What Is a Quant Trading Strategy? A Plain-English Guide

"Quant" sounds like something only math PhDs at big funds do. It isn't.

A quant trading strategy is just a set of trading rules written down so clearly that a computer could follow them. No gut feel. No "it looked like it was about to pump." If the rule says buy, you buy. If it says sell, you sell.

That's the whole idea. The math can get fancy later. The core is simple: clear rules you can test.

Quant vs. discretionary trading

Most people start as discretionary traders. They look at a chart, read the news, and decide. That can work, but it has two problems.

First, it's hard to repeat. If you can't explain exactly why you took a trade, you can't do it the same way next time.

Second, it's hard to test. You can't run your mood through five years of price data.

A quant strategy fixes both. Because every rule is written down, you can run it on past data, then on new data, and see what happens. You can also see when it stops working, instead of guessing.

The parts of a quant strategy

Almost every quant strategy, simple or complex, has the same five parts.

1. Market and timeframe. What you trade and how often you look. For example: BTC, on 4-hour candles.
2. Entry rule. The exact condition that opens a trade.
3. Exit rule. When you close: a target, a stop, a time limit, or a signal flip.
4. Position size. How much goes into each trade. (We wrote a whole post on position sizing for beginners.)
5. Costs. Fees, slippage, and funding. Leave these out and almost anything looks good.

If any of the five is fuzzy, it's not a quant strategy yet. It's an idea.

A simple example

Here's a made-up example, just to show the shape. It is not a recommendation.

• Market: BTC, 4-hour candles.
• Entry: When the RSI (a common momentum indicator) drops below 30 and then closes back above 30, buy at the open of the next candle.
• Exit: Sell when RSI goes above 60, or if price falls 3% below the entry, or after 10 candles, whichever comes first.
• Size: Risk 1% of the account per trade.
• Costs: Count the exchange fee on both the entry and the exit.

Notice how boring that is. Every word can be turned into code. Two people following it would make the exact same trades. That's what makes it testable.

Also notice "buy at the open of the next candle." You only know a candle closed above 30 after it closes. Buying at that same close would mean using information you didn't have yet. That small detail is the difference between an honest test and a fake one. We go deeper on this in point-in-time truth.

Common types of quant strategies

Most strategies fall into a few families:

• Trend following. Buy things going up, sell things going down. Moving average crossovers are the classic example.
• Mean reversion. Bet that a stretched price snaps back. The RSI example above is one.
• Breakout. Trade when price leaves a range it has been stuck in.
• Carry and funding. In crypto, perpetual futures pay or charge funding. Some strategies are built around that.
• Relative value. Trade two related coins against each other when their usual gap gets too wide.

None of these is "the best." Each one works in some market conditions and fails in others. A big part of quant research is learning when a strategy should be on and when it should be off.

How you test a quant strategy

Writing the rules is the easy part. Testing them honestly is where most people go wrong.

Step 1: Backtest. Run the rules on past data and record every trade. Count fees. Look at how many trades it made, how many won, and the worst drop. Our guide on how to backtest without fooling yourself covers the traps.

Step 2: Forward test on paper. Run the same rules on live market data going forward, with no real money. This is the real test. A backtest shows what would have worked. Forward testing shows whether the idea survives data it has never seen. We explain the difference in backtest vs forward observation.

Step 3: Compare. Does the paper version behave like the backtest? About the same number of trades? A similar win rate? If the backtest said 40 trades a month and paper shows 6, something is off.

Only after that does it make sense to even think about real money. And when you report results, keep it honest. Something like "4 of 10 strategies net positive (simulated)" tells people far more than one big percentage.

Mistakes that trip up beginners

Tweaking until it looks good. Change the settings enough times and something will fit the past by luck. That's overfitting. Here are the warning signs.

Testing hundreds of ideas and keeping the best one. The more you try, the more likely your "winner" is just noise. Our post on the deflated Sharpe ratio shows how fast this happens.

Ignoring fees. A strategy that trades a lot can lose most of its gains to costs.

Too few trades. Ten trades tell you almost nothing. You need enough trades for the numbers to mean something.

Trusting records you can't check. If you can't prove an entry and exit, don't count it. That's why we refuse to count trades we can't verify.

Where AI fits in

AI tools like ChatGPT, Claude, Grok, or Cursor are good at one part of this: coming up with ideas and turning them into clear rules. Ask one for a mean reversion strategy and you'll get something in seconds.

But an AI can't tell you if the idea holds up. Only testing can. And because AI produces ideas so fast, the testing has to be strict, or you just fool yourself faster.

That's the gap we're working on. We're not building an AI that tells you what to buy. We're building the infrastructure that forces AI trading ideas to prove themselves.

How Pickled Lizard tests a strategy

At Pickled Lizard, a strategy starts with a backtest on the full data set we have. Then it moves to forward Paper trading, on market data it has never seen, with fees counted. It's marked Verified only when it reaches 80% of its own backtest's trade rate and 80% of its backtest win rate. Until then, it shows as still collecting.

Live trading is off during the beta. Everything is paper, and every result is labeled simulated. The more we trade, the smarter it gets.

Try it

The beta is free. Bring an idea, or one your AI came up with, write it as clear rules, and watch how it does on Paper.

Test it free at pickledlizardz.com/join

Paper trading only. Results are simulated. Not financial advice.