Déjà Vu Trade
August 8, 2026  ·  Week 33 of 2026
DOW · TORONTO · TAIPEI Fires Thursday, Aug 13

The Mid-Summer Base

Three index longs fire at Thursday's close — the Dow into January, and the book's first two international markets: Toronto into January, Taipei into February. Each has won 20 of 25 years and 9 of the last 10. When the same date works in three different markets, the driver is the calendar.

The book going into Week 33. The soybean short (Jul 9 → Aug 8) completes its window this weekend — the first second-half position to run its full course. Everything else is unchanged, plus Thursday's new heating oil leg:
The Setup

Three index longs · Aug 13 at the close

AssetWindowHoldHit rateLast decadeMedian
INDUAug 13 → Jan 12~5 mo20 / 259 / 10+6.6%
TSXAug 13 → Jan 12~5 mo20 / 259 / 10+5.4%
TWSEAug 13 → Feb 11~6 mo20 / 259 / 10+5.8%

Mid-August is the base camp of the equity year's second half. The dog days — this week is typically the lowest-volume window of the entire year — carve out a seasonal trough, and from it the long autumn-winter current begins. On August 13 the screen puts three index longs on the tape at once: the Dow Jones Industrial Average and Canada's S&P/TSX Composite both into January 12, and Taiwan's TAIEX on a six-month window into February 11. Three markets, three continents, one date. When a pattern repeats across unrelated exchanges on the same calendar day, the driver is not any single market's story — it is the rhythm of the trading year itself.

The records are nearly identical, and that is the tell. Each leg has won 20 of 25 years and 9 of the last 10. More telling still: they lose together. The shared red years are 2008, 2015, and 2018 — global macro shocks, not local failures — and the single miss of the last decade is the same year for all three (2018's fourth-quarter selloff). Read that honestly: this is one bet on second-half risk appetite, expressed three ways. Size the trio as a single position, not three independent ones.

What the book actually gains here is geography. We are already long U.S. equities three ways — the Late-May Launchpad, the June growth leg, and the Independence Day Launch. The Dow leg is close kin to those; if you carry the earlier entries, it adds a value-and-industrials tilt but little true diversification. Toronto and Taipei are the genuine additions. The TSX brings resources and financials — a sector mix nothing on the book currently touches — and tends to ride both the equity current and the autumn firming in energy. The TAIEX is the world's most semiconductor-heavy major index; its window captures the autumn electronics build and runs through the seasonal strength that precedes Lunar New Year. Its tail is longer in both directions (+37.9% in 2025, −37.9% in 2008), which is what a chip-cycle market looks like.

Vehicles: DIA or /YM futures for the Dow; EWC for Toronto; EWT for Taipei. Note the international ETFs are unhedged — a long in EWC or EWT carries the Canadian dollar and Taiwan dollar along with the index, which has historically helped as often as hurt but is part of the position. For defined risk, January-dated calls fit the Dow and Toronto windows; February-dated for Taiwan.

What this is not: protection. In 2008 the three legs lost 26%, 31%, and 38% respectively — in a genuine crash they are one trade, and it is the wrong one. The four-in-five base rate is the edge; the shared tail is the cost of it.

The two weeks ahead

DayNotesTrades active
MonAug 10Week 33 opens — dog days, thin tape. A quiet week to stage entries, not chase them.
ThuAug 13 The Mid-Summer Base fires at the close — three index longs, one date.
LONGINDU+6.6%
LONGTSX+5.4%
LONGTWSE+5.8%
Exit Jan 12 / Jan 12 / Feb 11
ThuAug 20Nikkei short exits at the close. Jackson Hole symposium window opens — the event risk of late August.
FriAug 21August opex — third Friday.
SatAug 22Yen long completes its window; the summer risk-off pair is fully off the book.

Jackson Hole (late August) is the one scheduled event inside the entry window's first fortnight. The pattern's record already contains 25 of them — it needs no forecast, only the discipline to hold through the headline.

Full history — 25 years, three markets

LONG P&L. INDU and TSX: Aug 13 close → Jan 12 close. TWSE: Aug 13 close → Feb 11 close. The losses cluster in the same years — one macro bet, three expressions.

YearINDUTSXTWSE
2001-4.1%+0.9%+31.1%
2002+3.6%+4.2%-4.1%
2003+13.1%+13.7%+18.6%
2004+7.6%+10.9%+12.4%
2005+3.4%+8.5%+3.8%
2006+13.2%+6.1%+19.6%
2007-4.8%+1.5%-14.2%
2008-26.1%-31.0%-37.9%
2009+13.1%+9.2%+5.8%
2010+14.1%+16.8%+9.1%
2011+10.7%-2.1%+2.9%
2012+2.1%+6.0%+6.3%
2013+6.4%+8.7%+5.6%
2014+5.9%-6.5%+2.5%
2015-5.1%-13.1%-3.0%
2016+7.4%+5.0%+5.6%
2017+18.0%+8.5%+0.4%
2018-4.7%-8.1%-6.9%
2019+9.7%+5.4%+12.6%
2020+10.8%+8.2%+24.7%
2021+2.2%+4.3%+7.8%
2022+1.3%+0.2%+1.9%
2023+6.6%+2.9%+9.0%
2024+6.6%+10.6%+6.8%
2025+10.4%+17.4%+37.9%

Read. Since 2019, every leg has finished green every year — twenty-one consecutive winners across the three markets. Before that, the misses are years you remember from the front page, not from the chart. Enter at Thursday's close, size the trio as one position, and let the second half do what it has done four years in five for a quarter century.

← Last post: The Autumn Distillate Build Next: The Late-Summer Growth Re-Entry →

Get the next signal by email.

Posts arrive when a new trade is about to fire — not every week, only when something matters.