thesolidoke

My backtest said +0.15% to +0.31% per trade. Real trading gave me a tenth of that.

The strategy: fading the premium on tokenised stocks on Binance futures. On paper it looked great.

Then I ran it forward on a paper account using real bid/ask prices:
• 520 trades
• +257 USD total
• about +0.025% per trade, roughly 1/10 of what the backtest promised

Where did the other 90% go? The spread. Backtests usually assume you get the "mid" price. Real orders pay the gap between buyers and sellers, every time.

If a strategy's edge is only a few hundredths of a percent, the spread will eat it alive.

Does your backtest use bid/ask, or the mid price? 👇

Not financial advice. Paper trading results, shared openly.

#TradingBots #algoTrading #tradingfees #bitcoin $BTC

"Just risk more and you'll make more." I tested it.

Same crypto trend bot, 2020 to 2026, real fees, starting with 154 USDT:

• 1.5% risk per trade → 1,105 USDT, worst drop −25%
• 3% → kill switch triggered in 2021
• 5–10% → wiped out within 3–7 months
• 25% with no safety switches → −99.9%. Account gone.

Same strategy, same signals. The only difference was the bet size.

Most people don't lose because their strategy is bad. They lose because they bet too big for the strategy they have.

What risk per trade do you use? 👇

Not financial advice.

#TradingBots #algoTrading #tradingfees #bitcoin $BTC

I tested 42 trading rules on 14 metals markets. Not one made money.

Copper, aluminium, steel, iron ore, platinum, palladium, plus the miners' ETFs. Trend rules, breakout rules, long-only rules. All with real costs.

Result: no rule was positive after costs. Copper lost even before costs.

The plot twist: a few "winners" looked amazing at first. Then I checked the data. Those markets had stale prices, the same open/high/low/close on 62-93% of days. The bot was "trading" a price that wasn't moving.

And the kicker: simply buying and holding the miners beat every single rule (XME +198%, COPX +149% over 2021-26).

Lessons:
1. Check your data before you trust your results.
2. Sometimes the boring option wins.

Have you ever had a backtest that was too good to be true?

Not financial advice.

#TradingBots #algoTrading #tradingfees #bitcoin $BTC

Gemini said my new strategy would make +1.15% to +1.35% a day. On real data it made +0.11%: statistically nothing.

I send every backtest to two AIs (Gemini and GPT) for a second opinion. They're useful: they catch mistakes, suggest tests and argue with each other.

But:
• That "+1.35% a day" strategy was Gemini's own design.
• It once called my 2026 test dates "impossible" because its knowledge stopped earlier.
• Both sometimes state numbers with total confidence that aren't in my data.

My rule now: AI suggests, the data decides.

Sold an "AI trading bot"? Ask: who checked the AI?

Do you let AI touch your trading, or not?

Not financial advice.

#AI #TradingBots #algoTrading #BitcoinFiveWeeklyLosses $BTC

Would I let an AI see my bank account? It already sees my trading bots.

For a month an AI assistant has run my bot research: backtests, GPU runs, post-mortems. Here's where I drew the lines:
✅ It reads results, logs and backtests
✅ It drafts, I approve
❌ It can't move money
❌ Nothing goes live without my yes

Read-only access is the easy part. The hard part: AI is sometimes confidently wrong. Every claim gets checked against the raw numbers.

So would I try Claude's Personal Finance beta with read-only bank access? Same rules: read, never move.

The valuation? I test bots, not IPOs. Ask me in a year.

#AnthropicIPO #AnthropicRejectsIPODelayValuationinQuestion #AI #TradingBots $BTC

My first trading bot made money. Then I paid the fees.

Before my current bots, I tested a fast scalper on crypto.

363 trades.
Profit before fees: +$3.56 ✅
Fees paid: $22.13 ❌

The strategy was "right" slightly more than it was wrong, and still lost money. Every trade paid the exchange, and the exchange always wins that game.

What else the post-mortem showed:
• Stop-losses hit 140 times, targets only 89
• The only profitable exits were the time-based ones
• The shorts caused most of the loss

Lesson: an edge smaller than your fees isn't an edge. Now every test I run includes fees, then double fees, to see if it survives.

The scalper stays off. My current bots trade far less often, on purpose.

Not financial advice. Real numbers from my own testing.

#TradingBots #AlgoTrading #tradingfees #bitcoin #BitcoinFiveWeeklyLosses
$BTC

A month ago I made a decision: stop guessing, start testing.

I'm a project manager by trade (construction and logistics). My job is plans, numbers and deadlines, not charts. But I kept seeing the same thing online: screenshots of huge wins and never the losses.

So I set myself one rule: test everything properly, and show all of it.

How I started:
• 11 Oct: a bot trading indices and gold, live with about $1,000 on a cent account
• 14 Oct: a crypto trend bot on Binance futures, live with about $157 and tiny risk per trade (under 1%)
• Both run 24/7 on a rented server that checks itself and restarts anything that crashes

My ground rules:
Lock the rules before the test. No tweaking afterwards to make results look better.

Two AIs review every result (Gemini and GPT), and I check their claims against the actual numbers.

No verdict before 6 months. One good week means nothing. Neither does one bad week.
Small money first. If it can't work with $150, it doesn't deserve $15,000.

Why tell you? Because "honest bot results" is the content I wanted to find and couldn't.

Next post: the first strategy I tested, and why it failed.

Not financial advice. Just my own testing, shared openly.

#TradingBots #AlgoTrading #tradingjourney #bitcoin #BitcoinFiveWeeklyLosses $BTC

I spent a month building trading bots with AI.

Here's the honest scorecard.
No signals and no "100x" here. I'm testing trading bots properly and posting the results, losers included.

What I've got running:
• A crypto trend bot on Binance futures, real money: −5.6% so far
• An index and gold bot, real money: +2.75%
• 26 setups on demo accounts, plus 22 variations of my own bot on paper
• Dozens of strategy ideas backtested with strict rules. Most failed.
The big experiment: AI price forecasting
I took Kronos, an open-source AI model trained on millions of price charts, and fine-tuned it on 150 coins using rented GPUs (about $90).

Then I tested it on 78 days it had never seen.
✅ It ranked coins better than chance. Its picks beat the average, consistently, for up to a week.

❌ It still couldn't make money. The edge was smaller than trading fees, and the biggest moves (the ones that actually pay) were the ones it missed.

What I learned:
A model can be "right" and still lose you money. Fees eat small edges alive.

If a backtest looks amazing, assume you've made a mistake until proven otherwise.

Most "AI trading bot" claims you see online wouldn't survive a month of honest testing.

Coming next: a bigger test across 420 days of different markets, and I'll post the result either way.

Follow if you want the real numbers, not the screenshots.

Not financial advice. I'm sharing my own testing, wins and losses.
#TradingBots #AITrading #CryptoTrading #bitcoin #BitcoinFiveWeeklyLosses $BTC