THE SYSTEM ANATOMY | EPISODE 5: THE ENSEMBLE APPROACH
We spend an extraordinary amount of effort searching for the best trading system.
Perhaps that is the wrong search.
Because no single rule sees the whole market. Every system has a point of view. It sees some trends early
THE SYSTEM ANATOMY SERIES | EPISODE 4 OF 7
Fifty markets can still be one trade.
Diversification is easy to count and much harder to understand.
Markets were never independent. They are different expressions of the same coupled system, and when the storm arrives
THIS WEEK IN TREND | 21 AUGUST 2026
The TTU Trend Barometer jumped 11 points to 50%. Bitcoin surged 21.86%. Commodities rallied together.
It looked like a powerful week for trend following.
But the SG Trend Index barely moved.
JULY 2026 | THE TRENDS THAT WOULD NOT HOLD
July had no shortage of movement.
The problem was that nothing wanted to keep going.
Trends emerged, strengthened, invited participation, then broke. The TTU Trend Barometer told the story:
SYSTEM ANATOMY 3 | THE SURVIVABLE POSITION
The best position size is not the one that maximises what you make when you are right.
It is the one that ensures you are still around after being wrong. Again. And again.
THE SYSTEM ANATOMY SERIES | EPISODE 2 OF 7
Exit Logic
Everyone wants to know when to get in.
But the real architecture of trend following reveals itself in what happens next.
One rule makes being wrong survivable. Another gives being spectacularly right enough room to matter.
A stock is up 19% over three months. Its 2x leveraged ETF is down 27% over exactly the same period.
Same underlying. Same three months. Very different journey. Very different destination.
Arithmetic tells you what should happen.
Great July note from @ConvexityDredge. :-)
The July Update is up at the Convex Strategies website - "Rogers' Paradox". We draw on cultural anthropology and discuss the importance of individual learning in keeping systems fit and adaptable.
https://t.co/09PipuCuDe
THE SYSTEM ANATOMY SERIES | EPISODE 1: ENTRY LOGIC
Every trade begins with a decision.
But perhaps we have been asking that decision to do far too much.
We want the entry to separate signal from noise. To recognise the genuine move. To keep us out of the failures and somehow
THIS WEEK IN TREND | 14 AUGUST 2026
Energy surged. Grains strengthened. Trend strength improved.
But the biggest moves were not necessarily the ones trend followers wanted.
SYSTEM ANATOMY | OPENING THE MACHINE
A trading system can have the right philosophy and still fail in the mechanics.
Entry. Exit. Position sizing. Portfolio construction. Ensembles. Rollover. Survival.
Seven components. Each solving a different problem. Each depending on the
THE COMPOUNDER’S DELIGHT | What Did All That Estimation Buy?
I wanted to test a simple question.
If we continually estimate the edge and use it to determine the mathematically optimal amount to bet, does all that extra intelligence actually improve the result?
CARVED BY IMPOSSIBILITY | The Empty Chair
The cushion still holds his shape. He died on my birthday in 2017, and the question that has never left me is the simplest and hardest of all: where did he
https://t.co/P3wqmV3wNG…
#CarvedByImpossibility#Grief#Philosophy#Mortality
RISK IS NOT VARIANCE | The Series Concludes
Five episodes. 68 global futures markets. 647,629 market-days. More than four decades of evidence.
We began with one simple question:
If variance really measures risk, what should we observe in real markets?
This Week in Trend: 7 August 2026
Metals swept all five contracts higher, the Equity Indices reached new highs and breadth widened to 34 of 49 contracts,
Full report: https://t.co/Wj0GmaqwEk…
#TrendFollowing#SystematicTrading#GlobalMacro#TTUTrendBarometer
Diversification Does Not Cure Fat Tails
Risk Is Not Variance, Episode Five
We set out to end this series by recommending a better ruler. Stop squaring, use a measure that does not magnify the extremes, and the problem goes away. We tested it across 68 markets and four decades.
The Unwritten Sequence: an addendum to Episode 4
Ten portfolios. One risk report.
Compound return. Standard deviation. Sharpe. Sortino. Skew. Value at Risk. Expected shortfall. Identical across all ten to fifteen decimal places.They start at the same value and close at the same