Ten G10 currencies give us 45 currency pairs. The number stays fixed. The concentration of their daily movements does not.
Full post on LinkedIn →
lnkd.in/p/duiWcD-i
1
16
FX liquidity is not the same as signal capacity.
Full post on LinkedIn →
lnkd.in/p/eS9jmUUJ
17
The historical numbers can be entirely accurate. Some sensitivity to costs and execution is expected in any signal. What matters is how much margin it leaves against variations that are plausible and relevant for the way it is meant to be used.
Full post →
lnkd.in/p/dnrDQrN3
1
33
A model can be promising on historical data. Understanding what happens when its assumptions meet real execution requires a different kind of work.
Full post on LinkedIn →
lnkd.in/p/e8UeGxeA
1
38
Hedging the same amount in a good week or a bad one changes the final cost. Hedge timing can be measured.
Full post →
lnkd.in/p/eQK8mv-2
29
A trading signal that only says buy or sell leaves out what matters most to an institutional desk: how strong the conviction is, and whether conditions justify taking the risk.
Full post on LinkedIn →
lnkd.in/p/e8xh4Qn8
46
The main obstacle to institutional adoption of AI in FX lies in evaluation before capital is committed, more than in the technology itself.
Full post →
lnkd.in/p/eXv8aRED
41
If the market regime changes tomorrow, how much of my portfolio will actually behave differently?
Full post →
linkedin.com/posts/sebastian…
40
Institutional quant finance isn't buying AI tools anymore. It's building its own AI labs.
Full post on Linkedin →
linkedin.com/posts/sebastian…
23
AI could displace 92M jobs by 2030 — while creating 170M new ones.
The real question isn’t whether AI will make decisions.
It’s how those decisions are designed.
lnkd.in/p/eRubVcgx
27
That correlation can break down exactly when you need it most is nothing new to anyone managing risk.
The real question is how you stop relying on a fixed historical correlation once the market shifts regime.
Full post →
lnkd.in/p/eiffN-2M
31
What happens when several leveraged funds need liquidity at the same time?
Full post →
lnkd.in/p/eKbHeHJH
26
Five years ago there was no market for an adaptive FX system. We built one anyway.
Full post on LinkedIn →
lnkd.in/p/etBDHQP8
44
In FX, most "alpha" is disguised beta.
Full post on Linkedin →
linkedin.com/posts/sebastian…
1
51
Capital is moving into systematic strategies faster than the industry itself is growing.
Full post on LinkedIn →
linkedin.com/posts/sebastian…
42
The real risk in automation is not that a system operates without a human. It is that it operates without being observable, constrained or interruptible.
Full post →
linkedin.com/posts/sebastian…
44
In systematic FX, the hardest problem is not fitting the past. It is remaining robust when the market regime changes.
Full post →
linkedin.com/posts/sebastian…
42



