Overtrading
Taking more positions than a method supports. Costs scale linearly with frequency while edge does not, so activity beyond a point is a direct transfer of capital to costs.
MadStockAlerts Research · Updated August 28, 2026
What to take away
- Every additional trade carries the full cost and a diluted edge.
- Revenge trading after a loss is the most damaging form of it.
- Boredom and the feeling of needing to act are common triggers.
- Doing nothing is a position, and it has no cost.
- The marginal trade is by definition a worse one than the last.
- Frequency is a risk parameter, because it scales the certain quantity (cost) and leaves the uncertain one (edge) alone.
MAD Academy Training Video · 0:46
Activity Feels Like Progress
Trading more feels productive and costs money at every step, which is why frequency needs a limit that is not based on how you feel.
This lesson is part of a Stock Alerts + Tools plan.
The arithmetic
Costs are certain and scale with the number of trades. Edge is uncertain and, past the best opportunities a method identifies, declines with each additional trade because the marginal trade is by definition a worse one.
net result = (trades x average edge) - (trades x average cost)
- adding trades whose edge is below the cost reduces the total, regardless of how many win
The uncomfortable implication is that a profitable method can be made unprofitable purely by running it more often, without any single decision being obviously wrong.
The triggers
| Trigger | The thought behind it |
|---|---|
| Revenge | Recovering a loss immediately, in the same session |
| Boredom | Being in the market feels like working; waiting does not |
| Overconfidence | A winning run read as skill rather than variance |
| Sunk effort | Hours of research feel wasted if no position results |
| Adrenaline | Volatility itself is engaging, independently of any opportunity |
Revenge trading
This is the most destructive of them, because it combines a larger size with a lower-quality setup at the exact moment judgement is most impaired. Most single-session disasters are this pattern rather than one bad initial decision.
Scroll the chart sideways to see all of it.
The sunk cost variant
Research already conducted is spent whether or not a position follows. Entering a marginal trade because a great deal of work went into the analysis is the classic sunk cost error, and it is common enough in markets to be worth naming.
The correct reading is the opposite: thorough work that concludes there is no trade has produced exactly the outcome it was for. A research process that only ever ends in a position is not a research process.
Cash is a position
Holding nothing has no cost, no slippage and no risk. There are extended periods in which a given method identifies nothing that qualifies, and taking nothing during them is the method working rather than failing.
This is easier to accept for a method with an explicit universe filter, because the absence of candidates is visible rather than inferred. A scan returning nothing is a result.
Frequency as a risk parameter
How often to trade is usually treated as a style preference and is more accurately a risk setting, because almost everything that determines an outcome scales with it. Costs scale linearly. Exposure to execution error scales with it. The number of independent decisions, each of which can be wrong, scales with it. What does not scale with it is the supply of situations worth acting on.
required edge per trade = annual cost drag / trades per year
- cost drag includes spread, commission, financing and the tax consequence of short holding periods
- the required edge is what is needed to break even, before any profit
The arithmetic is unforgiving at high frequencies. A round-trip cost of two tenths of a percent is negligible once and is forty percent of capital across two hundred round trips, which means every one of those trades has to clear a bar that a quarterly holding period never encounters. Nothing about the quality of the analysis changes; the hurdle does.
The asymmetry that makes this a risk parameter rather than a cost calculation: the costs are certain and recur on every trade, while the edge is uncertain and may not exist at all. Raising frequency raises the certain quantity and leaves the uncertain one unchanged.
What activity is substituting for
Trading more than a process calls for is rarely a belief that more opportunities exist. It is usually a response to something else, and the substitutions are consistent enough to be named.
| What is happening | What the trade is standing in for |
|---|---|
| A loss earlier in the session | Undoing it, which requires acting now rather than acting well |
| A long stretch with no setup | Evidence of working, since waiting produces none |
| A run of wins | Confirmation that the run was skill, tested by doing more |
| Research that produced no position | Recovering the effort, which the position cannot do |
| A large move happening elsewhere | Participation, independently of whether the situation is readable |
The practical value of naming them is that each has a different tell and none of them feel like the description above from the inside. In the moment, every one of these presents as a genuine opportunity that happens to have appeared at that time, which is the part that makes an after-the-fact record more useful than an in-the-moment judgement.