Getting Started with Automated Trading

The promise of automated trading is compelling: set up a system that trades for you, generating consistent returns while you focus on other priorities. But before diving into automation, understanding the prerequisites for success can mean the difference between profitable trading and costly lessons.

The Discretionary Trading Foundation

One of the most important questions new algorithmic traders face is whether they should learn discretionary trading first. The answer, according to experienced automated traders, is a resounding yes. Discretionary trading experience provides an essential foundation that dramatically improves your chances of automation success.

Why Discretionary Experience Matters

When you trade manually, you experience the full emotional and psychological cycle of trading. You watch markets move against you, feel the temptation to exit profitable trades too early, and develop an intuitive sense of market behavior that's difficult to quantify but invaluable for strategy development.

This experience teaches you about market structure in ways that reading about it cannot. You observe where activity concentrates, why certain price levels hold significance, and how retail traders typically behave. These observations form the foundation of profitable strategies.

Perhaps most importantly, discretionary trading forces you to develop a thesis about why markets move. What drives price action? What causes sudden reversals? When you can answer these questions from experience, you're far better equipped to design automated strategies that work.

Understanding Market Cycles

Markets don't behave uniformly across all conditions. A strategy that performs brilliantly in trending markets may struggle during sideways consolidation. One that excels in high volatility periods may underperform in calm markets. Discretionary experience exposes you to these varying conditions.

Experiencing the Vicissitudes of Markets

The term "vicissitudes" captures the ever-changing nature of market behavior. Through discretionary trading, you experience:

Trending markets where momentum strategies shine and mean reversion strategies struggle

Ranging markets where mean reversion works while trend following produces whipsaws

High volatility periods when stop losses get hit more frequently

Low volatility periods when profit targets become harder to reach

Shifts in correlation between related instruments

Changes in average daily range that affect position sizing

Without experiencing these different market regimes firsthand, you might design a strategy that works beautifully in backtests but fails in live trading because it's optimized for conditions that no longer exist.

The Automation Trap

The promise that "you don't need to know how to trade, just turn on the system" is dangerous marketing. These claims suggest that profitable trading requires no skill or knowledge—just the right software. This fundamentally misunderstands what makes trading profitable.

What Automation Cannot Do

Automation excels at execution consistency, removing emotional decision-making, and managing multiple positions or markets simultaneously. But automation cannot:

Determine whether a strategy has a genuine edge or is curve-fit to historical data

Recognize when market conditions have changed and the strategy no longer works

Adapt to new market regimes that weren't present in backtesting data

Distinguish between a normal drawdown and a broken strategy

These judgments require human understanding that comes only from experiencing markets directly. When you've manually traded through several market cycles, you develop intuition about when something feels wrong—when the rhythm of fills, slippage, or price action has changed in ways that suggest your edge has diminished.

Strategy Categories and Characteristics

Before automating any strategy, you should understand its fundamental category and the market characteristics that support its success. The two primary strategy categories are trend following (momentum) and mean reversion.

Trend Following Strategies

Trend following strategies assume that price movements persist. When an asset starts moving in one direction, these strategies bet that movement will continue. Successful trend following requires:

Markets with sufficient momentum to overcome transaction costs

Instruments that develop sustained directional moves rather than constant choppy action

Timeframes that match the typical duration of trends in your chosen market

Stop losses wide enough to avoid getting shaken out of good trends

Position sizing that allows you to capture large winning trades

Trend following strategies typically have lower win rates (often 35-50%) but larger average winners than losers. They suffer during ranging markets and require psychological tolerance for extended drawdown periods.

Mean Reversion Strategies

Mean reversion strategies assume that extreme price movements will reverse, with prices returning to an average or equilibrium level. These strategies require:

Markets that exhibit range-bound behavior or regular oscillation around key levels

Sufficient liquidity to enter and exit positions near turning points

Statistical measures to identify when price has moved "too far" from the mean

Risk management that limits loss when price continues moving away from the mean

Quick exits to capture profits before the next reversal

Mean reversion strategies often have higher win rates (sometimes 60-70%) but smaller average winners and occasional large losers when a genuine breakout occurs. They struggle in strongly trending markets.

The Importance of Categorizing Your Strategy

Knowing your strategy's category helps you:

Select appropriate markets and timeframes for testing and trading

Identify the market conditions where your strategy should perform well

Recognize when market conditions no longer favor your approach

Choose complementary strategies that perform well in different environments

Set realistic performance expectations based on the strategy type

Without this understanding, you might abandon a perfectly good trend following strategy during a ranging market period, or keep trading a mean reversion strategy long after a trend has begun.

Proper Backtesting Methodology

Many traders approach backtesting as a one-time validation before going live. This approach misses the true purpose of backtesting: understanding how your strategy behaves across various market conditions and timeframes.

Testing Across Multiple Time Periods

A robust backtest doesn't simply run your strategy from point A to point B and check if it's profitable. Instead, it examines performance across:

Multiple separate time periods to ensure the edge isn't specific to one market regime

Various market conditions including trending, ranging, high volatility, and low volatility periods

Different economic environments (bull markets, bear markets, sideways markets)

Periods both in-sample (used for strategy development) and out-of-sample (used for validation)

When you backtest across multiple periods, you're looking for consistency in the character of results, not just overall profitability. A strategy that makes 20% one year and loses 18% the next might have the same average annual return as one that makes 8% every year, but they're very different strategies with different risk profiles.

Walk-Forward Testing

Walk-forward testing addresses one of backtesting's fundamental weaknesses: the temptation to optimize your strategy on the same data you're testing it against. This optimization bias can create strategies that look phenomenal in backtests but fail in live trading.

The walk-forward process works like this:

Optimize your strategy on an in-sample period (for example, 2020-2022 data)

Test those optimized parameters on an out-of-sample period (2023 data) without any changes

Record the out-of-sample results

Roll forward your in-sample period (now 2021-2023) and repeat the process

This methodology simulates how your strategy would have performed if you were periodically reoptimizing it and trading it forward. If your strategy maintains an edge across multiple walk-forward iterations, you have greater confidence it's capturing a real market inefficiency rather than random patterns in historical data.

Performance Metrics That Matter

When evaluating backtest results, resist the temptation to focus solely on total return. More sophisticated metrics provide better insight into whether a strategy will work in live trading.

Metrics for Trend Following Strategies

For trend strategies, focus on:

Average win size compared to average loss size (should be well above 1.5:1 for trend strategies)

Maximum drawdown and time to recovery (trend strategies often have deep, extended drawdowns)

Win rate during trending periods versus ranging periods

Percentage of returns from the top 10% of trades (trend strategies often depend heavily on a few large winners)

Metrics for Mean Reversion Strategies

For mean reversion strategies, examine:

Win rate (should typically be above 60% for mean reversion)

Average win compared to average loss (often closer to 1:1 or even smaller wins)

Maximum consecutive losses (mean reversion can experience clusters of losses during trends)

Performance during trending periods (should show minimal trading activity or flat performance)

Avoiding Optimization Bias

One of the biggest dangers in strategy development is creating a system so precisely tuned to historical data that it has no predictive power for future results. This over-optimization, often called curve-fitting, produces backtest results that look exceptional but live performance that's disappointing or disastrous.

Warning Signs of Over-Optimization

Several indicators suggest your strategy might be over-fit:

Performance degrades dramatically when you change any parameter slightly

The strategy has many parameters with very specific optimal values

Out-of-sample performance is substantially worse than in-sample performance

The strategy's logic doesn't make intuitive sense based on market behavior

Performance relies heavily on exact entry or exit timing (down to specific bars)

Principles of Robust Strategy Design

To avoid over-optimization:

Start with a simple strategy based on a sound market principle

Add complexity only when it improves performance across multiple time periods

Use parameters that remain relatively stable across different optimization periods

Verify that the strategy's logic makes intuitive sense

Test on multiple markets or instruments to verify the approach generalizes

Remember that markets are noisy environments where exact precision is impossible. Strategies that work robustly across a range of parameter values are more likely to continue working than those requiring precise calibration.

From Backtest to Live Trading

Even a well-backtested strategy may perform differently in live markets. This difference stems from factors that are difficult or impossible to model accurately in backtests.

The Reality Gap

The gap between backtested and live performance comes from:

Slippage between the theoretical entry price and actual fill price

Market impact when your orders affect the prices you receive

Changes in market volatility or behavior patterns

Psychological factors that affect your willingness to follow signals

Technology issues like internet connectivity or platform downtime

To bridge this gap, experienced traders use a staged rollout:

Begin with paper trading to verify that orders execute as expected

Move to live trading with minimal position sizes to experience the emotional reality

Gradually increase position size as the strategy proves itself in live conditions

Continuously monitor for changes in fill quality or strategy performance

When to Stop a Strategy

One crucial skill discretionary experience teaches is recognizing when a strategy has stopped working. In live trading, you should halt a strategy when:

Drawdown exceeds your tested worst-case scenario by a significant margin

Fill quality deteriorates substantially from what backtests assumed

Market conditions have fundamentally changed (volatility regime shift, regulatory changes, etc.)

The strategy's performance statistics drift significantly from backtested expectations

Making these judgments requires understanding both the strategy's theoretical foundation and practical market behavior—both of which come from discretionary trading experience.

Position Sizing and Risk Management

A profitable strategy can still blow up an account if position sizing is inappropriate. Discretionary experience teaches you how different position sizes feel psychologically and how market volatility affects your risk.

Kelly Criterion and Practical Position Sizing

The Kelly Criterion provides a mathematical framework for position sizing based on your edge and the odds of success. However, most experienced traders use a fraction of the full Kelly size (often 25-50%) because:

Kelly assumes your probability estimates are perfect, which they never are

Full Kelly sizing leads to significant volatility in account equity

Psychological comfort typically requires more conservative sizing than Kelly suggests

Through discretionary trading, you learn what position size allows you to follow your strategy without intervention. If you find yourself constantly worrying about open positions or manually intervening, your position size is too large for your psychological comfort.

Risk of Ruin

Understanding your risk of ruin—the probability that your strategy will reduce your account below a viable trading level—is essential. This calculation depends on:

Your edge (expected value per trade)

The variability of your returns

Your position size relative to account size

The number of trades you'll execute

Even with a positive expected value, poor position sizing can create a significant risk of ruin. Discretionary experience helps you understand these concepts not just mathematically but practically.

Building a Trading Thesis

Perhaps the most valuable outcome of discretionary trading experience is developing a clear thesis about why your strategy should make money. This thesis forms the foundation of your automated approach.

Components of a Trading Thesis

A robust trading thesis explains:

What market inefficiency or behavior pattern your strategy exploits

Why that inefficiency exists (market structure, behavioral bias, liquidity provision, etc.)

What conditions allow that inefficiency to persist

What would cause that inefficiency to disappear

How your strategy captures value from that inefficiency

For example, a simple trend following thesis might be: "Institutional position building creates momentum that persists over multiple days because large orders must be split to avoid market impact. By identifying the early stages of this momentum, we can ride a portion of the position building. This works because information advantages and slow capital deployment are persistent features of institutional trading."

With this thesis, you can evaluate whether your strategy's live performance aligns with expectations. If the thesis is based on institutional position building but your strategy is triggered by retail trader activity, you might need to reconsider your approach.

The Role of Automation

After establishing discretionary proficiency, understanding your strategy category, testing thoroughly, and developing a clear thesis, automation becomes a powerful tool for improvement.

What Automation Improves

Automation enhances trading by:

Removing emotional decision-making that leads to premature exits or revenge trading

Ensuring consistent execution across all valid signals without cherry-picking

Managing multiple positions or markets simultaneously

Executing precisely at planned entry and exit points without hesitation

Maintaining trading discipline during drawdown periods

TradersPost excels at providing this automation layer, taking signals from your analysis (whether from TradingView indicators, custom scripts, or other sources) and executing them consistently across your connected broker accounts.

What Automation Requires from You

Successful automation requires that you:

Monitor strategy performance against expectations

Recognize when market conditions have changed

Adjust position sizing based on recent volatility

Stop strategies that have broken or market conditions that no longer suit them

Continue learning and refining your approach

The goal isn't to create a strategy and forget about it. The goal is to automate the mechanical aspects of execution while you focus on the strategic aspects of strategy management and development.

Starting Your Automation Journey

If you're ready to begin automating your trading, the path forward is clear:

Step One: Build Discretionary Proficiency

Trade manually using your intended strategy. Keep a detailed journal of your trades, including your reasoning, emotional state, and outcome. Trade through at least one complete market cycle—experiencing both favorable and unfavorable conditions.

Focus on developing a deep understanding of why your strategy works when it works and why it fails when it fails. This understanding is impossible to gain from backtests alone.

Step Two: Formalize Your Strategy

Convert your discretionary approach into explicit rules that a computer can follow. This process often reveals ambiguities in your manual trading—situations where you make intuitive judgments that are difficult to codify.

Don't worry about capturing every nuance of your discretionary approach immediately. Start with the core rules that generate most of your edge. You can refine and add sophistication over time.

Step Three: Test Thoroughly

Backtest your formalized strategy across multiple time periods and market conditions. Pay special attention to performance during conditions different from when you developed the strategy.

Use walk-forward testing to verify that your strategy maintains an edge when parameters are optimized on past data and then applied to future data. This is the closest you can come to simulating how the strategy will perform going forward.

Step Four: Paper Trade

Before risking capital, paper trade your automated strategy. This isn't primarily about verifying profitability—your backtests should have already done that. Paper trading verifies that:

Your automation platform executes orders as you expect

Your position sizing logic works correctly

Stop losses and profit targets execute properly

Your alerts or signals trigger at the right times

Step Five: Start Small and Scale Gradually

Begin live trading with the minimum position size you can trade. This allows you to experience the emotional reality of following your system with real money while risking minimal capital.

As the strategy proves itself and you gain confidence in both the strategy and your automation setup, gradually increase position size toward your planned level. This scaling process might take months, but it's far safer than immediately trading at full size.

Common Mistakes to Avoid

Learning from others' mistakes is more efficient than making every mistake yourself. Common pitfalls in automated trading include:

Automating a Strategy You Haven't Traded Manually

Without discretionary experience with your strategy, you won't recognize when it's broken versus when it's experiencing a normal drawdown. You'll lack the confidence to continue trading through difficult periods and the wisdom to stop when conditions genuinely change.

Over-Optimizing in Backtests

The temptation to tweak parameters until backtest results look perfect is strong. Resist it. Strategies optimized to historical data rarely perform as well going forward. Robust strategies work across a range of parameter values and make intuitive sense.

Ignoring Transaction Costs

Backtests that ignore commissions, slippage, and market impact can show profitability for strategies that lose money in live trading. Be conservative in your estimates. If your backtest shows 15% annual return before costs and 14% after costs, the strategy is probably too sensitive to execution quality.

Trading Too Many Strategies Simultaneously

Each strategy requires monitoring and understanding. Starting with five different automated strategies simultaneously is a recipe for confusion. Begin with one or two strategies maximum. Only add more after you've proven you can manage what you have.

Failing to Adapt Position Sizing to Volatility

Many strategies experience periods of increased or decreased volatility. Fixed position sizing can lead to taking too much risk during high volatility periods or too little during calm markets. Implement volatility-adjusted position sizing to maintain consistent risk.

The Future Is Hybrid

The most successful approach to algorithmic trading isn't purely automated or purely discretionary—it's a hybrid that leverages the strengths of both approaches.

Use discretionary analysis to:

• Identify market regime changes
• Recognize when conditions favor or disfavor your strategies
• Develop new strategy ideas
• Provide context that helps you stay committed during drawdowns

Use automation to:

• Execute your strategies consistently without emotional interference
• Manage multiple positions or markets beyond human capacity
• Maintain discipline during stressful market conditions
• Scale your trading across multiple accounts or markets efficiently

TradersPost enables this hybrid approach by automating execution while leaving strategic decisions in your hands. You decide when to trade, what strategies to use, and how much size to commit. The platform handles the mechanical work of order routing, position management, and multi-account execution.

Moving Forward

The path to successful automated trading isn't a shortcut around learning to trade—it's an enhancement of trading skills you've already developed. By combining discretionary experience with systematic testing and disciplined execution automation, you create a trading approach that's greater than the sum of its parts.

Start with discretionary proficiency, develop a clear thesis about why your strategy should work, test thoroughly across multiple market conditions, and only then automate the execution. This foundation gives you the knowledge to manage your automated strategies successfully, recognizing when they're working as expected and when conditions have changed.

Automation amplifies your edge—but you must develop that edge first through understanding markets, experiencing their behavior firsthand, and learning what actually works across different conditions. Once you have that foundation, automation becomes a powerful tool for consistent, disciplined, scalable trading.