The Carry Router: How a Bot Earns Money When Bitcoin Does Nothing

Right now, as you read this, our newest bot is paused.

It's holding zero Bitcoin. Zero Ethereum. Zero anything. Just dollars, sitting still.

That's not a bug. That's the bot doing its job.

In April 2026, we deployed our 11th paper-tracked bot: The Carry Router. It's a strategy that earns money from a quirk of crypto markets that most retail traders don't even know exists. And when the quirk disappears (like right now), the bot does the smart thing and waits.

This is the opposite of every other bot we've built. No price prediction. No momentum signal. No regime filter. Bitcoin can fly to $200,000 or crash to $20,000. The bot doesn't care.

Sound boring? Good. Boring is the entire point.

Why am I telling you this? Because the 30-year-old hedge fund "carry trade" is one of the cleanest, most boring, most academic alpha sources in finance. And almost no retail crypto trader knows you can run it on Bybit with $5,000.

In this article, I'll show you:

• The 8-hour payment that funds this whole strategy
• Why the bot is paused right now (and why that's correct)
• What roughly 5% per year actually buys you in a portfolio
• The single risk that could blow it all up

Let's go.

The 8-Hour Heartbeat

Here's something about Bitcoin perpetual futures most people don't notice. Every 8 hours, one trader pays another trader a small fee.

Bybit calls it the "funding rate." Binance calls it the same. It exists on every major exchange that lists perpetual futures.

Here's how it works. A perpetual future is a contract that lets you take a leveraged bet on Bitcoin without ever expiring. Regular futures expire on a date. Perpetuals don't. To keep the perpetual price tied to the actual spot price of Bitcoin, the exchange forces traders on the more crowded side to pay a small fee to the less crowded side.

When Bitcoin is in a hot bull market and everyone is leveraged long, the long traders pay the short traders. Every 8 hours. Forever, as long as the imbalance lasts.

In a typical year on BTC, the average funding rate has been around +0.01% per 8 hours. Doesn't sound like much. But:

That's the prize. A free 11% per year, paid out in 1095 small installments, just for being on the unpopular side of a one-sided market.

Still with me? Good. Because there's a catch.

You Can't Just Be Short

If you took a naked short position to collect the funding, you'd be exposed to Bitcoin's price going up. And historically, Bitcoin has gone up. A lot.

So here's the trick the carry trade uses, and it's almost embarrassing in its simplicity:

1. Buy $5,000 of Bitcoin spot (real Bitcoin, in a wallet)
2. Sell $5,000 of Bitcoin perpetual futures (a leveraged short position)
3. Hold both. Forever.

You're long Bitcoin in one place. Short Bitcoin in the other. Net price exposure: roughly zero.

Bitcoin doubles to $146,000? Your spot makes $5,000. Your perp loses $5,000. Net: $0.

Bitcoin crashes to $35,000? Your spot loses $2,500. Your perp makes $2,500. Net: $0.

But every 8 hours, the funding payment flows in. Quietly. Reliably. Independent of where Bitcoin's price went.

That's the carry trade. It's not crypto-specific. Foreign-exchange traders have done it since the 1990s. Commodity traders since forever. It's one of the most academically studied edges in all of finance.

The Carry Router is just an automated bot that does this on Bitcoin and Ethereum, sized small enough to be safe.

Why It's Paused Right Now

Here's the thing about funding rates: they're not always positive.

When Bitcoin enters a bear market or a long sideways period, traders get nervous. Some start betting against Bitcoin. The crowd shifts. The shorts become the majority. And now the math flips: shorts pay longs.

If our Carry Router is short while shorts are paying longs, the bot loses money on every funding cycle. That's bad.

So the bot has a rule. If the 7-day average funding rate drops below zero, exit. Wait. Don't enter again until it climbs back above 2% per year.

As of today, the 7-day average funding on Bitcoin is -1.22% per year. On Ethereum: -2.04% per year. Both below zero. Both paused.

I tested it. The bot's first cycle ran on April 28, 2026. It looked at the data and said: "No, thank you." It did nothing.

This is what makes me trust the bot. A strategy that won't force a trade when conditions are bad is doing 80% of the work most retail traders fail at.

Let me say that again, because it matters. Most retail traders lose money not because their strategy is bad, but because they refuse to do nothing. They need to be in the market. They need to feel busy. The Carry Router has no such need. It is comfortable holding cash.

What 5% Per Year Actually Buys You

Let me show you some honest math.

Backtest 2022-01 to 2026-04 (about 4 years), Bitcoin + Ethereum carry combined, conservative version:

Five percent per year doesn't sound exciting. But look at that drawdown number. Less than one percent. Over 4 years.

For comparison: holding Bitcoin over the same period had drawdowns above -50%. The S&P 500 had drawdowns above -25%. Most "high-yield" DeFi protocols had at least one -90% event.

The Carry Router isn't trying to make you rich. It's trying to convert your idle dollars into a slightly-better-than-T-bills return, with almost no drawdown risk.

That's what makes it useful. Not as a standalone get-rich strategy. As the base layer of a portfolio. The boring sleeve that works while the exciting sleeves are running their high-stakes plays.

Think of it like this. You have a kitchen. The Carry Router is the refrigerator. It's not the chef. It's not the recipe. It's the boring appliance that quietly does its job in the background, so the rest of the kitchen can do the interesting stuff.

The One Risk That Could Blow It Up

I'd be lying if I told you the Carry Router is risk-free. No strategy is.

The risk is concentration on a single exchange.

To run the carry trade, the bot needs both spot Bitcoin AND a perpetual short position on the same Bitcoin. Most retail traders only have access to one exchange that lets you do both: Bybit.

If Bybit has an FTX-style failure (withdrawals frozen, deposits seized, regulator action, hack), the entire $10,000 paper allocation is at risk. Not a small amount. The whole thing.

The professional version of this strategy would split capital across 3+ exchanges. We don't do that yet because we're paper-tracking a small allocation. But it's the real-money condition we'd need before deploying any actual capital.

This is the entire reason the bot stays paper-tracked under our all-paper policy for at least the next 6 months.

What's Next

The Carry Router runs daily at 11:35 AM Zurich time. It looks at the funding data. It decides: enter, exit, rebalance, or wait. Every day, the result lands on its bot card.

The first real-money decision is at least 6 months away. Earliest October 28, 2026. Assuming the bot delivers the kind of clean, low-drawdown numbers in live data that the backtest predicted.

If you want to follow along, the bot card on /bots updates daily. The Treasurer, our meta-bot, considers the Carry Router for allocation every day at noon. When funding turns positive again, you'll see the bot wake up and go back to work.

Until then: it sits in cash. Patient. Boring. Exactly as it should.

For quants: the raw numbers

Sources

• Bybit Funding History API
• Bot Card on /bots
• The Treasurer — meta-bot allocator
• Methodology — v2.1 multi-benchmark framework

This is not financial advice. The Carry Router is a paper-tracked strategy under our all-paper policy — no real money is allocated to it. All numbers shown are from backtests or paper-tracking, not from real-money deployment.

— Dominic, the guy who paper-tracks 11 bots so you don't have to.