ADA-USDT perpetual futures trading bot · one instrument, one venue

Name: APEX
Brand: Apexstoa
Price: 399.00 USD
Availability: InStock

APEX

Six and a third years, 6,576 trades, profitable in every calendar year, after real venue fees, real funding and a measured entry-queue cost. Everything below is regenerated from the backtest that ships with it; nothing here is a figure you have to take on trust.

Get APEX · $399 ↗ Compare with CARRY

Six and a third years

$1,000 → $20,513, uncapped

all positive

+140.2%

Best year, 2022

+7.6%

Worst year, 2025: still profitable

6,576

Trades behind the record

APEX, 2020 through 2026, after every measured cost, trading every signal with no limit on open positions. The software ships at eight, and the three settings below show what that dial does. The harness ships with the strategy and regenerates the headline figures from raw market data, so none of them is a screenshot you have to take on trust.

APEX

ADA-USDT perpetual · one instrument, one venue

2020–2026 · 6,576 trades

Backtest equity, log scale · Jan 2020 – May 2026 · $1,000 start, 0.5% risk per position, no limit on open positions · net of venue fees, real funding and measured adverse selection

+0.1431 RNet edge per trade, after all measured costs

4.84t-statistic, undeflated: the best of a 10-value reward-to-risk sweep

6,576Fills across 6.3 years

61.2%Compound annual return, on the uncapped backtest

−71.9%Maximum drawdown, uncapped

0.85Return over drawdown (MAR)

7 / 7Calendar years profitable, uncapped. Lowest year +7.6%

3.38Walk-forward t, train 2020–23 → test 2024–26, on the base short-drift, not this configuration’s full-sample figures, and priced at maker-0 fees. Charged this product’s own costs, the same test half is t 2.42 (+0.1138 R).

Three settings the software runs: you choose the risk and how many positions stay open

Maximum

0.5% · 12 positions

$10,000 became: $195,626
Compound return: 60.0% / yr
Worst drawdown: −71.6%
Worst month: −29.0%
Worst year: +8.8%

Growth

0.5% · 8 positions

$10,000 became: $138,192
Compound return: 51.4% / yr
Worst drawdown: −69.1%
Worst month: −29.0%
Worst year: +3.1%

Conservative

0.25% · 8 positions · ships

$10,000 became: $53,499
Compound return: 30.4% / yr
Worst drawdown: −42.0%
Worst month: −15.6%
Worst year: +7.3%

All three settings finished every calendar year profitable. Sizing is the dial you turn for comfort: risk and position count move return and drawdown together, and both are values in the config file.

The drawdown profile, in full

486 dLongest stretch below a previous equity high

−29.0%Worst single month

−71.9%Worst peak-to-trough drawdown

52%Of months finish positive

80Longest run of losing trades in a row, about four weeks

Worst drawdown −71.9%, in September 2025, at which point the account still held six and a half times its starting deposit. Against the original balance rather than the peak, the deepest it ever went was −39.6%, in the first four months, and it spent 91% of the record above where it started. Starting at a high-water mark gives you no such cushion, which is what the −71.9% is there to tell you.

Sizing is the dial, but it does not scale the way people expect. At half the risk per trade the returns roughly halve; the drawdowns do not. At 8 concurrent positions, halving the risk takes the worst drawdown from −69.1% to −42.0%, not to −34.6%. Read the row you would actually run rather than interpolating between them; every figure here is published at both settings. The full statistical profile (win rate, losing runs, and the deflation of the headline) is set out in the pre-sale agreement you accept before buying.

Availability Limited. Access closes when capacity is reached

Why capacity is finite.

The record

(every figure, from one backtest)

Every figure in the charts below comes from raw market data through the harness that ships with the strategy, not from a screenshot. Charts 06 to 08 are drawn by tools that ship too.

01/08

Account equity

Log scale, 2020 – 2026

$1,000 grows to $20,513 · 61.2% compound per year

02/08

Underwater

Every point below a previous high

Worst drawdown −71.9% · longest recovery 486 days

03/08

Return by year

Every calendar period profitable

2020 +21% · 2021 +126% · 2022 +140% · 2023 +41% · 2024 +55% · 2025 +8% · 2026 +33%

04/08

Month by month

75 months, up and down

39 up, 36 down · best +67.7% · worst −29.0%

05/08

Was the order lucky?

1,500 reshuffles of the same 6,576 trades

The equity curve says what happened; it cannot say whether the SEQUENCE was fortunate. So the same trades were dealt in a different order 1,500 times, through the same sizing walk and the same real concurrency. The published −71.9% is deeper than every one of them. Real losses cluster, and the headline figure is the pessimistic end rather than a lucky draw. The same clustering shows at the end point: because overlapping positions are sized off equity at entry, the shuffled orderings also finished higher than the real one.

06/08

The shape of the payoff

Every trade, in units of the risk taken

This is what a 21% win rate looks like from the inside, and it is the answer to why the number is not a problem. Four trades in five stop out for about −1R: the tall cluster on the left. The fifth runs to the right, and the tail is long enough to pay for the other four with +0.1431 R left over. A strategy built this way is wrong most of the time by design; the average is what matters and the average is positive.

07/08

Any twelve months, from any starting point

Rolling twelve-month return, every starting point

Every day in the record, asked the same question: what did the next twelve months return from here? 82% of starting points finished the year ahead. The worst of them is the drawdown you have already seen, viewed from a different angle: the late-2024 to September-2025 stretch, measured from its own beginning rather than from a peak.

08/08

The risk surface

Positions at once × risk per trade → max drawdown

Three-dimensional surface of maximum
drawdown against concurrency cap and risk per trade, bending steeply downward as
risk rises

Sizing is the one dial you turn, and it does not scale the way people expect. With no limit on open positions, at 0.25% the worst drawdown is −44.6%; doubling to 0.50% gives −71.9%, not −89%; doubling again gives −94.3%. Positions overlap in time, so the bets are not independent and the surface bends. Every point is the full walk over every trade the cap admits at that pair: 361 complete runs, not an interpolation. A tighter cap turns fewer signals into trades: 6,576 with no limit, 4,342 at two.

(you set the exposure)

How many trades at once is your decision.

Each open position carries 20% of your equity in notional at the 0.25% setting that ships, and 40% at 0.5%, so concurrency is really a leverage dial. Move it and see exactly what it costs and buys. Every figure is read off the same backtest.

Positions open at once 8

Return against cap, flat past about ten

Risk per trade

0.25% is the setting the download ships with

Signals captured: 98.5%
Compound return: 51.4% / yr
Worst drawdown: −69.1%
Worst month: −29.0%
Gross leverage: 3.2×
If every stop hits at once: −4.0%
Adverse gap that liquidates: 30.8%

At eight positions you take almost every signal at 3.2× leverage, and it would take a 31% gap through your stops to end the account.

Licensing

Access is limited on purpose.

Capacity is finite for a structural reason: the strategy works a single instrument, and its economics depend on queue position at a price level that more capital only crowds.

The same level, the same second

Every copy acts on the identical public data at the identical moment and competes for the same fills. They aren't independent traders. They're one queue.

Queue position is the edge

Filling at the front of the queue and filling at the back are worth several basis points to each other. Each additional copy ahead of you pushes you toward the back.

Capacity is finite and small

The instrument prints a bounded amount of taker flow per hour. That flow is what fills resting orders, and it does not grow because more people bought the strategy.

So access is capped

So access closes once capacity is reached, rather than staying open indefinitely. Everyone who already bought keeps everything they paid for.

That is why access is limited: it keeps the strategy’s capacity for the people who hold it.

Buying

What you actually receive.

Pay in crypto

USDC on Solana, Base, Arbitrum or Ethereum. One-time, per strategy. No subscription, no performance fee, no managed account. Apexstoa never holds your funds or your keys.

Get the whole box

Runnable engine, pre-registration, the full backtest harness, venue integration, and the research record behind the listing: findings, pre-registrations and decision notes. The headline backtest regenerates end to end from data you fetch yourself. The supporting studies ship as write-ups, not as runnable code: the order-book tape, the analysis scripts and the multi-coin generator stay with the seller, so those conclusions are documented rather than re-derivable.

Reproduce it yourself

The harness regenerates the headline figures from raw data. If a published figure it checks doesn't reproduce on your machine, that's a refund.

Verify it on your exchange

Both products check your exchange before they trade. APEX ships a probe that tests the seven capabilities it relies on against your own funded account, using a test position worth only a few dollars, and it arms on a venue once those capabilities are measured. For Bitunix the package includes our own measurements: six of the seven pass, and partial closes are the one step the package asks you to reconcile. Run the probe on your account to confirm the same. CARRY sizes real money only on a verified venue. Every check runs on your own machine.

Terminal · from the unpacked download

$ cd apex-1.0.0/engine
$ pip install numpy pandas pyarrow requests pytest
$ python3 tools/fetch_data.py  # once, about 150 MB, digest-checked
$ python3 -m apex.cli verify  # every checked figure, from raw data

The download

Every folder, before you pay.

A map of the release archive, names only. The files themselves arrive on purchase. The tree is generated from the APEX 1.0.0 download itself, so it cannot drift from what you receive. Open any folder.

apex-1.0.0

engine

apex

data

multicoin_measured.json

venue

__init__.py

base.py

binance_usdm.py

bitunix.py

brackets.py

bybit.py

mock.py

simcore.py

__init__.py

__main__.py

_frozen.py

cli.py

config.py

signal_feed.py

trader.py

universe.py

verify.py

docs

APEX_manual.pdf

APEX_manual.sources.sha256

DISCLOSURE.md

package_map.png

PREREGISTRATIONS.sha256

PREREGISTRATIONS.sha256.ots

README.md

RISK.md

SETUP.md

TROUBLESHOOTING.md

WALKTHROUGH.md

probe_results

capability_probe_bitunix_20260828T122144Z.json

capability_probe_bitunix_20260828T160825Z.json

capability_probe_bitunix_20260828T161546Z.json

README.md

tests

fixtures

ADAUSDT_3m_warmup.npz

_bars.py

test_bar_data.py

test_binance_algo_budget.py

test_bitunix_offline_paths.py

test_bitunix_parity.py

test_bybit_offline_paths.py

test_cli.py

test_config_is_read_only.py

test_docs_parity.py

test_live_arm_gates.py

test_medium_findings.py

test_probe_binance_testnet.py

test_probe_capabilities.py

test_state_machine.py

test_trader.py

test_universe.py

test_venue_conformance.py

test_verify.py

test_window_gates.py

tools

apex_mc.py

apex_series.py

fetch_data.py

probe_capabilities.py

probe_multi_bracket_binance.py

regen_universe_table.py

render_apex_charts.py

render_apex_surface.py

apex.toml

proof

02-shortleg-candidate-050

sl_lib.py

execution_pilot

apex_exits.py

apex_risk.py

bitunix_v1.py

concurrency_table.json

rebaseline_bitunix_full.py

signals.py

venue.py

forward_test

twoada_paper_trader.py

README.md

research

02-shortleg-candidate-050

FINAL_PORTFOLIO.md

FINAL_SPEC_CORRECTED.md

FINDINGS.md

STEP0_DECLARATIONS.md

05-breakentry-nofvg-20260725

FINDINGS.md

06-survivorship-basket-20260726

FINDINGS.md

STEP0_DECLARATIONS.md

08-ada-l1-fills-20260726

HARNESS_NOTES.md

READ1_20260804.md

READ1_DRYRUN_20260731.md

READ2_20260810.md

READ2_EARLY_20260806.md

READ_CANCEL_ML_20260807.md

README.md

SCOPE_DECISION_20260815.md

STEP0_B_placement_gate_20260729.md

STEP0_DECLARATIONS.md

STEP0B_FIRSTDELIVERABLE_20260731.md

execution_pilot

PILOT_CLOSING_REPORT_20260731.md

PILOT_STEP0_DECLARATIONS.md

forward_test

CHECKINS.md

FORWARD_TEST_TWOADA.md

CLOCKS.md

GOLDEN_CONFIGS.md

README.md

LICENCE.txt

QUICKSTART.md

README.md

SHA256SUMS

Get started

Read the disclosures. Then decide.

Everything that bears on the decision is published here and in the pre-sale agreement: the returns, the drawdowns and the terms. Read it once, then choose the setting that suits you.

Every performance figure for these products comes from a backtest, a simulation over historical data, and trading leveraged cryptocurrency derivatives can lose you more than your deposit. APEX has a maximum historical drawdown of −71.6% at its highest setting and −42.0% at the conservative setting that ships (−71.9% and −44.6% with no limit on open positions). CARRY’s figures come from two years of settlements, and the funding regime that produces its yield is decaying.

Nothing here is investment advice. You run the software on your own exchange account and keep custody of your funds. The complete risk statement is in the pre-sale agreement, which you accept before any payment.

Sales are final, with two exceptions set out in the pre-sale agreement: a published figure that the product's verify command checks and that doesn't reproduce is refunded in full, and a defect no update can fix may be refunded at the seller's discretion. Otherwise the source code and the research record are delivered on purchase and can't be returned, which is exactly why every number that bears on the decision is above, before you pay, rather than in terms you'd have to go looking for.

Buy APEX · $399 Buy CARRY · $199 Buy both · $499 Why access is limited

One-time, in crypto. Both are on sale now, and the bundle is $99 below buying them apart. Nothing here is a subscription and there is no performance fee on either.

Runs on Bitunix, whose fee schedule every published figure is priced with (prices and funding come from Binance’s public data). Adapters for Binance and Bybit ship too, and the bot checks your exchange can do what the strategy needs before it arms, so it only ever trades where it has been verified.