One model, five independent predictors, six panels. Bottom line up front: a tightening squeeze most likely resolving in a negotiated settlement across 2026–2027 — not a sudden collapse. v3.32: the v3.28 reverse-trigger fired: the US-Iran Hormuz deal collapsed the oil premium (Urals ~$42 in early July, back below the $59 cutoff). The strike campaign has hit all 11 of Russia's largest refineries (Syzran re-struck Jul 12), refining runs have broken through the 2005 floor (3.91M b/d, Bloomberg), a diesel export ban is now in force, the fuel crisis is functionally nationwide (Crimea: total civilian sales ban) and is transmitting into record wartime food-price inflation and a diesel-strained harvest, Ukraine opened a direct campaign on the shadow fleet (Kerch Strait transit halted), NATO pledged €140B for 2026-27, and insider reporting points to a possible October mobilization. No kill-criterion has fired; base case unchanged, but the sub-$40 oil criterion is the closest it has ever been and the test window compresses to Aug–Oct 2026. Navigate the evidence below.
The defensible window for reserve exhaustion is late 2026–2028, oil-price dependent; at the current ~$42 Urals path it narrows toward H2 2027. Ukraine's funded clock runs to end-2027 (EU loan + NATO Ankara pledge); Russia's decision point compresses to Aug–Oct 2026 (July oil crash hits the budget in August; harvest and winter sowing run on scarce diesel Aug–Oct; Duma elections Sept 18–20; insiders name October for a possible mobilization). No method points to sudden collapse.
An empty reserve fund is not bankruptcy, and bankruptcy is not surrender. Even at zero, the deficit can be covered by borrowing and money-printing — uglier, survivable for years. "Russia is about to run out of money" has been forecast every year since 2022 and been wrong every time.
Reserves and deposits are cushions, not off-switches. They matter for what they force — visible cuts to pensions, wages, regions — not because hitting zero ends the war by arithmetic.
All five independent methods point the same direction — a tightening squeeze resolving in a negotiated settlement, not collapse:
Directional agreement: 5 of 5 methods.
Deliberately no single collapse probability or numeric confidence score is given — the methods agree on shape and direction, not on a calibrated percentage. Quantifying it would be false precision.
How to read this dossier: each tab is one independent line of evidence. Confidence comes from their convergence — five methods sharing no inputs all describe the same shape. Where they'd diverge is where manipulation hides. See Method & Cadence for how often to refresh it.
The liquid National Wealth Fund — the state's spendable gold + FX — has lost roughly half its value, and its runway is almost entirely a function of the oil price.
Liquid NWF, US$ billions. The 2025→26 uptick is misleading — gold revaluation + a temporary Iran-war oil spike, not recovery. The Jul 26 point (~$46B, Finance Ministry) reflects the brief June windfall replenishment, the first refill in nearly a year; the July oil crash hits the budget from August.
Projected NWF from Q2 2026 (Gazprombank scenario bands). At $50/bbl ~2.5 yrs; at $40 just over a year; at $30–35 empties by end-2026. Curves bend, not crash — VAT hiked, defense cut 4%, regions squeezed.
Urals (tax-reference) vs the $59 fiscal cutoff (dashed). Winter months below cutoff were financed from reserves; the Iran-war Hormuz crisis flipped Urals above cutoff from March (peak ~$95 in May); the US-Iran Hormuz deal then deflated it to $60.92 in June, and early-July averages (~$42, Argus) are back below the cutoff. The July bar will be added when the month's tax figure settles; per series discipline only settled monthly figures are charted. See the v3.30 log entries below.
Russia's central bank sold ~22 tons of physical gold in early 2026 — reserves fell a fourth straight month, the largest sell-off since 2002 (~3.5× the prior record) — to cover a Q1 deficit of 4.6T rubles (~$61B), driven by Urals at ~$39 vs the $59 budget assumption. The NWF began 2026 with ~4.1T rubles (~$53B) liquid; VTB estimates up to ~60% of remaining liquid reserves may be spent in 2026.
Verification: anchored to Russian Central Bank data via Moscow Times; corroborated across United24, NV (T1/T2). Submitted source was a Ukrainian outlet retelling (T3) — accurate on facts. Nuance: one analysis frames the gold sale partly as a yuan-liquidity settlement fix, not pure distress — both can hold.
Verdict: fires the soft "Urals broke below band" signal → shifts estimate toward the earlier end of the late-2026–2028 window. Does NOT trip a hard kill-criterion (no NWF-rule change, mobilization, or ceasefire). Base case unchanged — reserve clock draining faster = austerity mode, exactly as modeled, not collapse.
M2 money supply has roughly doubled since the 2022 invasion (from ~60T rubles), confirming the "print to survive" mechanism. The tell on real inflation: the Central Bank's own surveys show perceived inflation at 15.1% against an official 5.5%. Opposition economist V. Milov (via Nacke) argues the flat official inflation reflects demand destruction from high rates, not control — plus a private-sector cash-flow crunch (cited Vedomosti survey: 53% of firms reporting liquidity gaps) and ~25–30% loss-making "zombie" firms. This is the inside view of the austerity mode the model predicts.
Verification: M2 doubling + level (~130.2T, Apr 1) and the strong-ruble exporter complaints confirmed at T1 (Bank of Russia, Trading Economics). Important counter-detail: M2 is now falling — down 0.6% in March, 1.4% in January, annual growth slowing to ~10–12% — which cuts against Milov's "ongoing explosive injection" framing. The cash-flow-gap survey, zombie-firm %, steel slump and "Power of Siberia 2 is dead" are T3 opposition analysis — plausible, internally coherent, not yet corroborated to T1/T2.
Verdict: first update to add structural texture rather than accelerate a clock — deepens confidence in the squeeze mechanism. But a liquidity crunch and zombie-firm overhang are what "uglier, survivable for years" looks like from inside. No hard kill-criterion. Base case unchanged.
The civilian-economy dimension of the squeeze, now with hard anchors. A state agency (Dom.RF) projects 2026 new-housing sales down ~15% by volume; Deputy PM Khusnullin says the sector exhausted its pre-2024 buffer and the builders' association warns 1-in-5 developers risks bankruptcy. The concrete trigger: Samolet, a top-2 developer, is on the brink and has asked the state for 50B rubles to repay bank loans — the exact bank-absorbs-toxic-developer-assets mechanism economist I. Lipsits describes. Systemic bad debt is rising: a Kremlin-aligned center (TsMAKP) puts troubled assets above 10% of total banking assets; the CBR separately estimates ~10T rubles (~$129B) of bad loans. Connects to the war-chest panel: Lipsits names "forced deposit-to-bond conversion" as Russia's alternative to negotiating — which is precisely the soft-confiscation lever this model already mapped.
Verification: housing contraction confirmed T1/T2 (Dom.RF, Moscow Times 25 May, Khusnullin); systemic bad-loan rise confirmed T2 (TsMAKP, CBR); Samolet distress T2. Tiering discipline: Lipsits's sharper specifics (−28.2% completions, −37% Moscow sales, 18.9% SME-loan figure attributed to CBR) logged as his T3 claims — same direction, not independently pinned. His "hyperinflation >50%" and "military communism" outcomes are explicitly his forecast scenarios, not current readings. Dissent retained: a 2025 RIDDLE analysis argues large-scale bankruptcies and panic price-drops have so far been avoided and the "investment-property overhang" overestimated — truth likely between crisis and managed-slowdown. China "Panda-bond debt trap" is plausible T3 inference; Power of Siberia 2 stall is real and consistent.
Key framing: "military communism" is not a collapse signal — it's an independent economist arriving at the same soft-confiscation survival path the model contains. A deteriorating civilian economy is what survivable austerity looks like from inside. No hard kill-criterion. Base case unchanged.
Deliberately logging a datapoint that points the opposite way from the recent run (refining lows, Belarus imports, Crimea coupons). On June 1, oil surged after Iran's Tasnim agency reported Tehran suspended US communications and is weighing a full Strait of Hormuz closure: Brent jumped ~6–7% intraday, trading above $95; WTI ~$92. Why it matters here: oil price is a direct input to the reserve clock. A Middle East supply shock that lifts global benchmarks drags Urals up with it — widening Russian revenue even at a discount, and partially offsetting the squeeze the last several entries documented. This is the single most powerful variable that can swing Russia's way.
Verification: spike confirmed at T1/T2 (Bloomberg via Korrespondent, OilPrice +6.62% Brent, Trading Economics above $95; CIBC's Babin quoted accurately). Three caveats that keep this from being over-read as Russian relief: (1) Spike within a downtrend — Brent fell ~19% in May (worst month since COVID), down ~20% from 2026 peaks on ceasefire optimism; this is a bounce, not a new sustained high. (2) Already partially reversing — Trump said talks continue and a Hormuz-reopening MOU could come within a week; prices pared from the 7% high. (3) Crude ≠ refining — higher global crude helps Russia's budget-revenue lever but does not fix its domestic fuel shortage (refining strikes, fuel imports); it helps one Russian problem while not touching the other.
Verdict: the rare update pointing toward relief for Russia, logged precisely because it cuts against the hopeful direction and the discipline requires surfacing it as carefully as pressure data. A sustained Hormuz closure would be a real partial offset to the squeeze. But it's a volatile spike inside a downtrend, half-reversed already, touching only the revenue lever. Base case unchanged — though this is the live reminder that the oil-price variable can swing Russia's way fast, and is worth watching as closely as the strike campaign.
Verifying a submitted June-3 US-Iran clip pulled a thread the model had under-weighted. The v3.21 entry framed the oil-price move as "a volatile spike inside a downtrend, half-reversed already." The fuller picture is that the 2026 Iran war (from 28 Feb) and the Strait-of-Hormuz crisis have kept Russian crude elevated for months, not days. For tax purposes — the figure that actually drives budget revenue — average Urals was ~$77 in March and $94.87 in May, the highest since October 2023; spot hit ~$116 at Primorsk on 2 April, a 13-year high and nearly double the $59 budget assumption, with the usual Urals-to-Brent discount briefly collapsing toward parity as Asian buyers cut off from Gulf barrels turned to Russian crude. Russia's main oil tax revenue roughly doubled to ~$9B in April; the FT called Russia the war's "biggest winner." This is the single most powerful variable that can swing Russia's way, and it swung.
Correction applied (with your sign-off): the Urals chart series — previously trailing off at ~$60 for spring — now carries the verified tax-reference path ([…45, 77, 80, 95] for Feb–May), flipping from red (below cutoff) to green (above it) from March. The Reserves banner, which still read "red path active · May 2026 · Urals ~$39," was stale (that $39 was the January low); it now states the Q1 red path was real but has been partly offset since. The NWF series is untouched.
Verification: tax-Urals $77 (Mar) and $94.87 (May, "highest since Oct 2023," ~7,300₽/bbl, +18% m/m, +60% y/y) confirmed T1/T2 (Bloomberg, Reuters, economy/finance-ministry data via Moscow Times); $116 Primorsk 13-year high (Argus via Bloomberg, RussiaPost); ~$9B doubled April oil tax (Reuters); discount-collapse + "biggest winner" (FT/CREA via RTÉ, CNBC/Vakulenko's ~$9B/month estimate). The June-3 US-Iran exchange that prompted this (Kuwait airport hit, CENTCOM strikes on Qeshm, a Hellfire on an Iran-bound tanker) confirmed T1/T2 (CNN, Al Arabiya); the submitting Naki video is T3 framing on top of those verified events. Counterweights kept, surfaced as carefully as the relief: (1) the windfall under-delivered — first-four-months 2026 oil-and-gas revenue was still RUB 2.3T, −38% year-on-year, and April's actual "war premium" was ~RUB 21B against a RUB 200–250B forecast; (2) Ukraine's strike campaign (v3.18/v3.22/v3.27) is wrecking the refining and export capacity needed to capitalize on high prices — high crude price, degraded ability to sell it; (3) transport/insurance costs exploded (Sechin: $2→$20/bbl, seizure risk — cf. the shadow-tanker detentions); (4) Moscow nonetheless cut its 2026 growth forecast to 0.4% from 1.3%; (5) on 4 June Brent was ~$97, down ~12% on the month — the Hormuz premium is real but volatile and could unwind on a ceasefire. Verdict: a self-correction in the spirit of the v3.14 process-correction — the model had logged pressure richly and under-weighted the biggest relief variable, and this fixes the weighting. It does not trip a hard kill-criterion; note elevated oil is the opposite of the "Urals sustained below ~$40" criterion, so it pushes away from the red path. Base case unchanged but recalibrated honestly: the squeeze eased materially in spring 2026; destination still a negotiated settlement, with the reserve drain now slower than even the optimistic scenario band on the projection chart. Watch item: whether a Hormuz reopening / ceasefire collapses the premium back toward the winter lows, which would re-arm the red path.
The v3.28 watch item was explicit: a Hormuz reopening or ceasefire collapsing the premium back toward winter lows would re-arm the red path. It has. After the US-Iran agreement restored shipping through the Strait of Hormuz, tax-reference Urals fell from the May peak ($94.87) to a monthly average of $60.92 in June, then crashed to an average of $41.66 in the first three days of July at western ports, back to pre-Middle-East-war levels and nearly half the April peak. The discount to Dated Brent blew out to $27.35. The June windfall did real work first: the NWF was replenished for the first time in nearly a year, and liquid assets stood at $46.4B (3.61T rubles) on July 1, up from $44.11B a month earlier. But the deficit ran 6T rubles (~$77B) in the first five months, roughly 60% above the full-year plan, and because oil taxes lag a month, the July crash lands on the budget in August.
Verification: $41.66 early-July average and $60.92 June average from Argus data via Bloomberg (T1); corroborated Meduza, OilPrice, Ukrinform retellings (T2/T3). NWF July-1 figure from Finance Ministry data via TASS (T1 on the number, interested party on framing); deficit figure via Bloomberg/Argus reporting. Spot readings since run $47–55 and volatile, so the July monthly tax figure is not yet settled. Kill-criterion check: "Urals sustained below ~$40 with an effective strike campaign." The campaign half is now satisfied beyond dispute (see the Oil panel); the price half sits ~$2 above the threshold for days, not the sustained months the criterion requires. NOT fired, but this is the closest approach in the project's history, and the two halves have never been satisfied simultaneously before. Verdict: watch item fired; red path re-armed. Base case unchanged, destination still a negotiated settlement, but the v3.28 recalibration ("squeeze eased materially") is fully reversed. New watch: the July and August monthly tax-Urals averages. Two consecutive months below ~$40 with the campaign sustained would fire the criterion.
Reuters reported (July 6) a European intelligence assessment warning that Russia's banking sector is becoming increasingly vulnerable after years of war-driven lending, with rising bad loans. This extends the v3.10 banking-strain thread (TsMAKP troubled assets above 10% of banking assets; CBR ~10T rubles of bad loans) with a higher-authority source and a new event: an intelligence assessment circulated as the EU's next sanctions package advanced.
Verification: Reuters primary (T1), corroborated via OilPrice this pass. Provenance note: this substance was first logged in a v3.29 build delivered in a prior session whose file never returned to the project; it is re-verified and re-logged here so that v3.30 supersedes v3.29 regardless of which copy is deployed. Cross-checked against v3.10 for double-counting: same mechanism, new source class and new event, so logged as an extension rather than a restatement. Verdict: raises the banking tail risk that structural theory (predictor ④) tracks; no hard kill-criterion; base case unchanged.
The fuel shortage is now transmitting into food prices and the harvest itself. Rosstat data, late May through July 6: potatoes and onions +27.7%, cabbage +18.8%, beets +17.2%, carrots +13.1%, chicken +6.1%; headline CPI jumped to 6.0% in June (highest since January, against the global disinflation trend), gasoline inflation hit 19.9% y/y, and the June 16–22 weekly gasoline rise (+3%) was the largest in 20 years of Rosstat weekly data. The central bank itself flags fuel as the key upside inflation risk, constraining rate cuts just as Putin pushes for easing; freight costs to Crimea are up 3–6x. On the harvest: threshing is 1–2 weeks behind (1.3–1.5M ha by July 1 vs 4.2–4.6M a year earlier); a combine burns up to 300L of diesel per shift against station caps of 100–200L, and grain must come in within 7–10 days of ripening (Sizov/SovEcon); the worst-hit south (Rostov, Krasnodar, Stavropol) grows a fifth of the grain; a Bloomberg survey found only 2 of 8 farmers fueled to finish. Second-order loop, pre-registered: winter sowing runs Aug–Oct; a persistent diesel shortage would shrink 2027 crop area, seeding next year's problem from this year's crisis.
Verification: price data T1 (Rosstat, via Moscow Times and TASS weekly prints; CPI/gasoline series via Trading Economics, CEPA); harvest lag T2 (Kommersant/IKAR/Rusagrotrans via Meduza; Bloomberg farmer survey; Forbes retellings); the claimed "up to 15% regional losses" is a consulting-firm estimate held as T3. The submitting source (a regional-media coalition post) was T3 and accurate on substance. Counterweights kept: (1) IKAR and the Agriculture Ministry attribute the harvest delay so far primarily to weather, with fuel compounding; (2) large agroholdings are reportedly covered on diesel through end-2026 (Kommersant sources); (3) SovEcon still forecasts ~88.9Mt of wheat, only −2.5% y/y; (4) prewar diesel output ran ~2x domestic need, and the state is prioritizing farm diesel and weighing a full diesel export ban [enacted ~Jul 8 — see the Oil panel v3.32 entry]; (5) May food inflation was just 2.9%, so this spike is new, not chronic. Calibration against collapse-date seduction, stated plainly: this is a price-and-logistics crisis, not a caloric one. Russia is the world's largest wheat exporter and can curb grain exports (as it did ammonium nitrate in March) long before domestic supply binds; famine or hunger-riot scenarios have no defensible path from current data, and the model declines to assign one. The real channel is political: food prices land in the heartland during the Sept 18–20 election window, feeding the same Aug–Oct decision point the fiscal and fuel clocks already converge on. Verdict: no hard kill-criterion; base case unchanged; the squeeze acquires a third convergent clock (fiscal, fuel, now food/harvest) on the Aug–Oct window. Pre-registered watches: weekly Rosstat food prints through the harvest; harvest completion vs diesel availability by end-August; winter-sowing area signals Sept–Oct; any grain-export restriction (which would cut export revenue, feeding the fiscal loop).
Household deposits dwarf the reserve fund — an enormous war chest on paper, but a one-shot weapon that destroys the system producing it.
Pools of money vs annual costs, US$ billions. Deposits (~$930B) could in theory cover the deficit ~15–18 yrs — but a hard seizure ends trust permanently. Soft extraction (forced OFZ-bond conversion, inflation tax, capital controls) extends the comfortable runway ~1–2 yrs while impoverishing households.
Early tremors: a record 1.6-trillion-ruble withdrawal in January 2026 and the first term-deposit outflow since the 2022 mobilization panic. Russians remember 1991 and 1998 — confidence is brittle.
The real contest isn't who hits zero first. Russia's clock is internal capacity (reserves + pain tolerance); Ukraine's is external political will (EU/US funding). Both are wound to roughly the same window.
Funded runway from mid-2026. Ukraine: €90B EU loan covers ~⅔ of 2026–27 needs through end-2027; ~€200B frozen Russian assets untapped as a strategic reserve. Russia: comfortable into ~2027 (oil-dependent), then borrowing/printing/austerity.
The NATO summit in Ankara (July 7–8) produced the largest multilateral funding pledge of the war: the summit declaration commits €70B in military equipment, assistance, and training for Ukraine in 2026, with sovereign commitments to at least equivalent levels in 2027, ~€140B total. The declaration also recognizes Ukraine as a "security contributor" to Euro-Atlantic security for the first time. Separately, Trump announced the US will grant Ukraine a license to produce Patriot interceptors, reversing more than six months of refusal.
Verification: declaration text T1 (NATO official text; European Pravda, Interfax-Ukraine, Kyiv Post corroboration). Calibrations the headline drops: (1) the pledge comes from European allies and Canada only, explicitly excluding the US (the euro denomination was deliberate signaling); (2) it is not all new money, since it folds in the EU loan's defense portion (~€28.3B in 2026) and existing bilateral programs, though the total still exceeds current commitments; (3) the "collective security document" is the standard summit declaration reaffirming Article 5, which all leaders including Trump sign, not a new Ukraine-specific guarantee. Patriot license, scoped: an air-defense input, not a funding event; a verbal announcement with Lockheed and Raytheon not yet formally informed, production realistically years away (Germany's licensed line took years), and no fix for the acute shortage (two nights before the summit, 29 Russian ballistic missiles reached Kyiv with zero intercepts). Verdict: the "EU/US funding lapse" kill-criterion moves further from firing; Ukraine's funded clock through end-2027 is reinforced from a second, overlapping source. The submitting Tsymbaliuk video is T3 framing over these verified events. No hard kill-criterion; base case unchanged.
Confidence comes from convergence — methods that share no inputs reaching the same conclusion. Five now do.
Prediction-market implied P(ceasefire) by deadline (Polymarket, pulled 12 Jul 2026). Near markets price a 2026 ceasefire as possible but unlikely (~20–30% banded across markets); the bulk of probability sits in 2027. Markets still price a deal, not collapse. *Aug-26 bar is the signed-peace-deal market (stricter question); end-27 bar is a May snapshot of a thin market, held pending a fresh read.
Buffer exhaustion late 2026–2028, oil-dependent. Austerity, not surrender.
Money-weighted crowds price a negotiated ceasefire as the likely 2026–27 outcome.
AIS/satellite oil flows confirm stress — but a "shadow fleet" keeps exports moving (since Jul 2026 under direct drone attack in the Azov; resilient but contested. See the Oil panel).
Fragility models flag risk factors but are notoriously bad at predicting when.
Funds front-run a deal: ruble +~30–36%, money into Russia-proxy assets & distressed bonds.
In 1998 a hedge fund (LTCM) was sure it had hedged Russian risk — then Russia defaulted, froze ruble trading, and the fund collapsed. Markets can price Russia catastrophically wrong. The "thaw" bet counts as evidence only because it agrees with four unrelated methods.
Predictor ④ (structural theory) updated: independent reporting confirms genuine elite discontent. Bloomberg and Meduza both report senior Kremlin officials see the war at a dead end while Putin, "obsessed with Donbas," pushes to finish by end-2026 on terms he calls victorious — and receives a distorted optimistic frontline picture. European intelligence reports coup/assassination fears since March 2026 (Shoigu named a destabilizing-actor risk), with security tightened around Putin. This elevates the elite-fracture risk factor that structural theory tracks.
Verification: "elite disappointment + Donbas fixation + distorted reporting" corroborated across Bloomberg, Meduza (inside-Russia), the Guardian (T2). Coup-fear/security-tightening from European intel (T2/T3) — with journalists' own caveat that the access implied is "remarkable" and unexplained. The load-bearing counterweight (quoting the next sentence, not just the dramatic one): the most authoritative voice in the cluster, an Estonian intelligence chief, told CNN there is "little sign of unrest given tight security-service control of dissent." Elite discontent has been documented since Jan 2026 with zero structural consequence — confirming strain's political inertness under this system.
Why it reinforces, not breaks, the base case: Putin's reported desired exit is "a security agreement with Europe acknowledging territorial gains" — a negotiated settlement framing, pointing the same direction as the other four predictors. Structural theory's verdict stands: real strain, poor at timing, no imminent fall. No hard kill-criterion. Base case unchanged.
There's no "destroy X% and Russia negotiates" number. Strikes don't zero out revenue — they raise the oil price Russia needs to break even.
Annual oil & gas revenue by Urals price. Without strikes Russia clears the self-funding line ~$90; with a sustained campaign, not until ~$120+. Net effect: strikes raise the fiscal breakeven by ~$30/barrel.
The decisive caveat: it only bites at low prices. At ~$115/bbl the same campaign barely dents the budget; at $40–50 it's devastating. Ukraine controls the destruction, not the price — so oil is necessary but not sufficient.
A two-week drone wave halted or sharply cut central-Russia refineries: NORSI's CDU-6 (53% of the plant, ~190k bpd) down after a May 20 strike; Kirishi/KINEF (20M t/yr) fully offline since May 5; Ryazan suspended May 15, Moscow refinery May 17. Combined: 83M+ t/yr — ~25% of total refining, 30%+ of gasoline, 25% of diesel. Russia banned gasoline exports through July 31; refineries targeted have doubled since January. Now stacked on the gold sell-off (v3.2) and $39 Urals.
Verification: confirmed across Reuters, Moscow Times, Meduza (T1/T2); submitted Ukrainian-outlet article (T3) accurate but understated the systemic scope. Two caveats held: (1) refining hit ≠ proportional export-revenue loss — crude still ships, so budget damage is smaller than the headline %; (2) repair-and-reroute is real — Ryazan resumed within days/weeks before by overloading spare CDUs. The decider is whether strikes now outpace the 6–18mo repair cycle.
Verdict: strengthens the red path materially — most consequential update yet — but still no hard kill-criterion. The export ban is the adaptive, survivable response the model predicts. Base case unchanged: fuel shortages pressure Russia toward the table, not toward sudden collapse. Two clocks now accelerating (reserves + oil).
⌖ Watch item: the next real escalation would be ~25%-offline sustained into June–July plus renewed Russian fuel-import requests (Kazakhstan/Belarus, as in past shortages) — evidence the repair race is being lost. That, not a higher one-day strike tally, would justify a v3.5 fiscal bump.
The specific v3.3 pre-registered watch item — "renewed fuel-import requests from Belarus/Kazakhstan" as evidence the repair race is being lost — has materialized. Reuters/Moscow Times: Russian gasoline imports from Belarus surged 4× month-on-month to ~49,000t (gasoline) + 33,000t (diesel); Russia preparing additional imports from China, South Korea, Singapore. Shortages have spread beyond initial regions: rationing in southern Russia, Far East, Siberia, Vladivostok, Crimea, plus reported pressure on central regions. Foreign Affairs notes the government is counting on Belarus for potentially up to 30% of Russia's gasoline needs. This is the trigger I named explicitly two updates ago — and it fired.
Verification: Belarus-imports surge and geographic spread of shortages confirmed at T2 (Reuters/Moscow Times/Foreign Affairs/Kyiv Independent). One Krutikhin claim flagged unverified: his "440 vs 511 million tons" production-discrepancy figure has no independent corroboration in this search (no Bloomberg/Kpler/IEA anchor) — logged as his claim, not as established fact. Same discipline that rejected the v4.0 "satellite source" fabrication: precise-and-damning isn't license to launder a number into the record. Belarus tolling framing is accurate; the Strait-of-Hormuz geopolitical commentary is opinion, out of scope.
Methodological note: this is the first update in many rounds whose central claim is new and falsifiable rather than a restatement — a pre-registered trigger materializing. That is the opposite of fitting everything into "confirms the squeeze"; it's a watched event happening. Verdict: escalates the oil clock with conviction. Still not a hard kill-criterion (no ceasefire, NWF-rule change, mobilization). Base case unchanged — Russia importing fuel is what survivable-but-degraded looks like, same shape, more pressure. The next pre-registered escalation now becomes a nationwide coupon-rationing decree or import dependency materially exceeding domestic capacity for sustained months.
Bloomberg/OilX: Russian refining fell to 4.58 million b/d in May 2026, down 700k from year-ago, lowest level since October 2009. May was a record strike-tempo month: 16 strikes on fuel-producing facilities, hitting 8 of Russia's 10 largest refineries (Yanos struck 3×, two Lukoil plants 2× each — repeat-hit strategy preventing repairs). Compounds the v3.3 (campaign at threshold) and v3.12 (Belarus imports triggered) entries: the April figure was 4.69 m b/d (Dec-2009 floor); May broke through to 4.58 (Oct-2009 floor). The repair race is being lost in measurable monthly increments.
New texture — retail-level spillover: AI-95 gasoline available volumes shrank to ~⅓ of last year's, wholesale prices up 20%+, independent stations struggling to buy stock (Bloomberg/OilX, Dagens). Deputy PM Novak chaired a meeting to close the jet-fuel export loophole — kerosene now being pulled off the export market (Reuters/Interfax). This is the squeeze visible at the gas pump and the wholesale market, not just in throughput statistics.
Verification: Bloomberg primary reporting (T1) — published June 1; cross-corroborated by Critical Threats Project (ISW affiliate, T1/T2) for the April 4.69 m b/d figure, Euromaidan Press citing IEA May report, and Dagens reporting the specific OilX figures. Strongest sourcing in the oil-panel series. Portnikov's "Russia will switch to mass-bombing cities" forecast and the "demographic war" framing held out as T3 commentary, separate from the verified data.
Verdict: confirmed escalation, not restatement. The oil clock now visibly compounds — strike tempo at all-time high, refining at 17-year low, retail spillover beginning, jet-fuel export ban tightening. Still no hard kill-criterion (no nationwide coupon decree, no NWF rule change, no ceasefire, no mobilization). Base case unchanged — destination still negotiated settlement; speed and depth of the squeeze visibly compounding. Pre-registered next escalation remains: nationwide coupon rationing, OR sustained import-dependency exceeding domestic capacity over multiple months.
The pre-registered coupon-rationing watch-item (named in v3.12 + v3.18) has partially materialized. Effective May 30, occupied Crimea introduced coupon rationing for AI-95 gasoline and capped AI-92 at 20 litres/person/day (Aksyonov); Sevastopol began coupons May 22; restrictions also reported in occupied Zaporizhzhia/Donetsk and Russia's Belgorod region. Cause confirmed: Ukrainian drone strikes on the R-280 "Novorossiya" land-bridge resupply route. This is the strongest form of the watch-item seen so far — actual coupons, not just caps.
Why "partially" — the load-bearing distinction: what was pre-registered as the hard trigger was a nationwide coupon decree. What fired is regional rationing in occupied Crimea/Sevastopol — a logistically isolated peninsula fed by one vulnerable land bridge plus the Kerch bridge (capped at 100 L/vehicle for security). Crimea is the predictable first place to crack, not evidence the mainland (Moscow, the industrial heartland) is cracking. Treating regional Crimea rationing as nationwide Russian rationing would be exactly the overreach the method guards against.
Verification: confirmed at T1/T2 across Reuters (on-the-ground witnesses, queues in Sevastopol), CNN, NYT, Kyiv Post, Kyiv Independent, TVP, and RFE/RL (Radio Svoboda, 31 May) — multi-source, not single-outlet. Historical precedent (RFE/RL via "Agentstvo"), which calibrates the read: fuel coupons have appeared in modern Russia only rarely — last regional case Magadan Oblast 2017, before that the 2011 fuel crisis in Tuva, Primorye, and Amur Oblast. All were regional/peripheral, never nationwide — so Crimea rationing fits the peripheral-region pattern repeating, reinforcing (not contradicting) the regional-vs-nationwide distinction below. Held as partisan emphasis: the "largest fuel crisis since 1991 / Soviet collapse" framing comes from a Ukrainian defence-ministry adviser (Sternenko) — directionally striking, logged as official emphasis, not as established fact. Aksyonov's "stabilize within 30 days" is noted by observers as unrealistic.
Verdict: watch-item moves from green to amber — partially materialized, escalating toward the threshold, not across it. Strengthens the oil clock. Still no hard kill-criterion: regional occupied-territory rationing is not the nationwide federal decree that would move the base case. Base case unchanged. The threshold to watch now: coupon rationing appearing in mainland Russian regions beyond the border/occupied zones (e.g. central or Volga federal districts), which would be the genuine nationwide signal.
On June 3, Ukrainian drones struck the St. Petersburg oil terminal — one of the largest on Russia's Baltic coast — and the Kronstadt naval base, after flying ~1,100 km, hours before Putin's flagship SPIEF economic forum opened in his hometown. This stacks on v3.18 (refining at 17-year low) but escalates the type of target: refinery strikes hit the fuel-supply lever (army + domestic market); Baltic export terminals hit the crude-revenue lever (the budget itself). Combined with the v3.21 oil-price counterweight, the revenue picture is now pressured from two sides — Ukraine attacking export infrastructure while the price floor stays volatile.
Verification: confirmed T1 — NPR, CNN, Moscow Times, Irish Times, Channels TV all corroborate the St. Petersburg oil-terminal + Kronstadt strikes June 3; Russia's MoD acknowledged downing 354 drones across 15+ regions. Strongest sourcing tier (Western wire services, same-day). The "20th oil facility in 33 days" framing from the source video is plausible and consistent with the v3.18 campaign but not independently itemized here — logged as directional, not a verified count.
Verdict: confirmed escalation onto the export-revenue lever. The campaign is real but Russia's crude export volume has historically been resilient (shadow fleet keeps flowing — predictor ③); terminal strikes raise friction and cost, they don't zero out revenue. No hard kill-criterion. Base case unchanged — pressure broadening from fuel to revenue, same destination.
The v3.19 pre-registered threshold was explicit: the genuine nationwide signal would be rationing appearing in mainland Russian regions beyond the border/occupied zones. That spread has now begun. Retail fuel-purchase restrictions are confirmed at gas stations in Novaya Moskva (inside Moscow city limits), the wider Moscow region, and St. Petersburg, plus Belgorod, Kursk and Pskov (border) and Novgorod, Karelia and Murmansk (north/northwest). Cause is the same documented chain: the v3.18 strike campaign drove refining to a 17-year low, and the resulting product shortfall is now visible at the pump in the political heartland, not just on the logistically-isolated Crimean peninsula.
The load-bearing distinction (why this is "advancing toward," not "across"): what crossed into the mainland is per-customer volume caps (ORTK 60L petrol / 100L diesel; Gazprom 100–150L; Kirishiavtoservis 50L/receipt in the Leningrad region; a 20L cap at one chain across Moscow/Tver/Yaroslavl), not coupon rationing. Coupons/vouchers proper remain specific to occupied Crimea and Sevastopol — and there the largest chain has just suspended vouchers as the shortage deepened. Most mainland caps (100–150L) sit well above a normal fill and barely bind for ordinary drivers. The pre-registered hard signal was nationwide coupon rationing; the mainland has reached the softer cap form. Geographic reach escalates; severity in the heartland stays mild.
Verification: submitted source was Korrespondent (T3 Ukrainian retelling) headlining "15 regions"; it cites a Moscow Times piece whose own headline says "20." Verified the substance independently at T2 across The Insider, Meduza, The Moscow Times, Kyiv Post and UNITED24 (ORTK/Gazprom/Kirishiavtoservis caps, Belgorod/Kursk no-canister rule, occupied-Luhansk 20L). Count banded, not asserted: current-episode tallies range from "at least four" (The Insider, conservative) through ~15 (Korrespondent) to ~20 (Moscow Times). Do not conflate with autumn 2025: the "20+/57 regions" figures in some search hits are from the October-2025 fuel-crisis wave (Chelyabinsk/Tyumen/Sverdlovsk), a different episode — excluded. Counterweights kept (per discipline, surfaced as carefully as the pressure): (1) RBC-Ukraine's own expert, Hennadii Riabtsev, says it is too early to call a nationwide Russian crisis ("they have plenty of gasoline... the local problem is in Crimea"); (2) a large oil-depot head told Fontanka via Meduza the problem is "isolated stations, unpopular ones, low throughput, remote areas," not global; (3) Russian officials Novak and (earlier) Tsivilev claim a "stable" market — interested-party T3, but noted; (4) the distant Far-East/Siberia caps the source lists (Krasnoyarsk, Tomsk, Kamchatka) are partly seasonal (Kamchatka's own stated reason is preserving stock until next year's winter road), so they are not folded into the strike-driven count. Verdict: the v3.19 watch-item advances — the spread into mainland federal territory including Moscow and St. Petersburg is the signal that was being watched, and it is real and multi-source. But it arrives as purchase caps, not the nationwide coupon decree pre-registered as the hard trigger, so it moves the dial toward the threshold without crossing it. Still no hard kill-criterion (no ceasefire, no sustained sub-$40 Urals, no NWF-rule change or mobilization, no funding lapse). Base case unchanged — the squeeze is now visible at the heartland pump, which is exactly what "uglier, survivable" looks like from inside; same destination, more pressure. Next threshold to watch: actual coupon rationing (not caps) in a core mainland region, or caps tightening to genuinely binding low levels (≤20–30L) across the heartland for sustained weeks.
The campaign crossed several structural lines in June–July. On July 6, drones struck the Omsk refinery, Russia's largest (22Mt/yr) and the last of the top 11 never attacked, at 2,500+ km, a new range record; two industry sources told Reuters it halted crude processing, and Saratov halted after a July 7 strike. The Moscow refinery at Kapotnya was critically damaged by paired June strikes (out an estimated 6 months); Ryazan has been suspended since May 15. Russia no longer has a single untouched major refinery. Reuters estimates fuel production fell 25% y/y in June and runs ~20% below domestic demand. Downstream: rationing measures in at least 56 regions (Moscow Times open-source count) and shortage reports in at least 78 of 83 regions (Novaya Gazeta Europe), with the FT estimating 50 million people affected; Putin publicly acknowledged the shortages; a full diesel export ban is under consideration [in force from ~Jul 8 — see the v3.32 entry below]; fuel-quality rules were loosened; and Russia, one of the world's largest oil producers, is importing gasoline by sea from Belarus, India, and Asian markets.
Verification: Omsk/Saratov halts T1 (Reuters); Kapotnya damage T2 (UNITED24, Carnegie); crisis geography T1/T2 (Moscow Times, Novaya Gazeta Europe, Al Jazeera on-ground, FT, Politico). Capacity offline, banded per discipline: Ukraine's General Staff claims 42.7% of design capacity disabled (T3, interested party); the IEA says "more than 20%"; the FT reports June output 28% below the 5-year average; Carnegie put the April–May output decline at ~13%. Logged as a T1-anchored 20–28% output reduction, with the 43% figure held as the Ukrainian design-capacity claim. The pre-registered threshold, called honestly: v3.19/v3.27 named a nationwide coupon decree as the hard signal. No single federal decree exists; what exists is a dense regional patchwork (20–40L caps, car-plate rationing, Irkutsk's "manual mode", priority tiers for officials). Substance crossed the threshold; the pre-registered form did not. Counterweights kept: (1) Vakulenko/Moscow Times (mid-June, pre-Omsk) argued acute shortages concentrate within ~250 km of the front, driven by strikes on trucks and logistics more than national production, with media portrayal exceeding most Russians' lived experience; (2) crude exports are the relief valve, running 3.83M b/d in the four weeks to June 14, the year's highest pace, as unrefinable barrels flow to export (now at a crashed price); (3) diesel was structurally overproduced prewar, so the gasoline crisis does not automatically become a diesel crisis. Verdict: confirmed structural escalation, the strongest in the series. No hard kill-criterion (no decree-form rationing, no ceasefire, price not yet sustained sub-$40). Base case unchanged; the "effective strike campaign" half of the oil kill-criterion is now satisfied beyond dispute.
The v3.19 watch item's strongest possible regional form. From June 21, occupation authorities halted all fuel sales to individuals and businesses, by cash, card, or coupon, reserving stock for what Aksyonov called state services ensuring the republic's functioning and security; a state of emergency was declared, the TES chain's voucher system was suspended, children's summer camps were paused, and residents report fuel effectively reserved for emergency services and utilities, alongside rolling blackouts and water rationing. Tourism, the peninsula's economic anchor, is taking cancellations in peak season. Aksyonov warns the situation "will continue for some time."
Verification: T1/T2 multi-source (Reuters witnesses, Moscow Times resident interviews, BBC, CNN, Kyiv Post). Cause chain confirmed: strikes on the R-280 corridor, fuel trucks, depots, and substations isolating the peninsula. The load-bearing distinction, retained from v3.19: Crimea remains the logistically isolated first-to-crack periphery, not evidence of equivalent mainland severity; mainland restrictions remain caps and rationing, not total bans. Verdict: the regional watch item has gone as far as it can go short of the mainland equivalent. No hard kill-criterion. Base case unchanged. New regional threshold to watch: a total-sales halt or state of emergency in a core mainland region.
Between July 6 and 12, Ukraine's Unmanned Systems Forces claim ~90 vessels hit in the Sea of Azov, mostly shadow-fleet tankers plus tugs, ferries, and cargo ships, with strikes averaging one every 112 minutes at the peak. The effects are independently confirmed: Russia suspended navigation on the Don-Azov Canal and stopped accepting Kerch Strait transit applications from July 10 (Reuters, via grain-industry sources); the Rostov governor confirmed tanker hits in Taganrog Bay; satellite imagery shows burning tankers near the Kerch bridge; and Euronext wheat futures jumped 4% on fears for the up-to-25% of Russian wheat exports that transit the Azov. In parallel, "Operation Crimean Switch Off" hit ~50 energy facilities in Crimea in the first week of July.
Why this is load-bearing: predictor ③'s standing counterweight ("the shadow fleet keeps exports flowing") and the v3.22 caveat ("crude export volume historically resilient") are now under direct, sustained attack for the first time. Combined with the fuel-supply logic (these tankers fed Crimea), this is a two-lever event: it tightens the Crimea siege and raises the cost and insurability risk of the sanctions-evasion fleet that underwrites the crude-export relief valve.
Verification: effects T1 (Reuters on the Don-Azov/Kerch suspensions, Rostov governor's confirmations, Vantor satellite imagery via TWZ, Euronext price move); vessel counts are Ukrainian military claims (T3) with only partial independent itemization (Ambrey/Equasis matched several named ships). Counterweights kept: (1) strikes concentrate in the shallow Azov, not (yet) the Baltic and Pacific export arteries that carry most crude volume, so the resilience assumption is challenged regionally, not broken globally; (2) Ambrey warns of a realistic possibility of Russian retaliation against Ukraine-linked shipping within days; (3) crude shipments overall were still running at the year's highest pace in mid-June. Verdict: no kill-criterion; base case unchanged. Predictor ③'s caveat is formally downgraded from "resilient" to "resilient but contested" (see the Validation panel).
Russia banned most diesel exports through the end of July, in force from ~July 8, stacking on the existing gasoline and jet-fuel export restrictions. The loss of diesel from a key global supplier pushed world diesel prices to multi-year highs in a market already tightened by Middle East disruption. Fiscal read: the ban protects the harvest and domestic supply at the direct cost of export revenue — the same fuel-to-fiscal loop pre-registered in the v3.31 food entry, now running through policy choice rather than price alone.
Verification: Reuters, the IEA monthly report, and Bloomberg (T1). This supersedes the three stale "under consideration" references carried in the v3.30/v3.31 text, which are bracket-annotated in place — the dated entries themselves are preserved as the historical record. Counterweights kept: (1) the ban is time-limited (to end-July) and reversible; (2) diesel was structurally overproduced prewar, so cutting exports frees real volume for domestic and farm use; (3) Russia used a comparable diesel export ban in autumn 2023 and lifted it without lasting fiscal damage. Verdict: no hard kill-criterion; base case unchanged; the fiscal loop tightens by policy choice, which is itself a signal of how binding the shortage has become.
Overnight July 12, HUR and the Unmanned Systems Forces re-struck the Syzran refinery (Rosneft, 8.5–8.9Mt/yr, ~3% of national refining), setting it ablaze; the same night's operation hit 10 tankers and 4 ferries in the Azov (an extension of the v3.30 shadow-fleet entry, not double-counted here). Syzran had already been crippled on May 21, when the AVT-6 unit (~70% of capacity) was seriously damaged. The aggregate floor broke the next day: Bloomberg (via EA Analytics, Jul 13) puts July refining runs at 3.91M b/d, the lowest since March 2005, more than 1.4M b/d below the year-ago average. The IEA says more than half of Russia's refining capacity has been affected since the start of May, with June processing at 3.8M b/d, down 1.6M b/d y/y.
Verification: strike T1 (Bloomberg; Ukrainian General Staff statement; fire footage geolocated by multiple OSINT channels). Unit-level claims held T3: Cyberboroshno's "100% of primary capacity damaged" mapping (AVT-5 ~30% + AVT-6 ~70%) is Telegram-channel analysis, not independently confirmed. Astra's "gasoline output covers ~65% of summer demand" held as a T2/T3 single-source figure. Structural caveat: Russia has classified its refinery statistics, so all run-rate figures are independent estimates (EA Analytics, IEA, Energy Aspects) — banded, never official. Verdict: the v3.18 "17-year low" is superseded by a 21-year low; no hard kill-criterion; base case unchanged.
The IEA cut its Russian supply forecast (Jul 10): 2026 down 85k b/d to 8.9M, 2027 down 150k b/d to 8.8M, citing continued strikes on refineries, storage, and transport infrastructure. June crude output: 8.86M b/d (IEA) to 8.928M b/d — the lowest in at least two and a half years (OPEC via Bloomberg) — roughly 900k b/d below the OPEC+ quota. Counterweight, deliberately prominent: seaborne crude exports hit 4.22M b/d in the four weeks to July 5, the highest since the 2022 invasion, and the IEA puts June total crude exports at 5.8M b/d, up 620k m/m. Unrefinable barrels are being exported, not shut in wholesale.
Calibration: the budget channel runs through the extraction tax and price, so record export volume at ~$40–41 Urals partially offsets the refining losses; price — not refining volume — remains the binding fiscal variable and the kill-criterion test. The claim (Portnikov and others) that mass well conservation is underway and will produce an irreversible long-term industry collapse is declined as a T3 forecast: the output decline is real and now institutional-forecast-confirmed, but the record-export data cut directly against the mass-shut-in mechanism, and no T1 confirmation of large-scale well mothballing was found.
Verification: IEA monthly report + OPEC MOMR via Bloomberg, OilPrice, Reuters (T1). Provenance: this entry originates in a verification pass on a submitted Portnikov video (a familiar T3 voice, already held out in v3.18); his "refining at 2005 levels, under 4M barrels" claim T1-verified precisely — the rare case of a T3 headline landing on the nose — while his thesis restatement ("aggression depends on revenue") was not logged, per the double-counting rule. Verdict: no hard kill-criterion; base case unchanged; the structural production decline is confirmed at forecast level, and the collapse framing stays declined.
Telegram channels circulated Wildberries order-cancellation notices citing regional delivery difficulties, with Rostov Oblast among the affected regions. The company's response cuts both ways: RVB (the merged Wildberries-Russ entity) officially denied via TASS (Jul 10) that any cancellations relate to fuel, calling the notices "standard service messages" — yet Wildberries separately raised seller commissions in some categories, explicitly citing fuel-driven increases in logistics costs (expert estimates put the rise at 5–20%).
Verification: cancellation screenshots T3 (Telegram channels); the denial is T1-on-the-record (TASS, carried by Kommersant, Fontanka, m24); the commission increase is T2 (Realnoe Vremya citing the WB press service). The denial is logged at full weight — automated cancellation notices genuinely are routine, and the screenshots alone prove nothing. The commission hike is the harder datapoint, because it is the company's own attribution of cost pressure to fuel. Verdict: logged as a contested early signal of fuel-crisis transmission into consumer logistics — a watch item, not established fact. Watch: courier or marketplace delivery failures reported at T1/T2, or fuel surcharges spreading across logistics operators.
The IISS / Gould-Davies thesis as one input — corroborated on diagnosis, flagged on its "collapse in months" timing.
"Death-nomics" failing — paying more, getting fewer. Recruitment fell ~20% to ~800/day even as the average bonus hit a record 1.47M rubles (Mar 2026). The video's "40% drop" is roughly double the verified ~20%.
Q1 2026 balance (thousands), with uncertainty whiskers. Recruited (~80k) sits just below Ukrainian-estimated casualties (~85k, upper-bound) and below the ~102k/quarter the stated target implies. Kluge: enough to offset losses, not expand. The white error bars show the uncertainty band — widest on the casualty figure (±12k, partisan upper-bound) and the target (±18k, inferred) — a deliberate visual reminder these are ranges, not exact numbers. The deficit is real but narrow; "decisive shortfall" overstates it.
The strain is real and independently confirmed (recruitment, bonuses, rising deaths, creeping mobilization, the internet clampdown). But "collapse in months" is the collapse-date seduction the method warns against — and the escape valve is that Russia can shrink the deficit by slowing offensives. A manpower problem pressures the timeline toward a deal, not a sudden break. True data, overconfident forecast.
Illicit Russian Starlink terminals were deactivated in Feb 2026. A declassified US defense-intelligence assessment (via Bloomberg) found this disrupted Russian command-and-control and helped Ukraine retake ~400 km². Why it touches this panel: the escape valve above is "Russia trades offensive tempo for lower casualties." If degraded C2 independently blunts effectiveness and costs ground, that valve narrows — harder to slow down gracefully when you're already losing territory. Marginally reinforces the manpower-strain side.
Verification: Starlink cutoff + territorial effect confirmed at T1/T2 (Reuters, Bloomberg, US intel) — unusually strong for a battlefield claim. Scope discipline: the wider "southern logistics collapse / Crimea trap / FSB-base strike" narrative (Kovalenko, Information Resistance) is T3 battlefield assessment — plausible but unconfirmed, and a different driver (can Russia hold ground) than this funding model (can Russia afford to fight). Not logged as fiscal data; experts also caution the Starlink loss is "unlikely to change the course of the war."
Verdict: a confirmed manpower-clock input, not a funding event. No hard kill-criterion. Base case unchanged.
Good-news half (manpower clock): ISW recorded Russia's first net loss of controlled territory since the Aug-2024 Kursk incursion — 116 km² lost in April 2026 — with average daily gains down to 2.9 km² from 9.76 a year earlier; net losses continued into May (~69 sq mi over Apr 21–May 19). DeepState: ~12% less territory taken than March, advance rate down ~80% ("sputtering"). Reconfirms the offensive is failing and the Starlink driver (v3.4).
Caution half (escalation risk): FT (two people in contact with Putin + Ukrainian intel) reports commanders have convinced Putin he can take all of Donbas by fall — the same distorted-reporting mechanism as the v3.8 elite-fracture log. A leader who believes the enemy is collapsing and that he holds a "superweapon" demands more assaults and mobilizations, not fewer. The facts favor Ukraine; a reality-detached Putin makes the path bloodier before any settlement.
Verification: 116 km² and the slowdown confirmed across ISW, DeepState, Kyiv Post, Russia Matters (T1/T2) — strongest battlefield sourcing in the series. Caveats the source dropped, kept here: (1) ISW's infiltration asterisk — including infiltration tactics, Russia's April net is +28 km²; the 116 km² loss is controlled-territory only. (2) ISW flags spring rasputitsa mud as a partial, seasonal cause — not all Ukrainian pressure. The Oreshnik anecdote and Z-blogger quotes are T3 commentary, illustrative not load-bearing.
Verdict: dual signal — reinforces manpower clock + deepens elite-fracture finding (now FT-sourced). No hard kill-criterion, but raises the probability of one of the watched triggers (a new mobilization). Base case unchanged — the destination is still a negotiated settlement, but the near-term path looks more violent.
Same scope as v3.4 (Starlink): confirmed battlefield input that touches the manpower clock indirectly, not a fiscal data point. On May 31, Brig. Gen. Biletsky (3rd Army Corps) announced "Logistics Lockdown" — drones of the 3rd Assault Brigade's unmanned-systems battalion now exercise fire-control over Russian supply routes across occupied Luhansk Oblast. Confirmed strike at the Izvaryne checkpoint 205 km from the front — a new record range (previous geolocated max ~150 km). Named cities under drone fire-control: Luhansk, Starobilsk, Alchevsk, Brianka, Kadiivka.
Why it matters for this panel: stacks with v3.4 (Starlink C2 cutoff) and v3.9 (sputtering offensive). Russian operational sustainability is now under pressure on three axes simultaneously — degraded C2, sputtering advance rate, and rear-area logistics under live drone fire 200+ km deep. That further narrows the v3.4 manpower escape valve: it's harder to "slow offensives to lower casualties" when your supply lines themselves are interdicted.
Verification: the event itself confirmed at T2 — Kyiv Independent, Euromaidan Press, Militarnyi, UA News all independently report the operation, the 205km figure, the named cities, and Biletsky's announcement on May 31. Held out as T3 (the source's interpretation): the "sudden front collapse" outcome forecast, the $100M figure, and the specific technical claims (Hornet drones + AI vision, balloon relays as the 100-200km bridge method) — coherent and plausible, not independently corroborated. Methodological note — verify before refuse: the initial response on this submission refused to log without first searching. Verification revealed the underlying event is well-corroborated; the refusal was over-strong. Correction applied here: confirmed-event logged, speculative-interpretation held out, scope-discipline (battlefield ≠ funding) preserved.
Verdict: a scoped manpower-clock input, not a funding event. No hard kill-criterion. Base case unchanged. The pre-registered trigger that would justify a real fiscal bump remains: sustained, ISW-confirmed Russian retreat from significant positions attributable specifically to logistics interdiction. Announcing fire-control is not the same as achieving strategic effect.
Mass mobilization is one of the four hard kill-criteria — it would change the war's structure, not just the squeeze's speed — so a credible signal toward it outweighs almost any fuel datapoint. On June 1, a post on the Telegram channel of Andrei Gurulyov (real State Duma deputy, retired lieutenant-general, ex-58th Army commander) stated a "fundamental decision has already been made" on large-scale mobilization in fall 2026. Gurulyov then claimed the channel was "stolen" and the texts spread by "enemies" — but didn't delete the post.
The autumn-2026 convergence: Russia's State Duma elections are Sept 20, 2026. The Kremlin's Telegram crackdown and internet controls have been tied by multiple analysts to pre-election nervousness. Mobilization into an election season is politically explosive — which is both why this post landed hard and why the regime would want it disavowed. Stacks with the v3.9 thread (reality-detached Putin escalating) and the v3.4/v3.14 manpower-pressure findings.
Verification: the post's existence and Gurulyov's identity confirmed at T2 (Meduza, Kyiv Post, Militarnyi). What is NOT confirmed: whether mobilization is actually decided (the post's claim) OR whether it was genuinely "hacked" (his disavowal) — a real event wrapping an unverified substance. Logged as a rumor with unusually high-placed sourcing, not as the criterion firing. Nielsen's four-scenario framework (Danish Defence Academy) noted as supporting T3 analysis; notably Nielsen calls freezing the most rational Kremlin choice — consistent with the base case.
Verdict — timeline effect: does NOT trip the kill-criterion (a walked-back Telegram post is not a decree). But it raises the probability that a hard kill-criterion fires this autumn and sharpens when the base case is most likely to be tested. Base case unchanged; autumn 2026 now flagged as the inflection window — either Russia moves toward the freeze/settlement the model projects, or escalates via mobilization (the riskier v3.9 path). Watch for an actual decree, not more rumors.
The v3.22 signal (Gurulyov's walked-back Telegram post) has been superseded by something categorically stronger: eight sources in the Presidential Administration and security services told iStories/Verstka that mobilization is being discussed "at all levels," with October, after the Sept 18–20 Duma elections, cited as a possible start, and preparations "for what will never be called mobilization" underway for months. Regional administrations were reportedly told to cut 10–15% of staff by October 1. The political backdrop: United Russia polls at a maximum of 35%, security officials reportedly urged Putin to postpone or cancel the elections, and Sberbank CEO Gref told shareholders he does not believe anyone in the country has a primary concern other than ending hostilities as soon as possible.
Verification: insider reporting T2 (Meduza, iStories, Verstka, Novaya Gazeta Europe, corroborating each other); the substance remains unconfirmed rumor with no documented decision. Gref quote T1 (public shareholder meeting). Counterweights kept: (1) Kluge's Q2 estimate shows contract recruitment recovering to ~1,000/day from Q1's three-year low of ~800/day, a real relief signal for the spiral chart; (2) forecasting analysis argues any October drive would overwhelmingly use the stealth legal mechanisms built since 2022 (reservist "special deployments," continuous conscription) precisely to avoid a formal decree, meaning the kill-criterion as written (mass mobilization declared) may never fire even if the substance occurs; that framing gap is flagged as a methodological watch of its own. Verdict: NOT the criterion firing; no decree exists. But the probability mass has concentrated hard on the pre-flagged autumn window, which now converges with the August budget impact of the July oil crash. Base case unchanged; the autumn-2026 inflection window is confirmed and compressed to roughly Aug–Oct 2026.
This dossier is built with the Convergence Method: reframe to the load-bearing variable, band it, cross-validate across independent methods, tier sources, ship something falsifiable. Refresh it on the cadence its drivers move — not a fixed calendar.
Refresh the fast dials only: Urals price, latest NWF figure, recruitment/casualty updates, prediction-market odds. Check whether any kill-criterion tripped. Most months nothing material moves.
Re-run every panel, re-tier sources, bump the version. Aligns with the data's own cadence — Kluge's recruitment data and Russian budget data both land quarterly.
The moment a kill-criterion fires, regardless of calendar: Urals breaks a price band, a ceasefire is signed or collapses, the NWF fiscal rule changes, mass mobilization is declared, or a major EU/US funding decision lands.
Nature of this work. This dossier is independent analytical commentary produced for general informational and educational purposes. It is a structured reading of publicly available information, not original reporting, intelligence product, or expert advice in any regulated field.
No advice. Nothing here constitutes financial, investment, legal, tax, accounting, medical, political, military, intelligence, or other professional advice. Readers must not act on the contents without independent verification and consultation with appropriately qualified professionals in the relevant jurisdiction. Reliance on this material is entirely at the reader's own risk.
No warranty. While reasonable care has been taken to verify and tier sources (see Methodology tab), the dossier is provided "as is" and "as available," without warranty of any kind, express or implied — including accuracy, completeness, fitness for any particular purpose, or non-infringement. Information about ongoing events may be incomplete, outdated, contested, or wrong. Source reliability is graded but not guaranteed.
Forward-looking statements. Any forecasts, scenarios, probability ranges, or "base case" framings are analytical estimates based on the methodology described herein. They are not predictions, guarantees, or representations of future outcomes. Reality may diverge materially from any scenario presented.
No endorsement, no political affiliation. This dossier does not endorse, support, advocate for, or align with any government, political party, military, commercial entity, ethnic or national group, candidate, or cause. Quotation, paraphrase, or analysis of a source does not constitute endorsement of that source. Inclusion of a perspective is not advocacy for it.
Third-party content. The dossier discusses, summarizes, and links to materials produced by third parties (news outlets, governments, analysts, market platforms, commentators). All trademarks, copyrights, and other rights in third-party content remain with their respective owners. Use of such material is for commentary, criticism, news reporting, scholarship, and research under applicable fair-use / fair-dealing principles. Where rights-holders identify content used inappropriately, the publisher will address concerns promptly upon written notice.
Limitation of Liability. To the maximum extent permitted by applicable law, in no event shall the author, the publisher, host entity, or any contributor be liable for any direct, indirect, incidental, special, consequential, exemplary, or punitive damages — including, without limitation, loss of profits, data, use, goodwill, opportunity, reputation, or other intangible losses — arising out of or related to your access to, use of, or inability to access this dossier or any linked material, whether based on warranty, contract, tort (including negligence), statute, or any other legal theory, even if advised of the possibility of such damages.
Independence from employers and affiliations. Views and analysis are the author's personal independent work, conducted on personal time, using exclusively publicly available sources. They do not represent the views, positions, or work product of any current or former employer, client, partner, or institution. No employer-owned data, systems, contacts, workflows, or proprietary information was used in the preparation of this material.
Jurisdictional notice & governing law. This dossier is governed by, and construed in accordance with, the laws of the Commonwealth of Pennsylvania, USA, without regard to its conflict-of-laws principles. Any dispute arising out of or relating to this material shall be brought exclusively in the state or federal courts located in the Commonwealth of Pennsylvania, and the reader consents to the jurisdiction of such courts. Readers are responsible for ensuring their access to and use of this material complies with the laws of their local jurisdiction. The dossier is not directed at any person in any jurisdiction where such access or use would be unlawful, restricted, or require registration.
Updates and corrections. The dossier is maintained on the cadence stated near the masthead. The version log near the bottom of this page records all material changes, including any corrections of prior errors. Documented self-corrections are a feature of the methodology, not an indicator of unreliability.
Contact, corrections & takedowns. Substantive corrections, source disputes, and formal intellectual property / DMCA removal requests are welcome and will receive immediate assessment. Please direct all official inquiries — including takedown notices pursuant to 17 U.S.C. § 512 — to claudflair.vwx82@dralias.com, with sufficient detail to identify the disputed material and the rights asserted. By accessing this page you acknowledge that you have read and accepted these terms.