The $13 Trillion Door Opens as Private Credit Cracks: Whose Money Arrives Last?

The Probability Desk — Wednesday, June 3, 2026

Desk view in one paragraph. On June 1, 2026, the public comment period closed on the US Department of Labor's proposed rule to open the roughly $13 trillion defined-contribution retirement system — 401(k)s and their kin — to private equity, private credit, infrastructure and other alternatives. It is the procedural gateway to the largest new pool of retail capital private markets have ever been offered. It closed in the same fortnight that Fitch reported the US private-credit default rate hitting a record 6.0%, that Bank of America warned the three largest business-development companies sit roughly two quarters from an investment-grade-to-junk downgrade if redemptions persist, and that the Financial Stability Board told supervisors they cannot actually see inside a $2 trillion market. For a universal owner — a sovereign fund, a public pension, an insurer that is already one of the largest private-credit allocators on earth — this is not a product-launch story. It is a question about sequencing: does a wall of retirement money arrive to deepen and stabilise an asset class that is finally being tested, or does it arrive at the top, just as the first real stress cycle exposes the liquidity and valuation machinery underneath? We probability-weight four ways to year-end 2027: Managed Democratization (base, 45%), Regulatory Caution (15%), Validated Expansion (upside, 15%), and Stress Cycle Breaks (tail, 25%).

Editorial scenario analysis only. Not investment, actuarial, or geopolitical advice.

The Trigger

On June 1, 2026, the comment period closed on the Employee Benefits Security Administration's proposed rule — released March 30, 2026 — establishing process-based safe harbours for 401(k) plan fiduciaries who choose to include alternative assets such as private equity, private credit, real estate and infrastructure in defined-contribution menus. The rule is the regulatory machinery answering the August 2025 Executive Order, "Democratizing Access to Alternative Assets for 401(k) Investors," which directed federal agencies to widen DC-plan access to non-publicly-traded investments. A final rule could be published by year-end, with implementation more likely in 2027. (Sources: US Department of Labor / EBSA news release, March 30, 2026; CNBC, March 30, 2026; Morrison Foerster and Shulman Rogers legal alerts, April 2026; FactSet, 2026.)

The number that makes this systemic is the size of the door. US defined-contribution plans hold on the order of $13 trillion. Even a low-single-digit allocation — the kind a target-date fund might route through a small, liquidity-managed sleeve — is a multi-hundred-billion-dollar demand curve aimed at asset classes that today raise capital primarily from institutions. Private-credit AUM is approximately $2 trillion in 2026 and is projected by Moody's to approach $4 trillion by 2030 (Moody's Private Credit Outlook, 2026; AIMA, 2026). A retail/retirement bid of that scale would not be a marginal flow into private credit; it would be a regime change in who funds it.

Why it lands on a decade-horizon balance sheet: the universal owner is already the incumbent lender. Public pensions and insurers are among the largest private-credit allocators in the world; UK and US defined-benefit schemes have become, in the FSB's words, a "major source of funding" for the market, and insurers affiliated with private-equity firms hold an estimated $1 trillion in assets sourced through those relationships (FSB, May 6, 2026; Pensions Age, 2026; Insurance Business, April 2026). When the marginal buyer of an asset class changes from a patient institution to a daily-priced retirement saver, the liquidity profile of the whole asset class changes — and the institution that got there first owns the consequences of that transition whether or not it participates in the new flows. That is the single most important capital-structure fact of the week.

The Forecast Question

Through December 31, 2027, does the democratization of private markets into US defined-contribution plans proceed into a finalized rule with material DC inflows — and does private credit clear its first real stress cycle without a systemic liquidity or valuation event?

We define four mutually exclusive end-states by two resolving axes — the policy/adoption axis (does the DC door actually open, and does money walk through it) and the credit-stress axis (does the current default-and-redemption cycle stay idiosyncratic or turn systemic):

Does the rule finalize in usable form and DC money begin flowing at material but throttled scale, while private-credit stress stays contained and idiosyncratic? → Managed Democratization.
Does the rule narrow, stall, or get tied up such that DC adoption is negligible through 2027, with private credit remaining an institutional asset class? → Regulatory Caution.
Do defaults peak and recede, the asset class proves resilient, and DC flows become a structural new demand base that rewards the incumbent managers? → Validated Expansion.
Does a systemic event — a downgrade cascade through the largest BDCs, widespread redemption gating, a valuation/mark scandal, or defaults pressing toward 8–10% — hit before or during the DC rollout, freezing the democratization narrative and repricing the complex? → Stress Cycle Breaks.

Resolution criteria, horizon, and tripwires below are concrete enough to grade in the calibration log when the window closes.

Prior / Base Rate — Three Analogues

No reference class fits perfectly, but three liquidity-transition episodes are instructive, and all three caution the same way: the asset tends to survive; the vehicle and the financing structure are where the break happens, and the last money in absorbs it.

Third Avenue Focused Credit, December 2015. An open-end mutual fund holding illiquid high-yield and distressed credit suspended redemptions and moved to liquidate when investors rushed the exit faster than the underlying could be sold. It was a small fund, but it crystallised the core hazard of offering daily or near-daily liquidity against assets that do not trade daily. Lesson: the liquidity-mismatch break is a structural feature of the wrapper, not a function of the credit's ultimate quality. Why it might mislead: it was a single, leveraged, distressed-tilted fund, not a diversified direct-lending complex.

BREIT, November 2022 – February 2024. Blackstone's non-traded real-estate vehicle hit its redemption limits in November 2022 as savers rushed the gate; requests peaked at roughly $5.3 billion in January 2023; the gate held, the sponsor managed the queue, and by February 2024 the fund was again meeting 100% of redemptions. Lesson, and it cuts both ways: a well-designed gate worked — it prevented a fire sale and the structure survived — but it also showed that the moment retail wants out of an illiquid wrapper, the gate is the product, and "permanent capital" is only permanent until it is tested. Why it might mislead: real estate is not direct lending, and BREIT's sponsor had the balance sheet to backstop confidence.

The 2014–2016 retailisation of leveraged credit. Bank-loan and high-yield retail funds drew enormous flows in the search for yield, then faced sharp outflow and liquidity-discount episodes when the cycle turned. Lesson: democratising a credit asset class amplifies the procyclicality of its flows — retail buys late and sells fast — so the same money that deepens the market in calm widens the dislocation in stress.

The base-rate read: liquidity-mismatch events in retail-facing illiquid-credit vehicles are recurrent and survivable but rarely systemic — every analogue was contained without broad contagion, yet every analogue also inflicted real losses on the latest entrants and forced gates. That history pulls our systemic-tail weight up relative to a benign-conditions extrapolation, but stops short of a base-case crisis: the modal outcome in the reference class is a contained, idiosyncratic stress that the structure absorbs.

Source Ledger (selected — 28 named, dated sources)

#	Source	Date	Data point used	Conf.	Moves model?
1	US DOL / EBSA news release	Mar 30 2026	Proposed rule; safe harbours for alts in 401(k)s	H	Yes (policy axis)
2	CNBC	Mar 30 2026	DC market ~$13T; comment period through Jun 1 2026	H	Yes
3	Morrison Foerster legal alert	Apr 3 2026	Rule scope: PE, private credit, RE, infra	H	Yes
4	White House EO	Aug 2025	"Democratizing Access to Alternative Assets for 401(k) Investors"	H	Context
5	Fitch Ratings (via press)	May 2026	US private-credit default rate record 6.0% (Apr); 5.7% Mar	H	Yes (stress axis)
6	Fitch Ratings	2025–26	PC-backed corporate borrowers 9.2% default in 2025; <$25m EBITDA 15.8%	M	Yes
7	Financial Stability Board	May 6 2026	Report on vulnerabilities in private credit; supervisory data gaps; $1.5–2T	H	Yes
8	Bank of America (via Bloomberg)	May 29 2026	3 largest BDCs ~2 quarters from IG→junk if 5%+ net redemptions persist	H	Yes
9	BofA (via PitchBook)	May 2026	BDC redemptions likely peak Q2 2026, ease into Q4, stay >5% all year	H	Yes
10	Moody's Private Credit Outlook	2026	PC AUM >$2T in 2026, toward ~$4T by 2030; defaults "easing, fragmented"	H	Yes
11	AIMA press release	2026	Private-credit market reaches ~$3.5T (broad definition)	M	Context
12	FinancialContent / market wires	Apr 6 2026	BDC capital formation / retail sales down ~40% YoY	M	Yes
13	Insurance Business	Apr 2026	Insurers among largest PC allocators; PE-affiliated insurers ~$1T assets	M	Yes
14	Pensions Age	2026	DB schemes "major source of funding" for private credit	M	Yes
15	Quinn Emanuel client alert	2026	Private-credit-under-stress litigation risk; manager-driven valuation	M	Yes (tail)
16	FRED DCOILBRENTEU	May 26 2026	Brent $102.75 (macro context)	H	Context
17	FRED DGS10	Jun 1 2026	US 10-yr 4.47%	H	Yes (discount rate)
18	FRED DGS2 / DGS30	Jun 1 2026	2-yr 4.05%; 30-yr 4.99%	H	Yes
19	FRED BAMLC0A0CM	Jun 2 2026	IG OAS 0.74% (benign credit conditions)	H	Yes
20	FRED VIXCLS	Jun 2 2026	VIX 15.77 (low realized stress)	H	Yes
21	Federal Reserve H.15	Jun 1 2026	Fed funds 3.50–3.75%; hawkish hold; FOMC Jun 16–17	H	Yes
22	SEC EDGAR (BDC 10-Qs)	Q1 FY2026	Non-traded BDC redemption and NAV disclosures	M	Yes
23	Pensions & Investments / Commercial Observer	2022–24	BREIT gate Nov 2022; peak Jan 2023 $5.3bn; 100% met Feb 2024	H	Yes (analogue)
24	Reuters / press (Third Avenue)	Dec 2015	Focused Credit Fund redemption suspension & liquidation	H	Yes (analogue)
25	CAIA Portfolio for the Future	Apr 2026	Redemptions, defaults, wrappers; retail-heavy vehicle fragility	M	Yes
26	The Asian Banker (FSB coverage)	May 2026	FSB flags supervisory data gaps across $2T market	M	Context
27	gomarkets / market commentary	Jun 2026	June macro drivers: inflation, rates, policy credibility	L	Context
28	Global SWF / S&P Global	2026	SWF private-market deal activity up; fixed-income share falling	M	Context

Confidence: H/M/L. Where two figures circulate for the same metric (e.g., the 6.0% headline Fitch rate vs. the 9.2% 2025 cohort default rate vs. the ~$2T vs. ~$3.5T market-size figures), both definitions are stated and dated rather than blended.

Key Data Table

Variable	Current reading	Prior / comparison	Direction	Source
US private-credit default rate	6.0% (Apr 2026, record)	5.7% Mar 2026; 5.8% Apr 2025	↑	Fitch
PC-backed borrowers 2025 cohort default	9.2% (2025)	<$25m EBITDA: 15.8%	↑	Fitch
Private-credit AUM	~$2.0T (2026)	~$1.7T (mid-2023)	↑	Moody's / Fed
US DC retirement assets	~$13T	—	flat–↑	DOL/CNBC
Largest BDCs to downgrade	~2 quarters (if 5%+ redemptions)	median non-traded BDC ~1 yr runway	↓ (risk ↑)	BofA
BDC retail capital formation	−40% YoY (Q1 2026)	sharpest contraction on record	↓	market wires
IG corporate OAS	0.74% (Jun 2 2026)	benign / tight	flat	FRED BAMLC0A0CM
VIX	15.77 (Jun 2 2026)	low	flat	FRED VIXCLS
US 10-yr Treasury	4.47% (Jun 1 2026)	2-yr 4.05%; 30-yr 4.99%	curve re-steepening	FRED
Fed funds target	3.50–3.75%	hawkish hold; FOMC Jun 16–17	flat	Fed H.15
Brent (macro context)	$102.75 (May 26 2026)	elevated post-Hormuz	↑	FRED

The picture the table draws is the tension at the centre of this scenario: a credit-stress signal flashing amber (record defaults, BDC downgrade risk, 40% collapse in retail BDC sales) against a market-conditions signal still flashing green (IG spreads at 0.74%, VIX below 16, no priced stress). The policy door is opening into exactly that contradiction.

Evidence-Update Table (Bayesian)

We start from a base-rate prior, drawn from the analogues, that a retail-facing illiquid-credit transition produces a contained, idiosyncratic stress that the structure absorbs over a ~19-month horizon — roughly a 50% modal "managed" outcome, with the residual split across stall, validation, and a systemic break. Each evidence item moves a specific posterior; no weight moves without the evidence beside it.

Evidence (dated, sourced)	Direction	Strength	Effect on posteriors
DOL/EBSA rule proposed Mar 30 2026, comment closed Jun 1 2026; final possible year-end (DOL, CNBC)	The door is genuinely opening; timeline real but not instant	High	Base ↑, Stall ↓
DC market ~$13T vs PC AUM ~$2T (CNBC; Moody's)	Even small allocation = regime-change demand	High	Base ↑, Upside ↑
Fitch PC default rate record 6.0% Apr 2026 (Fitch)	First real stress cycle is underway, not hypothetical	High	Tail ↑, Upside ↓
BofA: 3 largest BDCs ~2 quarters from junk if 5% redemptions persist (Bloomberg, May 29 2026)	Downgrade risk concentrated in the biggest, most-owned vehicles	High	Tail ↑
BofA: redemptions peak Q2 2026, ease into Q4 (PitchBook)	Stress may be cresting, not accelerating	Medium	Tail ↓, Base ↑
BDC retail capital formation −40% YoY (market wires)	Retail is already leaving before the DC door even opens	Medium–High	Stall ↑, Tail ↑
FSB: supervisors lack visibility into $2T market (FSB, May 6 2026)	Hidden leverage/interconnection raises blow-up severity, not timing	High	Tail severity ↑
Insurers ~$1T PE-affiliated assets; DB schemes major funders (Insurance Business; Pensions Age)	Losses transmit to pensions/annuitants — universal-owner channel	High	Tail systemic-reach ↑
Moody's: defaults "easing, fragmented not systemic" (Moody's 2026)	Distress is dispersed, not a single correlated shock	Medium	Tail ↓
IG OAS 0.74%, VIX 15.77, Jun 2 2026 (FRED)	No stress priced; financing cheap; calm regime	High	Tail near-term ↓, Base ↑
Litigation risk from manager-driven valuation (Quinn Emanuel 2026)	Mark-to-myth valuations are a latent trigger	Medium	Tail ↑
BREIT gate worked (2022–24); Third Avenue liquidated (2015) (P&I; Reuters)	Gates contain fire sales but confirm liquidity mismatch is real	High	Base ↑ (containment), Tail structural-floor

Net of the update: the secured-but-gradual policy timeline, the sheer scale of the DC pool, the benign current credit conditions, and the analogue record of containment hold a "managed" base case as the plurality outcome; the record default rate, the concentration of downgrade risk in the largest and most widely held BDCs, the hidden leverage the FSB cannot see, and the insurer/pension transmission channel fatten the tail well above a daily norm. Posterior, before reconciliation with the simulation: a base case in the 45–50% range and a systemic tail in the 20–28% range — materially elevated.

The Scenarios (to year-end 2027)

Base — Managed Democratization · 45%. The rule is finalized in a cautious form during 2027; plan sponsors adopt slowly, mostly through target-date funds with small (low-single-digit), liquidity-managed private-markets sleeves rather than standalone options. DC inflows into private markets reach a material but throttled scale (tens of billions cumulatively, not the full theoretical wall). Private-credit stress stays idiosyncratic: defaults grind around the 6–7% area, the weakest BDCs see downgrades or gating but the complex holds, and incumbents absorb the cycle. Resolves base if: a final rule is in force, identifiable DC private-markets flows are underway at modest scale, and no systemic event (defined in the tail) occurs by end-2027. Tripwire toward this: a final rule with conservative liquidity/valuation safe-harbours plus the first large recordkeeper/target-date launches.

Regulatory Caution · 15%. The rule narrows materially or stalls — fiduciary-liability fears, litigation, a change in administration posture, or a high-profile early loss freeze plan-sponsor appetite — and DC adoption is negligible through 2027. Private credit remains an institutional asset class. Not a crisis; a chosen plateau on the retail side. Resolves if: no final usable rule, or a final rule that plan sponsors broadly decline to adopt, with cumulative DC private-markets flow below roughly $25bn. Tripwire: a withdrawn/substantially-weakened rule, or major recordkeepers publicly declining to offer alternatives on liability grounds.

Upside — Validated Expansion · 15%. Defaults peak in 2026 and recede through 2027, the asset class proves its resilience through the stress test, and the DC door opens into a validated product — flows become a structural new demand base that compresses spreads, rewards the large diversified managers, and deepens market liquidity. Resolves if: the private-credit default rate falls back below ~6%, no systemic event, and DC private-markets flows exceed roughly $130bn cumulatively with broad sponsor adoption. Tripwire: two consecutive quarters of falling default rates alongside multiple large target-date private-markets launches.

Tail — Stress Cycle Breaks · 25%. A systemic private-credit event hits before or during the DC rollout: a downgrade cascade through one or more of the largest BDCs, widespread redemption gating, a valuation/mark scandal that forces broad write-downs, or defaults pressing toward 8–10%. The democratization narrative freezes; regulators pause or reverse; the listed alternative-asset managers and the credit complex reprice; and losses transmit through the insurer/pension channel to the universal owner and, ultimately, to retail savers who arrived last. Resolves if any of: an IG→junk downgrade of a top-three BDC's unsecured debt; gating at multiple large retail-facing vehicles simultaneously; the private-credit default rate exceeding 8%; or a forced multi-manager NAV write-down event. Tripwire: a single top-BDC downgrade or a marquee fund suspending redemptions. The 25% weight — well above our daily norm — reflects the record default rate, the concentration of downgrade risk in the most widely held vehicles, the leverage the FSB cannot measure, and the procyclical retail flows, set against benign current conditions, cresting-redemption signals, and the analogue record of containment.

Weights sum to 100%. They are an ensemble (below), not the raw simulation output.

The Monte Carlo

We ran a real simulation (mc.py, numpy, 50,000 paths, seed 20260603) of two correlated drivers over the horizon to year-end 2027 — the peak private-credit default rate and cumulative DC inflows into private markets — plus a stress branch and a policy-adoption branch, then classified each path into the four end-states by the resolving thresholds above.

Design. The default-rate driver is anchored at the current record 6.0% (Fitch, April 2026), with a gentle base drift and a large additive jump on recession paths (recession probability 25% over the horizon). Redemption pressure at retail-heavy vehicles is modelled as correlated with the default rate (each point of default above 6% adds ~1.3 points of quarterly redemption pressure), with BofA's ~5% downgrade-trigger line as the danger zone. A systemic stress event fires when defaults and sustained redemptions both clear cascade thresholds, or on an outright default blowout (≥9%). Policy adoption is a mixture: the rule finalizes usefully on ~72% of paths (driving real but uneven uptake) and stalls on the rest (token sleeves only); a systemic event dampens DC flows by 60% (the democratization narrative freezes). Anchors and distributions are documented in the code header; this is a scenario model, not a forecast of a single price.

Percentile outputs (to year-end 2027):

Metric	P10	P25	P50	P75	P90
Peak private-credit default rate (%)	5.6	6.2	7.0	8.2	10.6
Quarterly redemption pressure (%)	3.0	4.1	5.4	7.4	10.1
Cumulative DC inflows (USD bn)	9	26	65	122	156

Raw simulation classification: Managed Democratization 44%, Regulatory Caution 15%, Validated Expansion 13%, Stress Cycle Breaks 28%.

Threshold probabilities: P(peak default ≥ 8%) 27%; P(peak default ≥ 10%) 14%; P(defaults recede below 6%) 16%; P(systemic stress event) 28%; P(redemptions sustained > 5%) 58%; P(rule finalizes) 72%; P(material DC adoption > $80bn) 44%; P(adoption stalls < $25bn) 25%.

Reconciliation to the published ensemble. We soften the simulation's mechanical 28% systemic tail to 25% in the published weights, moving the difference to the base case (44% → 45%), for three logged reasons: (1) current credit conditions are benign and unpriced for stress — IG OAS 0.74%, VIX 15.77 on June 2, 2026 — which the model's recession-jump deliberately under-weights in the near term; (2) BofA's own read is that redemptions peak in Q2 2026 and ease thereafter, confining downgrade risk to the weakest BDCs rather than the complex; and (3) every analogue in the reference class (Third Avenue 2015, BREIT 2022–24) was contained without systemic contagion. The published tail of 25% therefore remains materially above the daily norm and close to the model, while crediting the containment base rate. Upside is nudged to 15% (from the model's 13%) within rounding to reflect the same benign-conditions case.

Market vs Desk View

What the market is pricing: very little stress. Investment-grade spreads at 0.74% and the VIX below 16 say the credit market is treating private-credit wobbles as idiosyncratic noise, not a systemic signal. Listed alternative-asset managers and BDC equity have de-rated from their peaks (BDC retail capital formation is down 40% year-on-year), but unsecured BDC bonds still trade at investment grade — the very downgrade BofA flags has not yet been priced. The consensus, echoing Moody's, is "defaults easing, fragmented, not systemic."

The Desk view: the market is probably right about timing and may be under-weighting severity and transmission. The near-term calm is real — our own tail softening leans on it. But three things look under-appreciated. First, the downgrade risk is concentrated in the largest, most widely owned vehicles, so a single top-three BDC downgrade would not be idiosyncratic — it would re-rate the benchmark. Second, the FSB's central finding is that supervisors cannot see the leverage and interconnection, which means the severity distribution has a fatter, less observable right tail than spreads imply. Third, and most specific to this audience: the policy door is opening procyclically — it invites the latest, least-informed capital into the asset class precisely as the cycle is being tested, and that money is the most likely to gate and sell at the bottom. Our highest-conviction mispricing is therefore not direction but convexity: the cost of tail protection on the credit complex looks cheap relative to a 25% systemic-break probability and an FSB-flagged unobservable severity.

Universal-Owner Portfolio Heatmap

Direction of the reprice by asset class, with magnitude band, across the scenarios (3-month / 12-month / 5-year horizon emphasis noted):

Asset class	Base (45%)	Upside (15%)	Tail (25%)
Private credit / direct lending (held)	Flat-to-mild markdown; dispersion ↑	Spreads compress, marks hold (+)	Sharp markdown, gating, −10 to −25% NAV stress
Listed alt-asset managers (Apollo/Blackstone/Ares/Blue Owl)	Range-bound, fee growth intact	Re-rate higher on DC demand (++)	De-rate hard (−20 to −40%)
BDC unsecured bonds	Stable IG, weakest names drift wider	Tighten (+)	Downgrade cascade, IG→junk on weakest
Insurance / annuity balance sheets	Earnings drag from credit costs	Benign	Capital strain via PE-affiliated assets
Public IG credit	Tight, benign	Tight	Modest contagion widening (+25–75bp)
High yield / leveraged loans	Stable	Tighten	Widen, cross-asset credit repricing
Public equities (broad)	Neutral	Mild +	Risk-off drawdown, financials-led
Rates / long duration	Curve re-steepening continues	Mild bear	Bull-flatten on flight-to-quality
Real assets / infrastructure	Stable; alt DC sleeve tailwind	+ (DC infra demand)	Liquidity discount on non-traded RE
Gold / reserves	Neutral	Neutral	Bid as hedge
Cash / liquidity buffers	Neutral	Slight drag (opportunity cost)	Premium — the asset that matters in a gate

The strategic read for a universal owner: the exposure that matters is not the direct-lending book in isolation but the correlation between it, the insurer/annuity book, the listed-manager equity, and — increasingly — core IG credit benchmarks that now carry more of this risk than they did a year ago. The tail is a correlated event across several line items that are managed as if independent.

Second- and Third-Order Effects

A democratization-meets-stress dynamic propagates well beyond the BDC tape. First order: if the door opens cleanly, the marginal buyer of private credit shifts from patient institutions to daily-valued retirement savers, structurally lowering the asset class's cost of capital but raising the procyclicality of its funding. Second order: managers redesign products for the DC channel — semi-liquid, gated, interval structures — which embed a gate as a standard feature and normalise "permanent-until-tested" capital; valuation practices come under regulatory and litigation scrutiny precisely because retail marks are now in 401(k) statements. Third order: the insurer-pension-retail transmission chain means a private-credit drawdown no longer stays in alternatives — it lands on annuity balance sheets, on DB funding ratios, and, post-rule, on DC account statements, turning a market event into a retirement-policy event with political consequences that could reverse the very democratization that created the exposure. For sovereign and pension allocators outside the US, the read-through is the global one: the same retailisation playbook is being exported, and the question of who provides the system's liquidity in a gate is becoming a portfolio-construction question, not a footnote.

Watch Dashboard

#	Indicator	Threshold that changes the model	Source
1	DOL final rule status	Finalized usable rule → Base/Upside; withdrawn/narrowed → Stall	DOL/EBSA
2	First large target-date / recordkeeper alts launch	Yes → Base/Upside; broad refusal → Stall	recordkeeper filings
3	Fitch private-credit default rate	>8% → Tail; <6% → Upside	Fitch
4	Top-3 BDC unsecured rating	Any IG→junk downgrade → Tail	Moody's/S&P/Fitch
5	BDC quarterly net redemptions	Sustained >5% into Q4 2026 → Tail risk; <5% → Base	BofA / BDC 10-Qs
6	Multi-vehicle gating events	≥2 large vehicles gate simultaneously → Tail	press / filings
7	IG corporate OAS	>50bp widening → stress confirming	FRED BAMLC0A0CM
8	VIX	Sustained >25 → regime shift	FRED VIXCLS
9	Private-credit AUM trajectory	Stalls/contracts → Stall; accelerates → Base/Upside	Moody's/Preqin
10	BDC retail capital formation YoY	Recovers from −40% → Base/Upside; deepens → Tail	market wires
11	Insurer PC valuation write-downs	Cluster of write-downs → Tail transmission	NAIC / filings
12	Litigation / valuation disputes	A marquee mark scandal → Tail trigger	court dockets
13	Fed funds path / FOMC Jun 16–17	Cuts ease refinancing stress (Base); higher-for-longer (Tail risk)	Fed
14	10-yr / curve shape	Sharp bull-flatten → flight-to-quality (Tail underway)	FRED DGS10
15	FSB / regulatory capital action	New CLO/affiliated-insurer capital rules → adoption friction	FSB / NAIC

Red-Team — How This Could Be Wrong

The tail is too high. The strongest counter-argument: current conditions are benign and improving — IG spreads at 0.74%, VIX under 16, BofA expecting redemptions to crest in Q2 and Moody's calling defaults "easing." A 25% systemic-break probability over 19 months may over-weight a stress cycle that is already cresting; if redemptions ease as BofA expects and the rule finalizes cleanly, Base or even Upside dominates and 25% will look alarmist. What would falsify our tail weight: the default rate rolling over below 6% and BDC redemptions falling under 5% by Q4 2026.

The policy timeline is slower and smaller than the headline. "Democratization" could prove almost entirely theoretical through 2027 — fiduciary liability is a powerful brake, target-date sleeves are tiny, and recordkeepers move slowly. If so, the whole "wall of retail money" framing is premature; the real story is just a normal credit cycle and Stall is underweighted at 15%. What would falsify the Base/Upside policy assumption: major recordkeepers publicly declining to offer alternatives on liability grounds.

The base rate may be the wrong reference class. We anchored on liquidity-mismatch episodes (Third Avenue, BREIT) that were contained. But private credit at $2T, embedded in insurer and pension balance sheets and soon in 401(k)s, is a categorically larger and more interconnected system than any analogue — the FSB's whole point. If the right reference class is "systemic shadow-banking growth," the containment base rate is too reassuring and the tail is, if anything, too low. The honest position: our analogues argue for containment; the structural scale argues against it; we split the difference and weight the tail at 25%, and we would revise up on the first sign that a single name's distress is correlating across the complex.

Specific falsifiers for the whole call: a clean final rule plus receding defaults (→ Upside); a withdrawn rule (→ Stall); or a top-three BDC downgrade (→ Tail). Each is observable on the dashboard above, which is the point — this scenario is built to be graded.

Methodology Box

The Probability Desk produces probability-weighted scenarios, not point forecasts. Weights are an ensemble of a stated historical base rate, dated and attributed expert priors, and a real Monte Carlo simulation (documented in mc.py), reconciled by a logged aggregation rule and sense-checked against the base rate. No probability is published without the evidence that moved it; no source, quote, number, or expert view is invented; live macro figures are pulled from FRED and cited with their dates. Every scenario carries a resolvable outcome, a horizon, and observable tripwires so the call can be graded later in our public calibration log. The Desk owns the number; the simulation engine is a tool, never the predictor.