Backtesting Trading Strategies

What is Backtesting?

Backtesting is determining how a trading strategy would have performed in the past. Backtesting is an essential element of developing an effective trading system. It can be done manually or systemically, and it aims to establish whether a trading strategy is worth implementing in the live market. 

The underlying principle is that a strategy that worked successfully in the past can be trusted to deliver profitability in the future. Of course, this assumes that price patterns in the markets tend to repeat themselves. However, this may always not be the case because markets are always dynamic and ever-changing. 

Nonetheless, backtesting remains a sound risk management activity, and it helps traders avoid implementing subpar trading strategies in the live market without ascertaining their potential effectiveness.

Good strategies after backtesting give traders the confidence to apply them in the markets, while a flopping strategy can be tweaked or abandoned altogether. 

Open an Account at AvaTrade and implement your winning strategies today!

Portfolio Backtesting

Backtesting a portfolio is a method of establishing how a specified portfolio asset allocation would have returned in the past over a certain period. A portfolio can consist of one asset class or span several asset classes.

For instance, an exclusive stock portfolio can consist of stocks from several industries such as Financials, Technology, and Industrials. Alternatively, a broad portfolio may consist of several asset classes such as Stocks, Bonds, REITs, and Commodities.

Portfolio backtesting is done to identify the optimal composition that can help investors achieve their desired objectives. The objective can be managing risk or achieving greater value. 

Portfolio backtesting utilises broad data sets and can include fundamental and economic events that occurred in the past, such as earnings reports, divestitures, mergers & acquisitions, regulatory changes, and interest rates. 

When portfolio backtesting is done, investors need to interpret its yield information. Some of the variables to analyse include Compound Annual Growth Rate, Standard Deviation, Maximum Drawdown, and Sharpe Ratio.

Portfolio backtesting should qualify a portfolio allocation strategy or help an investor identify the best weightings for each asset required to achieve their desired results. 

How to Backtest a Trading Strategy?

A backtest utilises historical data to qualify a strategy. The underlying trading style will determine how a trader will go back. A long-term investor may have to consider many years back, whereas a short-term trader can use data of several weeks or a couple of months. Backtesting can be done manually or systematically, and it will generally follow the steps below:

Define the strategy

Outline all the parameters that constitute your strategy. This includes the asset class you are trading and the chart timeframe. Different asset classes have different characteristics, and they will determine the amount of historical information to be collected. For instance, bonds can be analysed for as much as 20 years, whereas short-term currency traders can utilise data for up to a couple of weeks. 

Look for trades

Look for trades that meet the conditions of your strategy. Analyse and record the entry, and exit signals that the strategy would have generated had all the trades been taken. All valid trades should be recorded to determine the gross return. Valid trades constitute both winning and losing trades. 

Determine the net return

The net return is determined by factoring in other trading-related costs such as transaction costs, commissions, or relevant subscriptions. Compare the net return to the initial capital over the backtested period to find the net percentage return. 

When done effectively, backtesting will help determine whether a trading strategy could be profitable or not. A profitable strategy could be deployed in the live market with confidence, whereas a losing strategy may have its parameters tweaked and backtested again, or it can be abandoned altogether. 

Backtesting tips

Here are some tips to ensure effective backtesting:

• Consider different market scenarios. If you only backtest during bull markets, your strategy may perform poorly during bear or sideways markets.
• Aim to keep volatility as low as possible. A highly volatile strategy could especially be devastating in leveraged markets and expose you to potential margin calls. 
• Backtest using a relevant set of data. For instance, a trading strategy applied in manufacturing stocks may perform poorly when trading technology stocks. 
• Customise backtesting parameters to meet your specific needs to get accurate results. The parameters can include position sizes, margin requirements, and transaction costs. 
• Be careful about over-optimisation. The intention is to get a profitable strategy (more wins than losses sustainably) and not a perfect one.  

You should also be aware that, though useful, backtesting may not be the best way to determine whether a strategy will be successful or not. This is because markets keep changing, and past results do not provide a cue for future performance. 

Backtesting vs Forward Testing

Backtesting involves determining how a strategy would have performed in the past. However, historical data alone is not enough to establish the viability of a trading system. This is why it is important also to perform forward testing. 

Also known as paper trading, forward testing simulates trading using live market data. It’s called “paper trading” because the trades are entered on ‘paper’ and do not use deposited capital, so no money is lost. Forward testing helps assess how a strategy would perform under live market conditions. 

Limitations of Backtesting

Backtesting is a vital step in developing a robust trading strategy, but several common pitfalls can compromise its reliability. Traders must be aware of these challenges and take proactive measures to mitigate them.

Data Quality Issues

Historical Data Inaccuracies

• Pitfall: Using datasets that contain gaps, errors, or inconsistencies can lead to misleading backtest results.
• Solution: Always source historical data from reputable providers and verify its completeness. When backtesting equities, ensure that corporate actions such as dividends and stock splits are accounted for to maintain accuracy.

Survivorship Bias

• Pitfall: Analysing only assets that have survived to the present day ignores those that failed, potentially inflating performance metrics.
• Solution: Use datasets that include delisted securities to create a more realistic historical representation of market conditions.

Overfitting and Data-Snooping Bias

Overfitting

• Pitfall: Excessively optimising a strategy to fit historical data too perfectly can result in poor real-world performance.
• Solution: Implement out-of-sample testing and cross-validation techniques to ensure the model isn’t merely capturing historical noise.

Data-Snooping Bias

• Pitfall: Repeatedly testing multiple hypotheses on the same dataset increases the likelihood of false positives.
• Solution: Predefine your trading hypothesis and strategy criteria before running backtests. Consider using Monte Carlo simulations to evaluate the robustness of your results.

Ignoring Real-World Trading Costs

Transaction Costs and Slippage

• Pitfall: Neglecting the impact of trading fees, slippage and liquidity constraints can create an overly optimistic performance projection.
• Solution: Incorporate realistic estimates of these costs in your backtest. Simulating different market conditions can help gauge their potential impact on trade execution.

Misinterpretation of Metrics

Reliance on a Single Performance Metric

• Pitfall: Overemphasising one metric, such as total return, without considering risk exposure can be misleading.
• Solution: Evaluate multiple performance indicators, including the Sharpe ratio, maximum drawdown, and win-loss ratio, to gain a comprehensive view of strategy effectiveness.

Interpreting Backtesting Results

Understanding how to analyse backtesting results is crucial for refining a trading strategy. Below are key performance metrics and their significance:

Sharpe Ratio

• Definition: Measures risk-adjusted return by comparing excess returns (above a risk-free rate) to return volatility.
• Usage: A higher Sharpe ratio suggests better risk-adjusted performance, but excessively high values may indicate overfitting rather than sustainable profitability.

Check out our full Sharpe Ratio guide.

Drawdown

• Definition: Represents the percentage decline from a portfolio’s peak value to its lowest subsequent point.
• Usage: Assessing maximum drawdown helps traders evaluate potential capital requirements and risk exposure in adverse market conditions. For instance, a strategy with modest returns but severe drawdowns may be unsuitable for risk-averse traders.

Check out the full AvaTrade guide: Drawdown in Trading.

Win-Loss Ratio

• Definition: The ratio of profitable trades to losing trades.
• Usage: While a high win rate is generally positive, traders should also assess the average gain per trade versus the average loss. A high win rate with small gains but occasional large losses could still pose significant risks.

Additional Metrics to Consider

1. Profit Factor: Ratio of gross profit to gross loss. A value above 1 suggests profitability, but extremely high values may indicate over-optimisation.
2. Expectancy: The average expected return per trade, factoring in both winning and losing trades.
3. Volatility & Standard Deviation: Measures consistency of returns. High volatility may signal potential instability in the strategy’s performance.

Advanced Insights

1. Out-of-Sample Testing: Validating a strategy on unseen data helps ensure that backtest results are not due to historical anomalies.
2. Forward Testing: Running the strategy in a live simulated environment (paper trading) highlights potential discrepancies between backtest results and real-world execution.

Case Studies of Backtest Strategies

Examining real-world applications of backtesting can highlight both successful and unsuccessful strategies.

Successful Application: Momentum Trading

1. Background: A trader designs a momentum-based strategy, buying assets that exhibit strong recent performance and selling underperforming assets.
2. Backtesting Outcome: The backtest reveals consistent returns, a favourable Sharpe ratio, and manageable drawdowns.
3. Implementation & Results: The strategy performs well in forward testing under similar market conditions.
4. Lessons Learned:
   1. Incorporating risk management rules (such as stop-loss orders) is essential.
   2. Strategy success is often linked to prevailing market conditions, so traders must remain adaptable.

Unsuccessful Example: Overfitted Mean Reversion Strategy

1. Background: A trader develops an algorithm that exploits mean reversion patterns of a specific stock index.
2. Backtesting Outcome: Exceptional historical returns, but closer examination reveals excessive fine-tuning to past data.
3. Implementation & Results: The strategy underperforms in forward testing, incurring significant losses.
4. Lessons Learned:
   1. Avoid overfitting by limiting the number of optimised parameters.
   2. Recognise that real-world trading conditions, such as liquidity constraints and unexpected market events, cannot always be fully captured in historical data.

Addressing Key Questions in Backtesting

1. How Can Traders Ensure Data Quality and Reliability?

Use reputable data sources adjusted for corporate actions e.g. dividends, splits and delistings.
Cross-check historical data for errors and missing values to maintain accuracy.

2. What are the Limitations of Backtesting, and How Can They be Mitigated?

Limitation: Historical performance does not guarantee future results.
Mitigation: Combine backtesting with forward testing and live simulations (paper trading) to validate strategies.

Limitation: Backtesting does not fully capture market dynamics such as liquidity crises or geopolitical events.
Mitigation: Regularly monitor and adjust trading strategies to adapt to evolving market conditions.

3. How Do Backtesting Considerations Differ Across Asset Classes?

Stocks:
✔ Must account for corporate actions (dividends, buybacks, earnings reports).
✔ Liquidity can be lower in market downturns, impacting execution quality.

Forex:
✔ Operates 24/5 with typically higher liquidity than stocks.
✔ Spread fluctuations and differing market sessions e.g. London vs. New York, can influence trading performance.

Tip: Tailor backtesting methodologies based on the unique characteristics of each asset class.

4. Why is Forward Testing Important After Backtesting?

Importance: Forward testing helps identify execution challenges, such as slippage and unexpected market volatility, that backtests may not reveal.
Tip: Use forward testing as a final validation step before committing real capital to a strategy.

5. How to Adjust Strategies Based on Backtesting Outcomes?

Step 1: Analyse Performance Metrics – Review Sharpe ratio, drawdown and win-loss ratio to identify areas for improvement.
Step 2: Refine Parameters Cautiously – Avoid excessive optimisation and make incremental adjustments instead.
Step 3: Implement Risk Management – Adjust stop-loss levels, position sizing and diversification based on risk metrics.
Step 4: Stay Adaptive – Market conditions change, so continuous monitoring and adjustments are crucial for maintaining an edge.

Final Thoughts

Backtesting is a powerful technique, but its effectiveness depends on data quality, realistic cost assumptions, and proper interpretation of performance metrics. By avoiding common pitfalls, validating strategies through forward testing, and continuously refining trading approaches, traders can improve their ability to navigate the markets successfully.

Trading is an ever-evolving discipline, so staying adaptable and data-driven is key to long-term success.

See a trading opportunity? Open an account now!

FAQ

• What is backtesting?

  Backtesting is determining whether a strategy is viable based on historical data. It aims to observe how the strategy would have performed in the markets in the past.

   
• Does Backtesting work?

  Backtesting works as a method of establishing the viability of a trading strategy. Nevertheless, effective backtesting will also depend on the underlying strategy and awareness of the limitations of the process.

   

** Disclaimer –While due research has been undertaken to compile the above content, it remains an informational and educational piece only. None of the content provided constitutes any form of investment advice.