Cipher Digital Credit Risk: Thesis Test and Instrument Assessment

Thesis test and instrument assessment

Research date 16 October 2026 | USD millions unless stated otherwise

Executive assessment

Cipher is partly consistent with the financing-dependence thesis, but the evidence does not establish an attractive credit short. Its strongest vulnerability is construction-related cash demand at the parent before projects distribute cash. Its completed, performing Amazon-backed leases are a weaker fit for a thesis based on disappointing AI monetization.

Premise 1. AI investment can disappoint economically without immediately impairing a landlord’s contracted rent. To reach creditors, the disappointment must affect tenant payment, enforceability, delivery, refinancing or renewal. Cipher therefore requires a more specific transmission mechanism than slower AI adoption.

Premise 2. Construction spending is paid before the resulting rent is collected. Completion guarantees can move project shortfalls onto Cipher Digital Inc. Expensive changes, delays and competing development commitments can consume parent liquidity even where long-term project returns remain positive.

Premise 3. External funding and protections can interrupt that chain. The new convertible financing, project accounts, tenant guarantees, cost-sharing and equity issuance are material defenses. They must be applied to the right entity, period and obligation, rather than pooled as unrestricted cash.

Conclusion for Theseus

The strongest candidate hypothesis is a parent funding squeeze driven by completion obligations, restricted access to borrowing and limited project distributions. Barber Lake is the highest-priority project to investigate. A broad short of all Cipher project debt would ignore important differences in creditor protection.

The conditional model in this report produces no gap in its defensive case through Q2 2027. Its stress case falls below a $200m analytical cash buffer in Q1 2027 and below zero in Q2. Those dates are consequences of explicitly selected assumptions, not estimated probabilities or evidence of an actual forthcoming default.

No issuer-specific executable CDS quote, confirmed bond borrow or complete current recovery waterfall was obtained. The public-source search found a financing scheduled to settle today, but no primary closing confirmation. Financing proceeds therefore remain conditional throughout this report.

Borrowers and obligations

The group’s legal structure matters more than its consolidated debt headline. Project notes finance distinct collateral and payment streams; parent convertible holders depend on parent assets and value remaining after project claims. Completion support is not a blanket parent guarantee of every project-note payment.

Principal and parent/project mapping: June capitalization [2]. Issue mechanics: [4], [6], [8], [11], [13], [16]. Proposed transactions are not inserted into June reported debt.

Annual project coupon at unchanged principal is $123.48m for Barber Lake, $122.50m for Black Pearl and $48.60m for Stingray, totaling $294.58m before amortization, fees or parent obligations. This arithmetic is not consolidated cash interest guidance.

Payment dates are May/November for Barber Lake, February/August for Black Pearl and June/December for Stingray. Black Pearl initially amortizes at 7% of issue principal annually after all phases complete; Barber Lake follows its indenture schedule; Stingray amortization depends on NOI. The final project principal repayment dates arrive before the underlying long leases expire, leaving refinancing or repayment requirements. [8, 11, 13]

The planned exchange is debt relief through equity issuance, not cash raised. Even a small surviving 2030 convertible balance may matter for the revolver’s springing maturity; check retirement and amendment evidence rather than assuming the exchange eliminates that clause. [4, 6]

Construction funding and historical cash flow

Barber Lake cost reconciliation

October presentation, slides 10–11: approximately $750m additional equity; original capitalization includes $710m equity plus $1,733m debt, with reserves, construction interest and fees. Figures are rounded. Estimate at completion is total project cost, not spending still unpaid. The lower net-cost figure deducts reimbursement collected over twenty years. It is not a current cash requirement. [3]

The September contract disclosure assigns the first $359.3m above the original budget to Cipher; 50% above that threshold is reimbursed through additional rent. Delivery is phased through Q4 2026–Q1 2027. Long-term reimbursement supports economics but does not fund invoices automatically. [5]

Estimated new-offering cash is $835.1m after purchaser discounts less $89.7m capped calls = $745.4m before other expenses. That nearly matches the approximate equity need, rather than providing $850m of spare cash. Uses also include broader development. No closing confirmation was located; the $150m option is excluded. [4]

Historical consolidated cash flow control

June unrestricted cash was approximately $831.8m; project-restricted cash approximately $3,728m [1, 2]. Bitcoin sales of $123.428m sit in investing cash flow, so operating cash flow alone overstates economic operating burn. Cash capex was $964.261m; capex in payables/accruals was $542.730m. Supplier commitments of about $2,169m are not an independently reconciled remaining construction budget. Do not add these categories together without eliminating overlap. [1]

Liquidity access and loss transmission

The revolver cannot be treated as cash

The $200m revolver has a $50m pre-completion exposure cap. June letters of credit of $33.1m imply only $16.9m incremental room under that cap, before other conditions; the equivalent post-cap ceiling would be $166.9m at unchanged usage. These are dated sensitivities, not confirmed October availability. [1, 6, 7]

The agreement defines Commencement Date as lease cash flows across all phases. Initial Black Pearl rent does not itself prove the cap has lifted. Each credit event requires at least $3bn of defined fully diluted market capitalization. Liquidity includes qualifying unrestricted cash plus commitments less exposure; it is not identical to immediately drawable cash. [7, sections 1.01 and 4.02]

Quarter-end contractual liquidity floors are $100m, $150m and $200m as the two projects reach defined commencement, with equity cure rights. The $200m cash buffer used in our model is a separate analytical choice; crossing it is not a covenant-default finding. [6]

How project trouble could reach the parent

Project overruns → completion-guarantee funding demand → parent cash depletion → failed funding or guarantee payment → possible contractual remedies. A falling equity valuation could also restrict incremental revolver access. None of these conditions is established as occurring today.

The revolver’s Material Indebtedness definition excludes specified non-recourse project debt, but includes actually owed completion/carve-out claims more than 60 days past due, subject to the definition and $75m threshold. This is a concrete possible transmission route, not automatic consolidation of project defaults. [7, sections 1.01 and 7.01]

Who absorbs losses first

At a project, evaluate available reserves, tenant support, completion support and collateral together with their legal conditions. Parent equity in that project is residual. At the parent, unsecured recoveries depend on assets after secured claims and project liabilities. Do not count project collateral twice as both secured-note recovery and parent convertible recovery.

Barber Lake’s original indenture applies received lease-termination fees to note redemption and confines debt-service reserve use to debt service. Parent distributions therefore cannot be inferred from project NOI. The reserve formula changes over construction and with additional notes; current balances and amendments are required. [9, sections 3.11 and 4.22–4.23]

A valid credit-loss thesis needs either a parent liquidity failure after remedies or a project loss after enforceable support. An economic impairment of project equity is insufficient to establish either one.

Conditional parent cash flow through June 2027

This is a sensitivity model, not a GAAP forecast. Q3 2026 is already historical but unreported in the reviewed statements; it is modeled here. All spending amounts and timing below are analyst assumptions, not user-provided estimates or issuer forecasts. Opening cash is the rounded June parent/unrestricted proxy of $831.8m. [1, 2]

All cases assume $745.4m new financing in Q4, $750m incremental Barber Lake equity paid $450m in Q4 and $300m in Q1, none of that increment paid in Q3, no further financing and no revolver draws. If incremental equity was already funded, move it between quarters without charging it twice. Other net parent uses include overhead, working capital, parent debt costs, non-Barber contributions and net project distributions; project-funded capex and debt service are excluded.

Stress case cash bridge

Scenario comparison

Under stress, Q1 funding needed to restore the analytical buffer is $122.8m; zero-cash funding first arises in Q2 at $47.8m. Under severe assumptions, the Q1 zero-cash gap is $372.8m. Negative balances represent cumulative unmet funding, not feasible cash holdings. No case has an assigned probability.

Reverse stress: $831.8m + $745.4m − $750m = $827.2m before all other post-June net uses; preserving $200m leaves $627.2m of net-use capacity. Failure of the new financing reduces this capacity by $745.4m. Fresh equity, distributions, spending cuts and asset proceeds can change the result. Current cash and committed-use schedules are needed to replace assumptions.

Project protections and operating sensitivity

Amazon-backed projects weaken a broad AI-demand short

Black Pearl has a fifteen-year Amazon Data Services lease and Amazon.com support for rent and operating expenses. The disclosed construction allowance is $9.5m per IT MW, with tenant responsibility above it subject to exclusions such as unreasonable costs or landlord wrongdoing/negligence. The allowance includes an initial asset basis, so multiplying MW by the cap does not yield remaining cash capex. First rent began in August. [12, 18]

Stingray’s disclosed lease has a fifteen-year base term, Amazon support and a $10.5m per IT MW cost threshold, with initial rent targeted for April/May 2027. Its financing presentation uses a 1.25x DSCR-based amortization framework. Actual terms, exceptions and updated commencement control. [14, 15]

Inference: weaker AI utilization need not reduce collectible contractual rent if Amazon remains obligated and solvent. The relevant risks include delivery-related termination, uncovered costs, collection timing and refinancing. Cost protection may reduce ultimate losses while leaving an interim funding need.

Operating stress is not automatically a contractual loss

Calculations: Black Pearl uses unchanged $2bn principal × 6.125% plus $140m annual amortization; this deliberately ignores prior amortization and overstates interest relative to a declining balance. Stingray holds the original illustration’s $44m interest and $45m principal fixed. If its NOI-linked amortization adjusts, this fixed-payment stress overstates the deficit. NOI figures are pre-amendment illustrations, not refreshed projections. [11, 12, 14]

These haircuts require a reason rent becomes uncollectible despite the contracts; they cannot be inferred from lower tenant utilization. Debt-service coverage is also not equivalent to cash available for the parent after reserves, cash sweeps and distribution restrictions.

Barber Lake protection must be re-underwritten

The announced Google backstop expanded to approximately $1.73bn in November 2025 [10]. That is support for defined Fluidstack obligations, not all parent debt or all twenty years of amended rent. The September extension involves a leading AI lab; public summaries do not establish equivalent Google coverage for that extension [5]. Obtain the amended recognition agreement, effective dates, declining coverage schedule and termination exclusions.

A new completion budget must reconcile invoices already paid, remaining invoices, reserves by permitted use and equity funded to date. Total project cost minus original bond proceeds is not the remaining parent cash call.

Thesis verdict and instruments

Rank instruments by the risk they actually carry

First investigate Cipher Digital Inc. parent credit for completion-driven liquidity stress. The 0% October 2031 convertible is an existing unsecured parent obligation; its filed form identifies ISIN US17253JAB26, which must be matched to the currently trading line [16]. The new 2032 convertible, if closed, introduces a November 2030 cash put [4]. Both instruments embed equity optionality, so short returns require credit, stock and volatility analysis together.

Parent CDS would be a closer expression of a covered parent credit event if a dealer offers it. Public searches did not establish a Cipher single-name market, standard reference obligation or executable price. Absence of public evidence does not prove unavailability. A bespoke trade would require an eligible counterparty, legal terms, collateral and exit analysis.

Barber Lake project notes are a secondary candidate only if losses survive Google support, reserves and completion remedies. Black Pearl and Stingray rank lower for a generic AI-monetization short because of Amazon support. A short of any project bond needs issue verification, borrow availability, coupon/financing cost and recall terms. A secondary quote page was excluded from valuation because its issuance balance was stale and no firm market depth was supplied.

What would establish or reject the opportunity

Establish it: demonstrate unavoidable cash calls beyond usable liquidity before committed receipts; show why equity, asset sales, capex reductions and drawable debt cannot bridge them; map the resulting loss to the exact obligation; then price the position after carry, collateral and exit costs.

Reject or delay it: verified delivery within budget, timely collections and cost reimbursements, release of project distributions, sufficient net offering proceeds plus remaining cash, or protection pricing that already exceeds defensible expected losses. Do not force a negative-credit conclusion because construction costs rose.

Priority requests: current parent cash and LC utilization; financing closing and final fees; Barber equity paid versus remaining; project cost-to-complete and reserve accounts; all-phase commencement notices; current Google support; actual remaining principal schedules; live CDS terms or bond borrow. No counterparties were contacted.

Investment judgment: retain Cipher as a conditional parent-liquidity watchlist candidate. The evidence fits the construction-financing branch of the thesis more strongly than the demand-collapse branch, and currently supports further underwriting rather than a trade recommendation.

Sources and reproducibility

Primary documents were reviewed as of 16 October 2026. Historical disclosures are dated explicitly; no inaccessible exhibit, undisclosed current balance, dealer quote or closing confirmation is assumed. Calculations use rounded USD millions and are reproducible from the inputs on pages 2, 5 and 6.

[1] June 2026 10-Q: cash flow, commitments and debt

[2] August business presentation: capitalization

[3] October 13 presentation: Barber Lake cost bridge, slides 10–12

[4] October 14 convertible pricing and proposed exchange

[5] September 24 Barber Lake amendment disclosure

[6] March 23 revolver announcement

[7] March 23 revolving credit agreement

[8] Barber Lake notes and completion support

[9] Original Barber Lake indenture; amendments require reconciliation

[10] November 2025 Google backstop expansion

[11] Black Pearl notes issuance and amortization

[12] Black Pearl financing presentation and tenant protections

[13] Stingray issuance and completion guarantee

[14] Stingray financing presentation and illustrative cash flow

[15] Stingray August supplemental indenture

[16] Filed form of parent 0% convertible due 2031

[17] Issuer SEC filing index checked for closing confirmation

[18] Q2 release: early Black Pearl delivery and rent

NOI: net operating income. DSCR: debt-service coverage ratio. LC: letter of credit. EAC: estimated total cost at completion. CDS: credit default swap. A covered CDS event depends on its own contract and is not synonymous with a project covenant breach.