Fed, BOJ, triple witching: the 5 checks before a crash
In plain English
This week the US central bank (Wednesday) and Japan's central bank (Friday) are both expected to raise interest rates. On the same Friday, about $6 trillion of options and futures expire. Some people think this combination will crash stocks.
My view: probably not, but the risk is higher than normal. Here is why in simple terms:
Both rate hikes are already expected, so they won't shock anyone. The surprise would be what the banks say about future hikes.
The thing that can actually hurt stocks is the US 10-year bond yield sitting at 5%. Higher yields make stocks less attractive and loans more expensive.
The Japanese yen is the wild card. If it strengthens fast, traders who borrowed yen to buy US stocks have to sell. Right now the yen is getting weaker, not stronger, so that risk is not live.
Friday's expiry doesn't pick a direction. It just makes whatever move happens bigger.
I have 5 things that all need to happen, in order, for this to become a crash. Today none of them have happened.
Bottom line: I'm positioned for a drop on the Nasdaq of roughly 1,250 to 3,000 points (5,000 to 12,000 ticks on NQ), which takes it from 29,200 down toward 27,900 at the shallow end and 26,200 at the deep end. That is a correction inside a bull market, not the end of it. I'll be buying into it, and I've written down exactly what would make me stop. Details below.
---
Fed Wednesday. BOJ Friday. $6T triple witching Friday. Three events in 72 hours, and the crowd is watching the wrong one. Here is how I'm reading it on the two charts that matter.
The setup
Both hikes are priced. Fed funds futures put a 25bp hike at roughly 80-87%. BOJ to 1.25% is described as almost fully priced. The hikes are not the event. The surprise has to come from guidance: the dot plot on Wednesday, or the BOJ's pace and JGB purchase language on Friday.
The real stress variable is the US 10Y . It closed 4.97% Friday and printed 5.017% intraday, right at the Oct 2023 cycle high. The 2Y is up 44bp since Aug 26. This is a bond-led repricing hitting a market with:
VIX at 15.8
SPX skew at the 1st percentile (nobody is hedged)
CTA and vol-control exposure rebuilt from the July lows
Buybacks going into blackout from Sep 12
A dealer gamma reset on Friday's expiry
Cheap protection plus a gamma reset into a two-hike week. That is the fat left tail. It is not the base case.
USDJPY (4H, left chart)
The naive thesis: two hikes = yen rallies = carry unwind = August 2024 again. The chart disagrees. USDJPY is 154.40 and rising into the BOJ meeting, not falling. The US-Japan 10Y differential is about 198bp and has not narrowed, because both bond markets are selling off together. Late-July intervention already squeezed the yen shorts. The marginal seller of yen today is a Japanese pension rebuilding foreign assets, not a hedge fund.
Levels on the chart:
163.99 = July high. The 40-year yen low was 162.8 on Jul 1, then intervention
160.39 = the shelf that broke in early Sep
155.23 = broken support, now resistance (blue zone). Price is testing it from below right now. 4H EMA 154.20
154.50 = intraday pivot
152.89 = Sep 8 low. This is the line. A BOJ hike that takes USDJPY below 152.89 within a day is the only outcome that puts the carry channel in play. Anything holding above 153 and the yen leg of the thesis is dead
MNQ (Daily, right chart)
Structure: HH at 30,975 (mid-Jun), HL at 27,200 (early Aug). Price 29,192, inside the range, below the 29,812 pivot, above the 28,927 pivot. The pink Area of Interest (30,250-30,750) is where the last rally failed. The green one (26,750-27,250) sits on the HL and the rising daily MA.
Target: 5,000 to 12,000 ticks lower. From 29,192 that is 27,940 at the shallow end (the HL retest) and 26,190 at the deep end (through the 26,400 level into the lower green zone). The chain below decides which end prints.
What I'm watching:
29,812 reclaimed = hike absorbed, range continues, no trade
28,927 lost on the Wednesday close = first real signal
27,940 = 5,000 ticks. Minimum target if checks 1 and 2 print. HL retest zone
27,200 HL = the line between a correction and something worse. Two daily closes below with HY spreads widening is the crash setup
26,400 to 26,190 = 11,000 to 12,000 ticks. Full target if the whole chain completes. Needs all five checks
The chain: 5 checks, in order, all must print
1. Wed: Fed hikes AND 10Y closes above 5.02%. No = 2-3% dip, done.
2. Thu night: BOJ hikes AND USDJPY breaks 152.89 within a day. No = 4-6% correction max.
3. Fri close: VIX above 25. No = two-day shakeout.
4. Fri open: expiry gap down not reclaimed by 10:30am ET. No = high-volume day, nothing more.
5. Mon-Wed next week: HY OAS above 320bp and two closes below the HL. Yes = this is the crash.
Today: 0 of 5. Friday's expiry is the only piece already in place, and it only matters if Wed and Thu both print. Expiration changes the size of the move, not the direction.
My bias
Short NQ into the week, targeting 5,000 to 12,000 ticks. Shallow target 27,940 if the Fed and BOJ both print (checks 1 and 2). Deep target 26,190 only if all five checks print. Entry is Wednesday after the Fed, not before; a short is wrong if NQ closes back above 29,812.
Constructive on a 12-month view. Q2 earnings grew 33%. Midterm-year Septembers average a 6% pullback that bottoms in October and runs into year-end. Anything that breaks this month is a reset inside a bull market, so the short is a rental. The drop is the buying window, not the trade.
I cover on the first of:
Target hit: half at 27,940, rest at 26,400 or the HL breakdown failing
Any check fails
VIX above 35 intraday (that is the panic peak, not the start; Aug 2024 topped above 60 and SPX was back in 3 weeks)
10Y back below 4.85% on a down day (bonds cushioning again)
Wed Sep 23 close, regardless
The second leg: real estate
CMBS delinquency is 7.55% (9.5% counting matured loans still paying interest). $1.1T of CRE debt matures in 2026-27 against a 5% 10Y. REITs were up 18% YTD into that. If the equity chain prints, listed REITs and CMBS reprice in weeks. Private marks and housing are a 2027 story. Order of entry when the reset comes: equities first, listed REITs second, private property last and only once the Fed has turned.
Not financial advice. Levels are for testing the thesis, not a signal service.
TITradingView Ideas14 Sept