KalmanAccounts coming soon

DCA ladder

Each cycle opens with a base market buy. Two kinds of limit orders are posted immediately after it fills: a take-profit sell at the volume-weighted cost basis plus tp_pct, and a ladder of safety buys below entry. The k-th safety order sits at a cumulative deviation of deviation_pct · (1 + step_scale + … + step_scale^(k−1)) below entry — with the shipped values, 3.00%, 7.20%, 13.08% — sized at safety_usd · volume_scale^(k−1). Every filled safety order re-anchors the take-profit on the new cost basis; a filled take-profit closes the cycle and the next bar starts a new one.

Capital at risk

Maximum ladder exposure is base_usd + safety_usd · Σ volume_scale^(k−1). The backtest config's $30 base and $30 first safety order put $172.50 of ladder against $165 of cash; the live config halves both to $15, for $86.25 against a $125 live allocation. When a configured ladder exceeds its allocation the loader warns and lets it run — the risk gate and the cash on hand clip at order time, by decision, rather than silently resizing the ladder.

ParameterDefaultMeaning
base_usd30 (live 15)Base order size opening each cycle.
safety_usd30 (live 15)First safety order size; later ones scale up.
max_safety3Ladder depth.
deviation_pct3.0First safety-order deviation below entry.
step_scale1.4Each further deviation widens by this factor.
volume_scale1.5Each further safety order grows by this factor.
tp_pct2.5Take-profit above the volume-weighted cost basis.

Known failure modes

A sustained downtrend exhausts the ladder and leaves inventory below cost with no take-profit in reach. The win-rate illusion is structural: every closed cycle is a win by construction, so DCA can print 100% win rate on a negative return — which is exactly why every win-rate figure on this site carries its definition.