Every strategy we're testing, and how far each has got
We do systematic trading: every strategy is written down in advance, then followed exactly, with no judgement on the day. That's what lets us test them. This page shows all of them, not just the good ones, using the same checks you'll find across this site.
No real money has been traded. Everything here is paper (a computer replaying the strategy on recorded prices) or sim (the strategy running on a real trading platform on a demo account). These are not real trades. Past results, real or simulated, don't predict future ones, and nothing here is advice or a recommendation.
Updated 14 Nov 2026. Results run to the week ending 13 Nov 2026. A week appears once it has finished.
98
strategies in the test
1
passed all 4 paper gates
2
running on sim
0
trading real money
The funnel
A strategy has to clear each step to reach the next, so a shrinking bar is the system working. Most strategies are supposed to fail here. That's what the test is for.
In the paper test98
Every strategy in the test, losers included. Many are variations on each other.
With 30 or more paper trades55
Enough trades to start judging. Fewer than this and a result means very little.
Then: beat the coin-flip test1
Flip each trade’s direction at random, thousands of times. The real result has to beat 80% of those.
Then: survive double costs1
Still in profit with trading costs doubled.
Then: not getting worse (all 4 gates passed)1
Hasn’t started doing worse than its own earlier record.
The 1 that passed everything are one strategy and a variation of it. They aren't the same as the 2 on sim, below: those are that strategy and a second one, which is still short of one gate.
On the platform
Passing on paper isn't enough. A strategy then has to run on the real platform and show that its fills and timing match the paper version. That takes at least 20 sim trades and 6 weeks, whichever is longer. Only then can it trade one small real contract.
Strategy #118 of 20 trades
On sim since 5 Oct 2026
Strategy #26 of 20 trades
On sim since 29 Oct 2026
The running book
This is the two strategies on the platform, added together, one point a week. It's measured in R: one R is the amount a trade planned to risk, so a trade that hits its stop loses about 1R.
−8.4R
paper result since the dashed line, 6 weeks
8.4R
biggest drop from a high, week to week
About zero
correlation: the two are roughly unrelated
See the weekly numbers
Week ending
Week
Running total
19 Jun
+2.1R
+2.1R
26 Jun
+4.8R
+7.0R
3 Jul
−0.1R
+6.9R
10 Jul
−0.2R
+6.7R
17 Jul
+3.3R
+10.0R
24 Jul
+2.4R
+12.3R
31 Jul
−0.0R
+12.3R
7 Aug
−0.5R
+11.8R
14 Aug
−0.2R
+11.6R
21 Aug
+1.9R
+13.5R
28 Aug
−1.0R
+12.5R
4 Sep
−1.5R
+11.0R
11 Sep
−1.1R
+9.8R
18 Sep
+0.3R
+10.2R
25 Sep
−0.1R
+10.1R
2 Oct
+3.7R
+13.8R
9 Oct
−0.9R
+12.9R
16 Oct
−1.7R
+11.2R
23 Oct
+2.6R
+13.8R
30 Oct
−2.7R
+11.0R
6 Nov
−2.1R
+8.9R
13 Nov
−3.5R
+5.4R
What this doesn't mean. Both strategies were chosen after we'd seen most of this curve, so everything left of the dashed line is hindsight, and that includes most of the +5.4R total over all 22 weeks. The number to watch is the one since the line, and it needs many more weeks before it means much. Lower drops are better, but a small drop over a few months says little about how bad a bad year could get.
The same book, trade by trade
One point every 5 trades instead of every week, with no dates. It shows how rough the ride was without showing when any trade happened. It stops at the last full 5 trades, so it can end a little away from the total above. It's paper too, and the same hindsight applies.
The rules of this page
Losing weeks and months get published the same as winning ones.
A strategy that fails stays on the board. Its number is never reused.
Strategies are numbered, and how they work isn't shown. The method is shared; the strategies aren't.
You'll only ever see totals. Dated charts are weekly at the finest, and the finer chart has no dates. No single trades.