Craig Peters scans close to 10,000 stocks. He is fully algorithmic, trading global CFDs through Interactive Brokers across exchanges in the U.S., UK, France, Germany, and Asia. He typically has around 40 open trades at any time and logged 120 stock trades in a single month. His background is in physics and actuarial science, and he played professional football for Tottenham Hotspur in his teens before walking away from sport to pursue academia.
The result he shares in this episode is specific and measurable: after applying a single advance-decline indicator from Tomas’s Trading Market Internals course, Craig’s risk-adjusted returns improved by nearly 40% in out-of-sample testing.
One indicator. One idea. 40% improvement in risk-adjusted performance.
What makes Craig’s story especially useful is that he did not just apply the indicator as taught. He adapted it to fit his own situation – global stocks rather than U.S. futures – and then extended it in a direction that Tomas had not considered. That creative extension led to a second research avenue that is still generating ideas.
Craig’s Trading Background and Style
Craig found systematic trading through a work colleague who invited him to a technical analysis seminar in 2005. That was the turning point. He has been trading systematically, with increasing discipline, ever since.
Over time, he identified trend following as his natural style – not through reading about it theoretically, but through self-discovery. He realized he was psychologically comfortable taking many losses in a row because he understood that the payoff would come later. That tolerance for extended losing streaks is not common, and it is the defining psychological requirement for trend following. His winning trades on his main stock trend system typically last three to five weeks; his losers stop out within a day or two.
His current portfolio combines trend-following strategies with mean reversion systems across global stocks. He uses daily bars throughout, scans across all CFDs available on Interactive Brokers, and runs fully automated execution.
His primary performance target is not a return number. It is a drawdown ceiling. He targets a maximum drawdown of about 25%. The return goal is “whatever the market gives me” – he is not trying to hit a fixed percentage, he is trying to be positioned correctly when opportunities arise while keeping drawdown controlled. That is professional-grade thinking. Most beginning traders flip this completely, optimizing for return targets without a clear drawdown framework.
The Problem He Was Trying to Solve
Craig’s primary challenge was short-side breakout failures. In his trend-following system, he included short breakouts for diversification – having both long and short exposure reduces correlation and, theoretically, provides a buffer during down markets.
But in practice, short-side breakouts were costing him significantly during bull markets. The specific failure mode: a stock would break down in sympathy with a broad market sell-off. Everything was moving down together. The stock looked like a good short setup. He entered. Then the general market stabilized or reversed, and the stock – which had only broken down because the whole market dragged it lower – reversed and stopped him out.
The key insight is that this was not a stock-specific problem. The breakout was failing not because the stock was wrong, but because the market conditions that triggered the entry were temporary and externally driven. The short breakout was valid in isolation but invalid given the market context.
He had tried using market indices directly as a proxy for market condition – if the index is falling, maybe skip new short entries. It did not work well enough. The indices did not capture the right signal with enough precision.
The Advance-Decline Indicator Solution
The Trading Market Internals course gave Craig a different tool to think with: the advance-decline indicator. Instead of using a single index price as a market condition proxy, he could measure the breadth of market participation.
Craig’s custom implementation calculates this indicator independently for each exchange he trades:
- For each exchange (U.S., UK, France, Germany, Hong Kong, Singapore, Australia), count the number of stocks that advanced that day.
- Subtract the number of stocks that declined.
- Divide by the total number of stocks tracked on that exchange.
- This produces a normalized value between -1 and +1.
A reading of -1 means every stock on the exchange fell. A reading of +1 means every stock rose. A reading of zero means advances and declines were equal.
The insight Craig tested: when this value is sufficiently negative – when there is a large majority of stocks declining across the exchange – new short breakout entries have strongly negative expected value. These are the market-wide sell-off conditions where stocks break down in sympathy rather than on genuine individual weakness. Entering shorts in that environment almost guarantees you are on the wrong side of the eventual mean reversion.
Specifically, Craig found that filtering out any short breakout entry when the previous day’s indicator value was below approximately -0.5 eliminated around 30% of all short trade entries. These were the low-quality entries that were driving his losses. In out-of-sample back-testing, removing these trades improved risk-adjusted returns by nearly 40%.
The exact threshold – whether it is -0.5 or -0.4 or -0.6 – matters less than the concept. The precision of the signal comes from measuring actual breadth participation across the exchange, not from the arbitrary threshold value.
Why This Works – The Logic Behind the Filter
The mechanism is straightforward once you see it. A short breakout has genuine negative expectancy when it is being triggered by macro conditions rather than stock-specific weakness. In a market-wide sell-off, the advance-decline indicator goes sharply negative. A short entry triggered in that environment is not a stock you are shorting because it has broken technical support – it is a stock that is down because everything is down. When the selling pressure eases, it will recover regardless of its individual technical picture.
Market internals indicators measure market breadth – how many participants are moving in a particular direction – rather than just price direction. That breadth measurement is what makes them a leading signal for the kind of reversal Craig was experiencing, not a lagging one.
The approach scales naturally to global stock trading because Craig builds the indicator independently for each exchange, using only the stocks he is actually following on that exchange. He is not relying on published U.S. advance-decline data and hoping it generalizes – he is building a custom version appropriate to his actual universe of traded instruments.
The Creative Extension: Mean Reversion Entry Timing
After implementing the filter, Craig’s thinking went further. If a strongly negative advance-decline reading is bad for short breakouts, what is it good for?
His hypothesis: the same conditions that make short breakouts fail should be favorable for long mean reversion entries. When the indicator is deeply negative – when a large majority of stocks are declining – stocks that are in established uptrends may be temporarily oversold due to broad market pressure rather than fundamental deterioration. A limit entry below the recent close, on a stock above its 200-day moving average, after a deeply negative indicator day, could capture the mean reversion bounce.
The rules he tested:
- Stock must be above the 200-day simple moving average (established uptrend requirement)
- Previous day’s exchange advance-decline indicator must be below approximately -0.5 (broadly negative market)
- Entry: limit order a small distance below the latest closing price
- Exit: when the stock closes higher than the previous day’s close, exit the next morning at open
Out-of-sample testing with commissions and slippage: positive expectancy. The concept works. The average profit to maximum drawdown ratio (approximately 0.6) is not yet where Craig wants it – too many trades enter simultaneously when the indicator triggers, because the only filter is the market-level condition. Individual stock selection criteria need to be tightened. But the foundational idea is validated, and it is on his active research list.
What This Means for Other Traders
Craig’s approach demonstrates several things that are directly applicable regardless of what you trade:
| Principle | Craig’s Application |
|---|---|
| One idea, applied well, beats many ideas applied poorly | A single advance-decline filter produced 40% improvement in risk-adjusted returns |
| Adapt tools to your situation rather than using them as taught | He built exchange-specific indicators rather than relying on published U.S. data |
| The same signal can serve multiple purposes | A filter for bad short trades became a timing signal for long mean reversion trades |
| Drawdown thinking first | His target is max 25% drawdown; return is whatever the market gives him |
| Market internals generalize across instruments | The approach works on global stocks, not just U.S. futures, with the right breadth data |
Craig’s story also addresses a question that comes up often about market internals: can they be applied to stocks, and what about markets where standard advance-decline data is not published? His answer, demonstrated practically, is yes – but you need to build your own indicator. The data is available through your broker. The construction is straightforward. The difficult part is thinking about how to apply the concept, not the technical implementation.
The Broader Picture: Creativity in Trading
What Tomas particularly values about Craig’s approach is the creative thinking that turns a packaged solution into an extended research program. The Trading Market Internals course provides a framework and several specific techniques. Craig took one technique, understood its underlying logic well enough to adapt it to a different instrument universe, and then inverted it to find a complementary use case.
That kind of thinking – asking “what does this signal tell me about what else might be true?” – is what separates traders who get better over time from traders who plateau. There are many skilled coders. Good coding is not what separates successful systematic traders from unsuccessful ones. Ideas are.
For practical background on how breakout filters and market condition indicators interact, the trend hack that filters 80% of bad trades shows a similar logic applied to the breakout entry side. And for context on building the kind of robust strategy portfolio that market internals can improve, the Breakout Masterclass Part 1 is the right starting point.
Want to use market internals to improve your risk-adjusted returns? Check out Breakout Trading PRO. It’s a structured, visual strategy builder inside BreakoutOS, powered by the proven Mr. Breakouts Formula used by thousands of traders.