

Loan transfers are not always freely transferable.

A credit agreement may define:

• eligible assignees;
• disqualified institutions;
• borrower consent;
• agent consent;
• minimum transfer amounts;
• affiliate restrictions;
• competitor restrictions; and
• transfer documentation.

A secondary-market trade therefore has at least two stages:

Economic agreement to sell
    ->
Legal settlement / assignment

The Asset Passport should record the effective holder after the transfer legally settles.

An announced intention to buy is not yet the same as ownership.

Loan-state history is a sequence of events

A mature loan can pass through:

Originated
    ->
Syndicated
    ->
Traded
    ->
Repriced
    ->
Amended
    ->
Affiliate purchase
    ->
Extended
    ->
Partially repaid
    ->
Repurchased
    ->
Cancelled

Some loans never follow most of those steps.

The model should still support them.

The important principle is that each event changes specific fields.

An amendment may change maturity but not holder.

A trade changes holder but not maturity.

A repayment changes principal but not necessarily holder.

A cancellation removes principal.

A price move changes market value without changing contract.

Event-based state prevents these concepts from being confused.

A Loan Asset Passport should preserve both facility and position

For DaDepo, the Guggenheim example suggests a two-level model.

Facility layer

• borrower;
• issuer or financing entity;
• facility identifier;
• original principal;
• current outstanding principal;
• origination date;
• maturity;
• benchmark;
• spread;
• security;
• guarantee;
• covenants;
• administrative agent;
• amendment history;
• repayment history; and
• cancellation history.

Position layer

• holder;
• holder type;
• affiliate relationship;
• principal held;
• acquisition date;
• acquisition price;
• settlement date;
• lender-of-record status;
• participation status;
• voting eligibility;
• current carrying value;
• disposal date; and
• disposal price.

Market layer

• bid;
• offer;
• trade price;
• source;
• date;
• quantity;
• valuation; and
• market commentary source.

Relationship layer

• borrower group;
• parent;
• affiliate holder;
• investment manager;
• beneficial owner;
• guarantor;
• agent; and
• relevant effective dates.

Lifecycle layer

• origination;
• transfer;
• assignment;
• amendment;
• extension;
• repayment;
• repurchase;
• cancellation;
• restructuring;
• default;
• acceleration; and
• discharge.

Provenance

• credit agreement;
• assignment document;
• agent record;
• lender notice;
• SEC filing;
• public report;
• valuation source;
• user-entered information;
• review status; and
• last updated date.

This makes the difference between the loan and a holding in the loan explicit.

The balance should be derived from events

A common system design mistake is to store only:

Current principal: $880m

A stronger system stores:

Original principal: $1.18bn
Repayment event: -$100m
Cancellation event: -$200m
Current principal: $880m

The current state can be calculated.

The history remains auditable.

This matters when investors ask:

Why is the balance lower?

A repayment and a cancellation may both reduce principal.

They are not the same event.

Debt cancellation can have accounting and tax consequences

When debt is actually extinguished, accounting and tax treatment may become significant.

The consequences can depend on:

• who acquired the debt;
• whether borrower and holder are related;
• acquisition price;
• jurisdiction;
• accounting standard;
• tax rules; and
• transaction structure.

DaDepo should not determine those consequences.

It should preserve the facts professionals need to analyse them.

For example:

Face principal acquired: $100m
Purchase price: $70m
Purchaser: Affiliate
Borrower: Entity X
Held as investment: Yes
Cancelled: No

That factual record is far more useful than an inferred accounting conclusion.

Distressed debt makes provenance more important

When a loan trades near par, investors may tolerate small data ambiguities.

When the loan trades at 70, every state matters.

Investors want to know:

• who owns the debt;
• who is buying;
• whether the buyer is affiliated;
• whether the purchases reduce public float;
• whether voting power changes;
• whether principal is cancelled;
• whether the borrower is spending cash;
• whether another entity is funding the purchase; and
• what the latest market price represents.

A clean lifecycle record can answer these questions without reconstructing months of lender notices.

Loan trading creates an asset marketplace problem

DaDepo’s long-term marketplace logic is directly relevant.

A loan position offered for sale needs a defined object.

That object may be:

$5m position in Facility ABC

The buyer needs to know:

• current borrower;
• current outstanding facility;
• position size;
• interest terms;
• maturity;
• transfer restrictions;
• accrued interest;
• settlement mechanics;
• agent;
• security;
• covenant state;
• affiliate restrictions;
• latest documents; and
• ownership provenance.

The sale does not create a new loan.

It transfers a position in an existing one.

That distinction matters for marketplace architecture.

AI can reconcile holder history—but it should not infer extinguishment

AI can help extract:

• borrower;
• loan amount;
• maturity;
• benchmark;
• spread;
• security;
• assignment notices;
• holder names;
• transfer dates;
• prices;
• amendments;
• repayment amounts; and
• cancellation language.

Across a loan history, AI can flag:

• holder records that conflict;
• a transferred position still attributed to the seller;
• current principal inconsistent with repayments;
• an affiliate purchase described elsewhere as cancellation;
• a maturity inconsistent across documents; or
• market prices incorrectly stored as principal values.

That is useful.

AI should not independently decide:

• that an affiliate relationship exists;
• that a transfer is legally effective;
• that voting rights are restricted;
• that debt has been extinguished;
• that a restructuring has occurred; or
• that an accounting gain should be recognised.

Those conclusions require legal, accounting or transaction authority.

What DaDepo can contribute

DaDepo’s role is not to trade Guggenheim debt.

The broader opportunity is to make private-loan state explicit.

A loan Asset Passport can connect:

Borrower
    ->
Facility
    ->
Current terms
    ->
Lender positions
    ->
Transfers
    ->
Market prices
    ->
Affiliate relationships
    ->
Repayments
    ->
Cancellation events

That creates a reliable foundation for:

• private-credit portfolio management;
• secondary loan trading;
• restructuring;
• distressed investing;
• servicing;
• lender reporting;
• amendment consent; and
• due diligence.

The central principle is simple:

ownership events and obligation events must be modelled separately.

What DaDepo does—and does not do

Creating or reviewing a Loan Asset Passport does not mean that DaDepo has:

• confirmed the legal validity of a loan;
• authenticated every assignment;
• established lender-of-record status;
• determined beneficial ownership;
• determined affiliate status;
• interpreted voting rights;
• valued the loan;
• verified market prices;
• determined that debt has been extinguished;
• determined accounting treatment;
• determined tax treatment;
• executed a loan trade;
• acted as administrative agent;
• serviced the loan;
• restructured debt;
• provided financing;
• recommended an investment; or
• provided legal, tax, accounting, investment, financial, restructuring, credit-rating, underwriting or valuation advice.

Important: DaDepo provides technology and information tools. It does not provide legal, financial, investment, tax, accounting, lending, loan-trading, servicing, restructuring, agency, underwriting, credit-rating or valuation advice or services unless a specific service is expressly identified and lawfully provided. Whether a loan purchase results in cancellation, extinguishment, voting changes, accounting consequences or tax consequences depends on the transaction documents, governing law, applicable accounting and tax rules, and the facts of the transaction.

A practical loan-state checklist

Before describing a debt purchase as a “buyback”, ask:

1. Borrower: Which legal entity owes the debt?
2. Facility: Which exact loan or tranche is involved?
3. Original principal: What amount was originally borrowed?
4. Current principal: What amount remains contractually outstanding?
5. Seller: Who sold the position?
6. Purchaser: Which legal entity bought it?
7. Relationship: Is the purchaser affiliated with the borrower, and from what date?
8. Position size: How much principal did the purchaser acquire?
9. Purchase price: At what price did the position trade?
10. Settlement: Has the assignment legally settled?
11. Holder of record: Who is recognised by the agent?
12. Economic holder: Is there a participation or another beneficial-ownership layer?
13. Voting: What voting rights attach to the acquired position?
14. Outstanding status: Does the loan remain outstanding after the purchase?
15. Cancellation: Has any formal cancellation or retirement event occurred?
16. Repayment: Has the borrower paid principal?
17. Market price: Is a quoted price a bid, offer, trade or valuation mark?
18. Carrying value: What value is used by current holders for reporting?
19. Lifecycle: Are transfers, amendments, repayments and cancellations separately recorded?
20. Provenance: Can each state be traced to the credit agreement, agent record, trade, filing or authoritative transaction document?

If the answer to “who owns the debt?” changes, that does not automatically change the answer to “how much debt exists?”

Private credit needs an ownership ledger and an obligation ledger

The Guggenheim-linked transaction is a useful reminder that loan markets contain two parallel histories.

The first is the history of the obligation:

Borrowed
    ->
Amended
    ->
Repaid
    ->
Cancelled
    ->
Discharged

The second is the history of ownership:

Original lender
    ->
Secondary investor
    ->
Credit fund
    ->
Affiliate
    ->
New holder

Those histories interact.

They should not be collapsed.

A borrower can owe the same debt while the holder changes five times.

A holder can buy at 70 while the borrower still owes 100.

An affiliate can own the debt without extinguishing it.

A borrower can later cancel part of the facility and reduce principal.

These are basic distinctions.

They become surprisingly easy to lose once private-credit portfolios are managed through spreadsheets, valuation systems, agent records and news reports at the same time.

The solution is not a more descriptive status field.

It is a proper lifecycle.

For private credit, the market needs to know both:

Who owns the right today?

and:

Does the obligation still exist?

Those are different questions.

A reliable asset record answers both.

Further reading

• Financial Times: Guggenheim affiliate buys up debt linked to its asset management arm
• Bloomberg Law: Guggenheim Investments Says Affiliates May Buy Its Hard-Hit Loan
• Bloomberg Law: Guggenheim Loan Drops to New 73-Cent Low With Firm in Spotlight
• SEC filing: GIH Borrower LLC First Lien Term Loan
• SEC filing: GIH Borrower LLC First Lien Term Loan—Steele Creek Capital Corp.