Results
The purpose of this page is simple: to show that the idea behind GBS was not built from a handful of attractive chart examples. It was researched over a large historical sample and evaluated through a walkforward process designed to reflect changing market conditions.
Past performance does not guarantee future results.
On this page
PnL Charts
These visuals are the clearest summary of how the framework behaved through the walkforward record.
Sizing view
Switch between raw R results, fixed-dollar risk, or percentage-risk compounding.
R shows the raw account-independent research curve, where every trade is measured in multiples of risk.
Interactive PnL chart
Rendered directly from the latest public trade list so the curve can be refreshed by updating the data file.
Interactive drawdown chart
Red shows drawdown below zero while blue shows the number of days spent underwater above zero, both resetting when the equity curve makes a new high.
Research stats
A fuller statistical summary for readers who want to inspect the walkforward record beyond the charts.
The value of these results is not that they remove uncertainty. The value is that they show the framework behaving through time under a defined testing process, rather than relying on isolated examples or a one-off in-sample fit.
Assumptions
The research uses a locate assumption of $0.10 per share. That was chosen because it covers most real-world locate cases without pretending locates are free or always easy to source.
In live trading, locate costs can range from effectively zero to roughly $1.00 per share depending on broker, locate provider, stock availability, and current demand. Some trades in the real world have to be passed on entirely because the locate cost is simply too expensive.
The research uses slippage equal to 0.25 x ATR(3). The reason is that real-world slippage depends on how quickly the market is moving, not on one static dollar or percentage number.
Using the average true range of the last three candles gives the model a simple way to scale slippage with current volatility. It is not perfect, but it is more realistic than pretending the market moves with the same friction at all times.
The study also assumes an execution fee of $0.0035 per share. That keeps the research from implicitly treating entries and exits as free.
In practice, commissions and routing fees vary by broker and venue, but including a per-share fee helps the results stay closer to real trading conditions.
Validation
The research objective was not to produce a single polished backtest. The goal was to see whether the backside reversal framework could remain useful as the market evolved.
That matters because micro-cap gapper behaviour changes over time. A model that only looks good in one fixed regime is far less interesting than one that continues to make sense when conditions shift.
Why this matters
The framework was evaluated as market conditions changed, not just optimized once and left frozen in the past.
The research includes practical assumptions such as slippage and locate-related costs, which makes the results more useful than idealized fills.
The model is reviewed and re-optimized over time because the micro-cap market does not stay static, and the framework should not pretend otherwise.
Interpretation
Caveats
Real fills, borrow availability, halts, and fast market conditions can produce outcomes that differ materially from model assumptions.
Even a well-tested framework can degrade if market behaviour changes far enough. That is one reason GBS is maintained and re-optimized rather than treated as static.
The research record can support confidence in the framework, but it should not be interpreted as a promise of returns or a substitute for risk management.
