How fast an account climbs out is a compounding statistic, not a comfort statistic.

The bridge

Two halves of the account's life, measured by one metric.

An account's equity history alternates between advancing and recovering, and the two halves are usually analysed by different machinery — growth by expectancy and conversion instruments, drawdown by the gate ladder and the risk distributions. Acceleration is where they meet. The metric tracks the second derivative of equity, so it registers both how strongly the advancing periods climb and how quickly the recovering ones stop subtracting. An account with excellent expectancy and slow recoveries and an account with modest expectancy and fast ones can produce the same annual figure by very different routes, and acceleration is the reading that distinguishes them.

What slow recovery costs

Time spent regaining old ground is time not compounding.

The arithmetic is unforgiving in a way that depth arithmetic makes intuitive but duration arithmetic hides. Capital deployed while below a prior peak is working to restore a level the account already achieved, and none of that work advances the compounding curve — it repairs it. An account spending a large share of its year underwater is running an engine whose output is largely absorbed by repair, and its annual figure will understate its per-trade quality while accurately describing its growth. This is the honest reason recovery speed belongs in a growth metric rather than in a comfort discussion.

The designed slowness

Some of the cost is deliberate, and reading it as a fault is the error.

The system slows its own recoveries on purpose. Deeper gates cap the tiers that would climb fastest, so an account emerging from a genuine drawdown does so at compressed deployment — and the simulated distribution reflects exactly that, since every modelled path lived under the same caps. The result is a signature the raw statistics would call inefficient and the doctrine calls the purchase price: shallower bottoms bought with slower climbs. A recovery velocity inside the modelled distribution is therefore the system executing its trade correctly, however unsatisfying the pace feels, and no reading of the acceleration metric authorises accelerating it by override.

The reading that matters

Velocity against its distribution, not against impatience.

As with every other quantity in this cluster, the number is interpretable only against its modelled counterpart. The recovery-time distribution reports how long governed paths take to climb out from various depths, which converts a lived recovery into a placement. Faster than modelled, with deployment inside its normal range, is a favourable reading worth documenting. Slower than modelled is the finding that routes to the conversion investigation — under-deployment during the climb, friction, or a mix that neither loses nor gains. And slower than modelled while deployment sits below the cap is the specific, common, and correctable case.

The early-warning property

Fading acceleration precedes a flattening curve by weeks.

Acceleration's value is that it turns before the level does. Positive slope with fading acceleration describes a machine still climbing while losing thrust, which is the earliest visible signature of edge decay or friction creep — and it appears while the equity curve still looks entirely healthy to anyone watching the level. Recovery velocity contributes directly to that signal: an account whose climbs are progressively slower at comparable depths is producing the drawdown-side version of the same warning. Neither reading is a verdict on its own, and both feed the persistence machinery that decides whether anything real is happening.

• Recovery velocity is read against the modelled recovery-time distribution, never against patience.
• Slower-than-modelled with deployment under the cap is the common correctable case.
• No acceleration reading permits overriding the gate to speed a recovery.

The key idea

Survival and compounding are the same subject viewed from two ends.

The drawdown distributions look like risk instruments and the acceleration metric looks like a growth instrument, and recovery velocity is the quantity that reveals them as one thing. How deep the holes get, how long they last, and how fast the account climbs out are jointly the determinants of whether an edge becomes an equity curve. The gate ladder shapes all three deliberately, trading speed for shallowness, and the acceleration read is where the terms of that trade become visible as a number rather than as a feeling about pace.