The Quiet Power of an Edge
What a small, repeatable, Discord-logged advantage is actually worth — and the one honest condition the whole thing rests on. In plain English.
Signals are logged for observation and education. CTA registration (NFA / Series 3) is in progress. No capital is managed, and no investment is offered or solicited, before the structure and licensing gates clear.
THE MEASURED RECORD
Three Numbers
You don’t need a finance degree to evaluate this machine — you need three numbers. Measured, not promised.
-
62%
of trades win
about 62 of every 100 — 62.4% exactly
-
$1.69
collected per $1.00 lost
a 1.69 profit factor — won vs. lost, net of every cost
-
149
live trades
called live in our Discord channel — not a backtest
One sentence matters more than any number on this page: 149 trades is six months of one kind of market — promising, not yet proven. The back half of this page is how we prove it.
Every figure on this page is measured from one live futures engine — 149 trades called in real time in our Discord channel, Feb–Aug 2026, net of all costs. Past performance is not indicative of future results.
Check every figure against the live instrumentation These three numbers are a fixed snapshot of the first 149 trades. The dashboard reads the same journal, live, and keeps counting — so its totals will have moved past them.
The record so far
Six months, in the open
+41.0%
returned, net of all costs
−5.8%
deepest drawdown along the way
Both measured on the $20,000 portfolio the record was traded on — the same dollars are a different percentage on a different balance, so the balance is stated. The drawdown is the worst peak-to-trough on the realised curve, not a modelled one. That $20,000 is the balance this record happened to be traded on; it is not a projection for an allocation of the same size, and no allocation is offered.
- Feb 4 first logged call
- Aug 10 most recent
- Days live
- 187
- Trades logged
- 149
- Market regimes
- One
One window, one regime, one account. This is a realised result over 187 days, not a rate of return being offered or projected forward — six months is too short to annualise honestly, so we don’t. Measured from 149 trades called in real time in our Discord channel, Feb–Aug 2026, net of all costs. Past performance is not indicative of future results.
WHAT AN EDGE ACTUALLY IS
The House, Not the Gambler
An ‘edge’ isn’t a prediction. It’s a structural tilt — the casino’s is the cleanest on earth. We copied its four rules.
| Rule | THE CASINO | THIS EDGE — LIVE RECORD |
|---|---|---|
| THE TILT | THE CASINO ~5.3¢ of every $1 staked, by geometry | THIS EDGE — LIVE RECORD ~27¢ kept per $1 risked (+0.27R), net of all costs |
| THE WIN RATE | THE CASINO 52.6% of even-money spins go to the house | THIS EDGE — LIVE RECORD 62.4% of trades win |
| THE CAP | THE CASINO table limits no single bet can ever hurt the house | THIS EDGE — LIVE RECORD one fixed chip (R) per trade, set before entry |
| THE VOLUME | THE CASINO thousands of spins a night make luck irrelevant | THIS EDGE — LIVE RECORD ~291 trades a year |
Las Vegas was built on a nickel per dollar — this tilt measured ~5× sharper, with far fewer spins.
A casino’s tilt is geometry — fixed by the wheel. Ours is measured from 149 trades called live in our Discord channel, Feb–Aug 2026, net of all costs — markets adapt; hence the tripwires.
THE PAYOFF SHAPE
It Bends the Payoff
Most assets sit on one line — more reward means more risk. A real edge steps off it.
Conceptual diagram · unquantified axes
-
conventional frontier — more reward is paid for with more risk of permanent loss.
-
no edge / broken discipline — below the line: more risk of permanent loss, less reward.
-
THIS EDGE — above the line: Sharpe ~1.4–1.7, loss capped ~1R.
Bounded loss + positive expectancy = a payoff shape almost nothing else has. Conceptual placement, conditional on the edge being real and disciplined. The Sharpe estimate (~1.4–1.7) is trade-derived from 149 trades called live in our Discord channel, Feb–Aug 2026, net of all costs, and is conditional on the edge persisting — not a measured long-horizon result.
HOW RUIN IS ENGINEERED OUT
Built the Way Aircraft Are Built
Aviation engineered the crash ≈100× rarer since 1960. Ruin here is engineered out of the math before any trade — not ‘we try to avoid ruin,’ but a simulated number: 200,000 paths, five years, at the 2.5% per-trade ceiling.
<1in200,000
simulated 5-year risk of ruin at the 2.5% per-trade ceiling
Effectively zero — computed, not promised.
AVIATION ALREADY DID THIS
≈28 1960
~0.2 2024
-
LOAD LIMIT
How much stress before it breaks?
A wing is rated far beyond any storm — you never fly near the limit. Here: one capped chip (R) per trade, deliberate over-capitalization, a hard 20% ceiling on total risk at any moment, and drawdown caps of $1,000 in any week and $2,000 in any month on the $20,000 portfolio.
Those caps are enforced at the broker, and will be at Interactive Brokers where member accounts are opened. They were set after this record was logged — the record already sits inside them. Across the 149 trades the worst week lost $571 and the worst month $155. Neither cap was reached, let alone crossed.
-
THE BACKUPS
Will anything catch a failure?
Two independent checks that don’t share a failure mode — one on the instrument, one on the pilot. No single point of failure, including us.
-
THE SAFE DEFAULT
What happens automatically when something fails?
The failure state is smaller, not larger — by rule, not willpower. Nothing about recovery depends on anyone being heroic in the moment.
-
THE LIVING TEST
The question the sky never asks.
A wing’s physics never changes — certify it once, done. A market adapts. So the system keeps testing itself, every day, with kill-criteria that never retire. The first three keep the account alive — this one keeps it right.
Ruin and drawdown figures are Monte-Carlo simulations over 200,000 five-year paths, with win/loss outcomes drawn from the same 149 Discord-logged trades (Feb–Aug 2026, net of all costs) and conditional on the edge being real out-of-sample. Live results can exceed the model. The 2.5% ceiling is the rule the simulation is run at and the rule going forward; 10 of the 149 logged trades were sized above it, the largest at 5.0%, against an average risk of $261 (1.3%).
WHAT KILLS IT
Two Ways It Dies. Two Tripwires.
An edge can fade — or its operator can drift. Each failure mode has its own alarm, and neither is ours to silence.
EXPRESSION GATE
Is the edge still positive?
If rolling-100-trade expectancy crosses $0, new trading stops — research-only until the edge is re-proven out-of-sample.
FIDELITY GATE
Are we still flying by the checklist?
- Every trade is attributed.
- Every rule change is countersigned and disclosed before it runs.
- The daily routine is logged.
A tripped gate means smaller size or a full stand-down — by rule, announced within 24 hours.
INDEPENDENT WITNESS
Who can overrule us?
The call on a breach is not ours to make. A modification we want is not ours to approve. Failure is as visible as success.
Manish Dharod
Binding authority over modifications and breach calls
No competitor volunteers the failure rate of their own instrument class — ours is on this page. The kill switch fires on realized losses, not on anyone’s opinion of them.
THE CEILING
Capped in Dollars, Not People
Capacity is consumed by capital, not headcount. The ceiling is $5M.
Two rows, each standing for the same $5,000,000 of capacity at identical size. The first is divided into 166 equal slots, the most that fit at the $30,000 minimum allocation. The second shows the same room filled by a handful of larger allocations of uneven size.
At the $30,000 minimum,
166 fit.
The same room, arriving larger,
far fewer do.
Minimum allocation $30,000.
A handful of large allocations could close the door on everyone else.
Why a low ceiling is the point
Index funds scale forever because they have no edge. Sharp edges live in fleeting inefficiencies a large order erases — so a low ceiling is exactly what a real edge looks like. This edge is Medallion’s category, just over three orders of magnitude earlier.
RENAISSANCE MEDALLION
≈$10B · capped
THIS EDGE
$5M
≈ 2,000× smaller than Medallion
Capacity is the strategy’s own forward estimate, not a measured figure; Medallion’s ≈$10B is reported third-party context. The win rate, profit factor and tilt cited elsewhere on this page come from 149 trades called live in our Discord channel, Feb–Aug 2026, net of all costs.
Observation only — the pre-licensing status is stated in full at the foot of this page.
ONE EDGE, MANY EXPRESSIONS
The same edge, expressed three ways
The edge is the primitive. Everything below is a configuration of it — none of them exist yet, and none of them are offered here.
- A monthly cash-flow check
- Lumpy returns levelled into a regular distribution by a reserve.
- A portfolio overlay
- The edge run beside an existing market position rather than instead of it.
- A managed program
- The edge run for allocated accounts under CTA registration.
All three are gate-conditional and none is available today.
No figures for these configurations appear on this page, because none has been measured live.
FROM PROOF TO PORTFOLIO
Four stages, each with a gate
How a Discord-logged edge becomes a managed program. Dates we control are commitments. Dates regulators control are estimates — labelled as such.
-
Live now
1PROVE
Signals logged and timestamped in Discord — the audit trail, not the trading vehicle. The tripwires decide whether we advance.
-
Indicative: 60–90 days
2STRUCTURE
Execution runs through Interactive Brokers: members open their own account, and the vehicle is either a managed account (trade authorization granted at the broker) or copy trading (master-account mirroring). Capital stays in the member’s own name at the broker. Decision and criteria shared with the cohort.
-
Indicative: 90–180 days
3LICENSE
CTA registration path — NFA membership and Series 3 — with compliance counsel engaged before the first managed dollar. Regulator timing is an estimate, not a commitment.
-
Gate-conditional
4TRADE
Allocated accounts open, subject to both gates. Observation confers no priority and no claim on allocation.
No capital is managed, and no investment is offered or solicited, before the structure and licensing gates clear.
Managed-account and copy-trading execution of futures both constitute CTA activity; final structure and timing are subject to counsel review. The 60–90 and 90–180 day ranges are indicative estimates, not commitments.
THE ONLY OPEN POSITION
You don’t take the numbers on faith.
Every call is timestamped in Discord, in real time, before the outcome is known. Wins and losses post the same way.
During licensing, no one can allocate. Observation is the only open position.
Observation and education only. Joining is not an application, a subscription, or a claim on future allocation. There is no waitlist.
Signals are logged for observation and education. CTA registration (NFA / Series 3) is in progress. No capital is managed, and no investment is offered or solicited, before the structure and licensing gates clear.
Hiren Desai Founder & Chief Investment Officer hd@ekantikcapital.com (763) 482-1077