@NightshiftAILab

AI proposes. Risk rules decide. Every trade, rejection and result recorded. Building a trading lab in public. Paper trading now.

nightshiftresearch.com
Joined November 2021
  • Tweets40
  • Following0
  • Followers12
24Media
NIGHTSHIFT · PAPER · 2026-10-20 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: $0.00 net · 0 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
2
16
NIGHTSHIFT · PAPER · 2026-10-19 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: $0.00 net · 0 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
2
27
NIGHTSHIFT · PAPER · 2026-10-18 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: $1.02 net · 1 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
2
22
NIGHTSHIFT · PAPER · 2026-10-17 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: -$5.51 net · 2 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
1
25
NIGHTSHIFT · PAPER Worst unposted close · prior 12h SOL · Aggressive Net $-2.99 · -2.04% of entry equity After modeled costs; funding/ops excluded. #8109 · Full journal: nightshiftresearch.com
1
1
17
NIGHTSHIFT · PAPER Best unposted close · prior 12h ETH · Aggressive Net $0.20 · 0.14% of entry equity After modeled costs; funding/ops excluded. #7406 · Full journal: nightshiftresearch.com
1
1
10
NIGHTSHIFT · PAPER · 2026-10-16 UTC Daily recap Conservative: -$1.72 net · 1 closes Aggressive: -$6.54 net · 3 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
2
16
NIGHTSHIFT · PAPER Worst unposted close · prior 12h ETH · Aggressive Net $-2.68 · -1.80% of entry equity After modeled costs; funding/ops excluded. #6654 · Full journal: nightshiftresearch.com
1
1
6
NIGHTSHIFT · PAPER Best unposted close · prior 12h ETH · Aggressive Net $-3.26 · -2.08% of entry equity After modeled costs; funding/ops excluded. #6035 · Full journal: nightshiftresearch.com
1
1
4
NIGHTSHIFT · PAPER · 2026-10-15 UTC Daily recap Conservative: -$1.11 net · 1 closes Aggressive: $2.20 net · 4 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
1
4
NIGHTSHIFT · PAPER Worst unposted close · prior 12h SOL · Aggressive Net $-3.29 · -2.06% of entry equity After modeled costs; funding/ops excluded. #5266 · Full journal: nightshiftresearch.com
1
1
11
NIGHTSHIFT · PAPER Best unposted close · prior 12h ETH · Aggressive Net $8.58 · 5.68% of entry equity After modeled costs; funding/ops excluded. #4850 · Full journal: nightshiftresearch.com
1
6
NIGHTSHIFT · PAPER · 2026-10-14 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: -$2.23 net · 4 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
1
11
NIGHTSHIFT · PAPER Worst unposted close · prior 12h SOL · Aggressive Net $6.20 · 4.14% of entry equity After modeled costs; funding/ops excluded. #3885 · Full journal: nightshiftresearch.com
1
7
NIGHTSHIFT · PAPER Best unposted close · prior 12h ETH · Aggressive Net $-3.62 · -2.36% of entry equity After modeled costs; funding/ops excluded. #3374 · Full journal: nightshiftresearch.com
1
2
9
NIGHTSHIFT · PAPER · 2026-10-13 UTC Daily recap Conservative: -$1.98 net · 1 closes Aggressive: -$5.19 net · 6 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
1
3
11
NIGHTSHIFT · PAPER Worst unposted close · prior 12h SOL · Aggressive Net $-3.69 · -2.24% of entry equity After modeled costs; funding/ops excluded. #2199 · Full journal: nightshiftresearch.com
1
2
14
NIGHTSHIFT · PAPER Best unposted close · prior 12h SOL · Aggressive Net $9.84 · 6.21% of entry equity After modeled costs; funding/ops excluded. #1532 · Full journal: nightshiftresearch.com
1
2
20
NIGHTSHIFT · PAPER · 2026-10-12 UTC Daily recap Conservative: $0.00 net · 0 closes Aggressive: -$22.49 net · 4 closes Modeled costs included; funding/ops excluded. Full journal: nightshiftresearch.com
2
10
New experiment added: arbitrage. Sounds simple. Buy an asset somewhere cheaper, sell it somewhere more expensive, and keep the difference. I want to see how often that difference survives costs and timing—and let everyone inspect the evidence. Here’s how it works:
1
1
2
19
more replies
The scanner uses no additional AI calls, and we keep storage bounded to control costs. This is another question we’re putting into public view: how much apparent arbitrage survives a closer look?
1
2
If the opportunities disappear after costs, that’s useful evidence. If something survives consistently, we have something concrete to investigate further. Follow @nightshiftailab to watch the experiment develop. Every improvement should make the evidence easier to understand.
1
Load more