The convenience trap
A fixed twenty-dollar stop is easy to reason about and almost always wrong. It is too tight in high volatility, where ordinary noise removes you from correct positions, and too wide in quiet conditions, where it accepts far more risk than the setup justifies. Fixed distances make risk a function of the calendar rather than the market.
Volatility-scaled placement
Scaling the stop to recent realised volatility means the same logical distance expresses differently depending on conditions. Practically this means measuring a rolling average range and placing the invalidation beyond a multiple of it, then adjusting lot size so the cash risk stays constant even as the point distance changes.
- —Measure volatility on the trading timeframe with a lookback long enough to be stable but short enough to adapt.
- —Place the stop beyond structure — the swing that invalidates the idea — not at a round multiple of the range.
- —Recompute lot size from the stop distance so risk in currency terms is unchanged.
- —Respect the broker's minimum stop distance, and refuse the trade if structure sits inside it.
Why stops should never widen
The single most consequential rule in our architecture is that a stop can move toward profit and never away from it. Widening a stop converts a defined loss into an undefined one and is the mechanism behind almost every catastrophic retail account we have reviewed. It always feels justified in the moment, which is exactly why it must be structurally impossible rather than discouraged.
Trailing logic that protects gains without strangling trades
A trailing stop is a trade-off made explicit: every unit of protection costs a unit of room, and a trail that is too tight converts winning positions into small wins with impressive reliability. The useful design question is not how tightly to trail but at what point in a trade's life the objective shifts from participation to preservation.
Stage the trail rather than running it constantly
- —Leave the initial stop at structure until the trade has covered its own cost.
- —Move to breakeven only when doing so does not sit inside normal noise.
- —Trail on structure or a volatility multiple, never on a fixed point distance.
- —Never allow the trail to move against the position, under any code path.
Partial exits, targets and the arithmetic of scaling out
Taking partial profit feels prudent and mathematically reduces expectancy in most trend-following contexts, because it removes size from precisely the trades that would have paid for the losers. It can still be correct — but as a psychological instrument with a known cost, not as a free improvement.
Decide what the partial is for
Everything discussed here is applied on a public, third-party verified account — updated continuously, losing weeks included.
