eli terminal — March 15, 2026

The Twelve Hands

Every forced actor, every mechanical obligation, every collision point in the next 30 days
March 15, 2026 · 08:40 UTC · FOMC in 3 days · Triple Witching in 5 · Earnings in 29
"Nobody is doing what they want. Everybody is doing what they must." — The state of every actor in this market, March 2026

This report is different. It's not an analysis of one theme — it's a map of every mechanical obligation in the market over the next 30 days. Twelve actors. None of them choosing. All of them forced. Each action creates the conditions that force the next actor's hand. The game board looks like free will. It's clockwork.

Inversion Theory says: map the forced responses. The game tree is smaller than it looks. This report takes that literally. Every hand below is a MUST, not a MIGHT. The question isn't what they'll do — it's what happens when twelve "musts" collide in the same six-week window.

The Buyers (Forced Long)

1
The 401(k) Machine
MUST BUY ~$3.6B/day in equities
$19.3T in passive funds · $903B/year net inflows · 160M workers contributing biweekly
Continuous. Every two weeks. Cannot stop unless employment collapses.

The floor. Target-date funds allocate 60-80% to equities automatically. No human decides. The money enters payroll → 401k → Vanguard/Fidelity → index → market. It buys at $695 and $662 and $600 and $500 with equal conviction. It's the reason this market can't crash (#88). It's also the reason it can't be rescued — the floor descends with the market.

2
The SPX Spec Shorts
MUST COVER -358,096 contracts · Covering +53K/week
~$36B notional short position · Already covered +119K from -477K trough
Continuous. Accelerates into FOMC (risk management) and OpEx (gamma unwind).

The most uncomfortable trade on the board. Net short 358K contracts at VIX 27 — every morning risks a short squeeze. They've been covering at +53K/week, buying $5.3B in SPX exposure. But SPY is still falling. The covering is being absorbed by sellers further up this list. These are forced buyers whose buying isn't working.

3
The Treasury Auction Bidders
MUST BUY $1.8T/year in new Treasury supply
10Y auction Mar 11: 4.217% yield, B/C 2.45, indirect bidders 74.3%
~$150B/month in issuance. Next major auctions: 20Y Mar 19 (day after FOMC).

Foreign central banks (74.3% indirect) MUST buy Treasuries because they hold dollar reserves and need to roll maturing holdings. But the price they're willing to pay keeps falling — the 10Y at 4.22% is the highest in months. Each auction is a forced purchase at a worse price. The bid-to-cover at 2.45 is adequate but not enthusiastic. The 20Y auction the day AFTER FOMC is a collision point — if the dot plot is hawkish, the auction lands in a falling bond market.

The Sellers (Forced Short)

4
The Put Holders (1.85M contracts)
MUST RESOLVE by March 20 (triple witching)
1,848,914 SPY puts expiring Mar 20 · P/C ratio 2.59 · Max pain $682 ($20 above spot)
Decision point: March 18-19. Roll, exercise, or let expire. 48 hours to decide.

Nearly 2 million put contracts, each controlling 100 shares. 185 million shares of downside insurance expiring in five days. The dealers who sold these puts are delta-hedging by holding short stock. After expiration, the hedge unwinds — creating a mechanical buying impulse. BUT: if holders roll their puts to April, the selling pressure transfers forward instead of releasing. The Mar 18 FOMC decision determines which path: hawkish → roll → selling continues. Dovish → expire → buying impulse.

5
Active Fund De-risking
MUST SELL $189B/year net outflows from active funds
$17.4T in active strategies · Continuous redemptions
Accelerates at quarter-end (March 31) for window dressing.

Active fund managers are in secular decline — passive has overtaken active at $19.3T vs $17.4T. Every dollar leaving an active fund creates a sell order. At -$189B/year, that's $750M/day of forced selling. It's not a view on the market — it's a structural migration from active to passive. The managers leaving aren't bearish; they're obsolete. The selling is the obituary of a business model, not a market call.

6
Risk Parity Deleveraging
MUST SELL stocks AND bonds when correlation flips positive
~$200B in risk parity strategies · RPAR ETF -3.3% (1mo)
Triggered when SPY/TLT correlation sustained above zero. Currently +0.13 and climbing.

When stocks and bonds move together, risk parity's total portfolio risk more than doubles (IMF). The response is mechanical deleveraging: sell both sides until risk is back within mandate. RPAR -3.3% in one month is the visible signal. The invisible signal: Bridgewater's All Weather fund (~$150B) adjusting leverage that flows through swap dealers into futures markets. This selling appears in no COT report because it's done through OTC derivatives.

7
Foreign Central Banks
MUST SELL Treasuries to raise USD for oil imports
China + Japan hold $2.1T in Treasuries · DXY +3.7% (1mo)
Continuous. Intensifies when oil rises (more USD needed).

Oil is priced in dollars. When oil is $99, importing nations need more dollars. They get dollars by selling the dollar-denominated asset they hold: Treasuries. This creates the paradox: foreign demand at Treasury auctions (74.3% indirect) coexists with foreign selling in the secondary market. They're buying new issuance at auction (must roll) while selling older holdings for USD liquidity. Net effect: bond prices fall despite "strong" auction results.

The Frozen (Forced to Do Nothing)

8
Corporate Buyback Desks
MUST NOT BUY — Earnings blackout window
$4.8B/day in demand removed · 80-90% of S&P 500 in blackout
March 15 through late April. 6-week absence.

The only actor whose forced action is inaction. $1.2T annualized, 8.4% of daily SPY volume, 100% of net equity purchases since 2000 — all of it frozen. The autoimmune system in scheduled shutdown. Returns when Q1 earnings are reported, but may return at lower pace if companies are preserving cash ($99 oil hitting margins).

9
Gold Specs
FROZEN at +98,399 contracts · Change: +482 (effectively zero)
Position unchanged for 6+ weeks · Neither adding nor cutting
Resolves when USD direction becomes clear post-FOMC.

The most interesting non-action on the board. Gold specs are sitting on a massive long at +98K contracts and haven't moved in six weeks (+482 change is rounding error). They're waiting for a signal. If FOMC is dovish → gold rallies → they add. If hawkish → gold falls → they're trapped and forced to sell. The frozen hand becomes a forced hand the moment Powell speaks.

The Catalysts (Forced to Signal)

10
Powell + FOMC (March 18, 2:00 PM)
MUST PUBLISH dot plot, statement, and press conference
Modal expectation: 1 cut in 2026 (30%). Zero cuts: 24%. Two cuts: 24%.
March 18, 2:00 PM ET. 72 hours from now.

Powell can't hide. The dot plot forces 19 FOMC members to commit their rate projections to paper. Waller and Miran will dissent dovish (73% and 97% prediction market odds). The median dot is the signal that reprices everything downstream. The market prices 1 cut (30%) with near-equal odds of zero (24%) and two (24%) — maximum uncertainty. Whatever the dot plot shows will instantly reprice the rate path, which reprices bonds, which reprices risk parity leverage, which reprices equities. This is the domino that starts the chain.

11
Oil Specs (Short at $99)
SHORT -28,145 contracts · Adding -11K more last week
Commercials opposite: +114,697 (LONG). 143K disagreement.
Resolves on geopolitical event or demand data. Continuous short squeeze risk.

Specs are SHORT oil at $99 and adding to the position. Commercials — the people who physically handle crude — are massively long (+115K). A 143K-contract disagreement between financial traders and physical operators. If oil goes to $110, the specs face $1.1B in losses per $1 move. They're betting on a ceasefire or demand collapse. The physical market disagrees. This is either the smartest trade in the market or a short squeeze waiting to happen.

12
Margin Borrowers ($1.28T)
MUST SELL if portfolio falls below 25% maintenance
$1.28T margin debt (8th consecutive record) · Credit balance -$878B (record low)
Triggered at different levels depending on leverage. 6x funds already in warning zone at -4.8%.

The sleeping hand. Retail 2x leverage is safe until -33%. But hedge funds at 6x leverage have lost 29% of equity at the current -4.8% drawdown — WARNING territory. At -10% from peak (SPY ~$626), 6x funds lose 60% of equity. That's forced liquidation. The danger zone is SPY $626. At the current -0.14%/day pace, that's ~36 trading days away (late April). It arrives during bank earnings week.

The Collision Map

Net Forced Flow by Direction — Buyers vs Sellers vs Frozen
Collision 1: March 18 (FOMC) → March 20 (Triple Witching)

Powell publishes dot plot. Within 48 hours, 1.85M SPY puts must resolve. If dot plot shows ≤1 cut (most likely), puts get rolled forward → selling pressure persists into April. If dot plot shows 2+ cuts (surprise), puts expire → dealer hedges unwind → mechanical buying impulse. Gold specs (+98K) unfreeze in the direction of the dot plot. Euro specs (+5K, collapsed from +50K) either rebuild or capitulate to zero. Everything depends on the median dot.

Collision 2: March 31 (Quarter End)

Fund managers MUST window-dress. Active funds selling losers (tech, consumer discretionary). Pension funds rebalancing — with both stocks and bonds down, they're selling whichever fell less and buying whichever fell more. But in a "nowhere trade" where everything fell, the rebalancing generates selling across the board. Quarter-end flows collide with buyback blackout — no corporate bid to absorb.

Collision 3: April 13-16 (Bank Earnings) → Late April (Tech Earnings)

JPM, GS, WFC, BAC, MS report Q1 — built on assumptions of $60 oil, 3.50% Fed funds, and no Hormuz crisis. Actual: $99 oil, 4.27% 10Y, Hormuz closed. Loan loss provisions for CRE (12.3% delinquency), consumer credit (4.8% CC delinquency, $1.28T CC debt). If banks guide down, margin borrowers at 6x leverage ($1.28T margin debt) enter forced liquidation territory. Bank earnings ignite the margin cascade that the slow bleed prevented from happening organically.

The Force Balance Sheet

Daily Forced Buying
~$8.9B
401k $3.6B + Spec cover $5.3B
Daily Forced Selling
~$5.5B+
Active $0.75B + Risk parity + Foreign CBs
Removed Buyer
-$4.8B
Buyback blackout
Net Balance
~-$1.4B
Slight seller advantage → slow bleed

The daily arithmetic explains the -0.14%/day pace. Forced buying ($8.9B from passive flows and spec covering) slightly exceeds visible forced selling ($5.5B from active outflows, risk parity, and foreign CBs). But the removed buyback bid ($4.8B) tips the balance to net selling of ~$1.4B/day. At $57B average daily volume, $1.4B of net selling produces a -0.14%/day decline.

This is the mechanical explanation for the slow bleed. It's not fear. It's not narrative. It's not a view on the economy. It's twelve hands, none of them choosing, producing a flow imbalance of $1.4 billion per day that grinds SPY down at exactly the pace we observe.

Cross-Asset COT Positioning — Who Is Forced Where

The Game Theory

Twelve hands. Seven are selling or frozen. Five are buying. The buyers are unconditional (passive) or uncomfortable (spec covering). The sellers are structural (active migration, risk parity, foreign CBs) or absent (buybacks). The net: -$1.4B/day, producing -0.14%/day.

The only hand that can change the balance is #10 — Powell. The dot plot on March 18 is the single input that reprices every other hand's obligation. A dovish surprise (2+ cuts) unfreezes gold specs into buying, triggers put expiration (releasing dealer hedges into buying), signals that cash yields will fall (pulling $8.27T from money markets toward risk assets), and might even pause risk parity deleveraging (if bonds rally, correlation flips negative).

A hawkish confirmation (≤1 cut) locks in the current force balance for another three months. Puts roll. Gold specs liquidate. Cash stays in money markets. Risk parity keeps selling. The slow bleed continues at -0.14%/day until it reaches the margin cascade trigger at SPY ~$626 in late April.

Twelve hands. One switch. Seventy-two hours.