What is quantitative analysis
Quantitative analysis is the use of statistics, mathematics and computing to study market data and test trading or investment ideas with measurable rules.
People who do it are called "quants". Quant trading means trading where entries, exits and position size come from rules that were defined and tested on data in advance. It is a research method. It does not guarantee results, and leveraged forex and CFD trading based on any method can lose money fast.
Formal definition
In finance, quantitative analysis is the building and testing of models that describe price, return, volatility or risk as variables. Each model has stated inputs, a stated rule and a measurable output. The output is tested against historical data. The model is accepted only if it holds on data it was not built on.
Quantitative analysis differs from technical analysis and fundamental analysis in how it works, not in the data it uses. A chart pattern or an inflation release can feed a quant model. The model only counts as quantitative if its rule is exact and its result is measured statistically.
Ready to try it on a real or demo account?
Partner link · trading carries a high risk of loss
Worked example (hypothetical)
A researcher asks this question: "After a day with an unusually wide range on EUR/USD, is the next day's range also wider than average?"
- Data. Download a long history of daily high, low and close prices.
- Variable. Compute each day's range and compare it with the average daily range over a fixed lookback window.
- Split. Keep the older part of the history for building the rule. Hold back the newer part for testing.
- Measure. On the older part, calculate the correlation coefficient between today's relative range and tomorrow's relative range.
- Test. Repeat the calculation on the held-back part. If the relationship weakens sharply or disappears, it was probably noise or overfitting.
- Apply to risk. If the relationship holds, a trader might use it to adjust stop distance or position size. Volatility clustering of this kind is a common input for risk management.
The result describes the past. It does not tell anyone what to buy or sell.
Where you see the term
- Platforms. MT4 and MT5 let users code Expert Advisors and indicators and run backtests in the Strategy Tester. Backtest reports show net result, drawdown and the equity curve.
- Reports. Fund factsheets list alpha, beta, annualized volatility and annualized return. All of these are quantitative measures.
- News. Coverage of "quant funds", "systematic strategies" or "algorithmic selling" during sharp market moves.
Edge cases and limits
- Overfitting. A model tuned too closely to past data fits the noise in that data and then fails on new data.
- Look-ahead bias. The test uses information that was not available at the time of the trade.
- Survivorship bias. Delisted stocks or closed instruments are missing from the dataset, so results look better than they were.
- Costs. Spreads, swaps and slippage can turn a small statistical edge into a loss.
- Regime change. A relationship can break after a policy shift or a black swan event. Fat tails make large moves more frequent than simple models assume.
Related terms
Frequent questions
What is the meaning of quant?
A quant is a person who applies mathematics and statistics to financial data. "Quant" is also used as a short form of quantitative analysis.
Is quant trading the same as algorithmic trading?
Not exactly. Quant trading is about how the rules are researched. Algorithmic trading is about executing orders automatically. They often overlap, but a quant rule can be executed by hand.
Do I need programming to do quantitative analysis?
Simple tests can be done in a spreadsheet. Larger datasets and backtests usually need code, such as MQL in MetaTrader or a general-purpose language.
Does a good backtest mean the strategy will work?
No. A backtest shows how rules behaved on past data. Live results can be worse because of costs, changing market conditions and biases in the test.