Risk. Trading cryptocurrency carries a risk of losing money. Past results do not guarantee future results. This site is not investment advice and does not promise any income.
All strategies trade spot only and follow the common rules: each purchase is a separate lot, sell orders only above the purchase price including fees, every order passes the risk controller. They differ in where and how they place orders.
Test on historical data, not a guarantee. BTC/USDT and ETH/USDT, Binance 5-minute candles, 01.01.2018 to 31.10.2026, a $10,000 budget, a 0.1% fee per side, profit without compounding. "Realized profit" counts closed lots only; coins on hand are not included. The test counts an order as filled when the price touched the level; on a real exchange other orders queue ahead of yours, so real results may be lower. Source of the figures: the project's internal reports of 05.11.2026.
Some strategies trade small swings, others wait for rare deep falls or follow the trend. The market behaves differently over time and no single scheme fits always.
Idle money
In tests, individual strategies leave on average 35% to 67% of the money free, and lots bought at peaks wait 2–4 years. Different strategies are busy at different times.
Checked on paper
The strategies run side by side on separate paper accounts, on Binance and on Bybit, under equal conditions. The choice for real money will be made from live data and by the owner.
We do not promise that a set of strategies beats any single one. That is what is being tested.
1Dual-speed grid
Two grids with different rhythm and a reserve for a deep fall.
fast grid: 10 orders, small step
slow grid: 5 orders, large step
deep reserve: −15…−70% from the high
Not to scale: the dashed line is the current price, buy levels are below it.
When the price drops sharply (a close 1% below the 15-minute high), two grids start. The fast one: 10 orders with a 0.6–1.2% step; a lot is sold at +1.2% net. The slow one: 5 orders with a 2–4% step, target +3%. A separate deep reserve buys only on a big fall from the 30-day high (five levels from −15% to −70%), target +5%.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
14,7%
11,8%
Months with profit
62%
58%
Worst drawdown
−36%
−45%
Average free USDT
39%
40%
Test on historical data, not a guarantee.
Strengths
Drawdown in the test is lower than for most grids.
Targets are "net", after both fees.
Weaknesses
On average about 40% of the money is free and idle.
While one lot is stuck, the other fast-grid levels stay at old prices; the tested variant that shifts levels did not help.
2Grid with a coin stock
Part of the money is held in coins in advance.
buy grid: up to 8 orders, 1% step
Not to scale: the dashed line is the current price, buy levels are below it.
About 40% of the budget is bought in coins at once (8 lots). Each lot is sold at +1.2% net, and with the proceeds a buy-back order is placed 1% below the sell price. Another 40% is an ordinary buy grid with a 1% step; 20% is an untouched reserve.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
9,8%
9,4%
Months with profit
23%
14%
Worst drawdown
−52%
−49%
Average free USDT
49%
48%
Test on historical data, not a guarantee.
Strengths
Coins are already bought, so selling starts right away.
Weaknesses
Fewest profitable months among the grids (23% and 14%).
The 20% reserve is not used. A variant with a 20% stock instead of 40% gave 13.0% and 12.1% a year in the test; it is being checked separately.
3Dynamic grid
A grid that is rebuilt to follow the price.
12 buy steps below the center
Not to scale: the dashed line is the current price, buy levels are below it.
12 buy steps below a center. Every 6 hours, if the price has moved 3 or more steps away from the center, unfilled orders are re-placed around the new price (bought lots are not touched). Target +1.5% from the buy price. Half of the profit is added to the size of the next purchases (50% reinvestment).
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
20,1%
27,2%
Months with profit
31%
25%
Worst drawdown
−53%
−58%
Average free USDT
46%
45%
Table figures include 50% reinvestment; without it 13.1% / 16.0% per year. Test on historical data, not a guarantee.
Strengths
On ETH: the highest profit per year of all nine strategies (27.2%).
Weaknesses
The figures include reinvestment. Without it: 13.1% (BTC) and 16.0% (ETH); only those are comparable with the others.
Drawdown up to −58%; 30% of the budget sits in an idle reserve.
4Multi-level grid
A working grid plus deep levels.
working part: up to 25 small lots
deep part: 8 levels −15…−75%
Not to scale: the dashed line is the current price, buy levels are below it.
Working part (80% of budget): up to 25 small lots, step of at least 0.8%; it buys after a 1% drop within 15 minutes, target +2%. Deep part (20%): 8 levels from −15% to −75% from the 30-day high; it sells halfway back to the high, but never below +5%.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
17,2%
19,4%
Months with profit
89%
75%
Worst drawdown
−52%
−69%
Average free USDT
49%
28%
Test on historical data, not a guarantee.
Strengths
The most profitable months: 89% (BTC) and 75% (ETH).
Weaknesses
On ETH the drawdown is −69%, one of the deepest.
There were periods when almost no free money was left.
5Sharp-drop buyer
Buys after a sudden fall in price.
entry and 3 add-ons: −1.5%, −3%, −4.5%
Not to scale: the dashed line is the current price, buy levels are below it.
Waits for a fast fall: −1.5% in 15 minutes or −2.5% in an hour. On the signal it buys 5% of the budget with a limit order and, if the fall continues, adds at 1.5%, 3% and 4.5% below the first price. Target +1.5%. The next signal for a pair comes no earlier than 6 hours later.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
5,5%
6,2%
Months with profit
44%
30%
Worst drawdown
−22%
−25%
Average free USDT
77%
74%
Test on historical data, not a guarantee.
Strengths
The calmest in drawdown (−22% and −25%).
Few trades.
Weaknesses
Three quarters of the money sits idle (74–77% free).
Lots wait very long: up to three years (up to 1,105 days in the test).
6ETH/BTC deviation
The ETH/BTC pair: buys when ETH is unusually cheap against BTC.
entry and add-ons −2%, −4%, −6%, −8%
Not to scale: the dashed line is the current price, buy levels are below it.
The budget is counted in BTC. The strategy checks how far ETH/BTC has moved from its weekly average (a Z-score). At Z ≤ −1.5 it buys ETH for BTC and adds at 2, 4, 6 and 8% below the first price. Target +1% net in BTC.
Test on historical data (2018–2026)
ETH/BTC, profit in BTC
Realized profit per year
0,06–0,73% в BTC
Months with profit
2–21%
Worst drawdown
−7…−25%
Average free USDT
70–91%
Test on historical data, not a guarantee.
Strengths
With a wide window the drawdown is small (down to −7%).
Weaknesses
The income is negligible: 0.06–0.73% a year in BTC, depending on the window.
ETH/BTC fell from 0.054 to 0.029 and cycles got stuck. It stays a control strategy.
7Dense grid
Many orders with a very small step.
frequent steps with a small gap
Not to scale: the dashed line is the current price, buy levels are below it.
A buy grid with frequent steps. The next purchase is placed one step below the cheapest open lot; lots are small and targets are close. There are many trades, and profit builds up from frequent small sales.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
22,5%
24,3%
Months with profit
53%
31%
Worst drawdown
−61%
−65%
Average free USDT
27%
11%
Test on historical data, not a guarantee.
Strengths
High profit in the test: 22.5% and 24.3% a year.
Weaknesses
Drawdown −61% and −65%; money is almost always in use (average free 27% and 11%).
It gains more than others from the test assumption "an order fills when the price touches the level". On a real exchange that gain will be smaller.
8Trailing grid
Levels are tied to the high of the last day.
working part: −2…−10% from the daily high
deep part: 8 levels −15…−85%
Not to scale: the dashed line is the current price, buy levels are below it.
Working part (70%): 5 buy levels at 2, 4, 6, 8 and 10% below the 24-hour high; unfilled orders are recalculated every 15 minutes, target +2%. Deep part (30%): 8 equal lots from −15% to −85% from the high, target +5%.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
18,4%
18,6%
Months with profit
71%
72%
Worst drawdown
−59%
−59%
Average free USDT
22%
14%
Test on historical data, not a guarantee.
Strengths
71% and 72% of months with profit; 18.4% and 18.6% a year.
Weaknesses
Drawdown −59%; at the end of the test the unrealized loss was about $2.7k (BTC) and $4.1k (ETH).
It stays a control strategy: we do not offer it without a limit on the share of coins.
9Trend, core and grid
Trades actively only in a rising market.
main grid: 5 buys, 3% step
reserve: 5 buys, 6% step
Not to scale: the dashed line is the current price, buy levels are below it.
Checks whether the price is above its 200-day average. If so, a buy grid works with up to 50% of capital: 5 buys with a 3% step (target +2%) and 5 reserve buys with a 6% step (target +3%). If below, it cancels unfilled orders and buys almost nothing.
Test on historical data (2018–2026)
BTC
ETH
Realized profit per year
≈1%
≈1%
Months with profit
13% (BTC)
Worst drawdown
−28%
−30%
Average free USDT
88%
87%
Figures follow the first-version rules. Test on historical data, not a guarantee.
Strengths
Small drawdown (−28% and −30%).
Weaknesses
By the first-version rules it hardly trades: about 1% a year, 88% of the money idle.
A variant anchored to the high gave 8.7% (BTC) and 10.8% (ETH) a year in the test; it is being checked on paper and is not enabled yet.
An honest comparison: just buy the coin
If at the start of 2018 you bought the coin with the whole $10,000 budget and left it, at the end of the test (October 2026) you would have $56,565 in BTC and $30,773 in ETH. That is more than any grid: for example, the final capital of No. 1 is $21,180 (BTC) and $16,293 (ETH), No. 4: $24,540 and $24,423.
Grids do not try to beat the market's growth. Their aim is to take a little profit every day from price swings, and they do it because a large part of the money sits in USDT. They miss strong growth. This is a calculation on the past, not a forecast.
What else is being tested
A deep-level ladder for No. 2 and No. 3 and level shifting for No. 1: hypotheses; their historical tests are described in internal reports; they are not enabled on real accounts.
A "fill only when the price passes through the level" mode on paper accounts, to estimate how much of the test profit comes from a simple touch.
An alternative base for the next purchase (the "near" variant) for No. 2, 4 and 7: its profit gain in the test is unstable (better than the ladder only in some periods), so we do not enable it first.
Trade NF