TWR Series (time-weighted)

The TWR (time-weighted return) series is EquityTruth's cashflow-neutral equity curve, the GIPS standard for comparing managers. It removes the distortion of deposits and withdrawals so the curve reflects pure trading skill — use it to compare strategies, not your personal money-weighted return.

Computed from
Data series
Scope
Single report
Range
Any real number
Direction
Context-dependent

Variants: close-to-closelow-to-low

The TWR series (time-weighted return) is the equity curve with your deposits and withdrawals (cashflows) stripped out — the standard professional fund managers are measured by (GIPS). It answers "how did the strategy itself perform," regardless of when you moved money in or out. When you compare two track records, this is the curve to use — but it is not the curve that tells you what you personally earned (see below).

How it's built

r_t = (V_t − V_{t-1} − CF_t) / (V_{t-1} + CF_t)
TWR = (∏(1 + r_t) − 1) × 100%
V_t
account value at bar t
CF_t
net cashflow (deposit +, withdrawal −) at bar t
r_t
the bar's return with the cashflow removed from both numerator and denominator
the product symbol — multiply all the (1 + rₜ) terms together; chaining those sub-period returns is what cancels the cashflows out
In this product: Computed at M1 bar resolution, with a fresh sub-period started at each cashflow event — that is exactly how the cashflow is "removed." Before the first bar the portfolio didn't exist (V₋₁ = 0).

By breaking the curve into sub-periods at every deposit/withdrawal and chaining their returns, the size and timing of the cashflow cancel out — what's left is the return the strategy produced on whatever capital was present.

Time-weighted does not mean your entry/exit timing is removed. It removes the distortion of money moving in and out of the account. The skill, luck, and timing of the trades themselves are fully reflected — that's the whole point.

A great TWR does not mean you made money. TWR scores the strategy; it says nothing about your dollar result, which depends on how much capital you had in at each point. Classic trap: a curve that's flat, then rips higher, shows a strong TWR — but someone who deposited heavily right before the flat stretch (or at the top) can still be down in real money. For "did the strategy perform?" read TWR; for "did I make money?" read the Return series (money-weighted). Neither is "better" — they answer different questions.

The negative-equity fallback (why TWR and Return can diverge)

Edge case — skip this unless an account in your comparison ever went to zero or negative. Normal accounts are unaffected.

There's one rare case worth understanding. If floating losses drive equity below zero (a margin-call situation), 1 + r_t goes negative and the cumulative product flips sign — which would permanently corrupt the TWR chain. At the first bar where that happens, EquityTruth splices the TWR index onto the Return (Gain) index, scaled so the junction is continuous. On a cashflow-free account the two are mathematically identical right up to that splice point (verified to 0.000000% on real data); after it, TWR tracks Return dynamics. This only ever activates for accounts that go negative — it's the one place TWR and Return legitimately diverge for that reason.

Close-to-close vs low-to-low

Like every equity series, TWR has a close-to-close variant (one value per day-close; used by Time Under Water) and a low-to-low variant (intraday low to intraday low; used by Sharpe, Sortino, volatility). See the full explanation on the Return series.

The Return series is the money-weighted counterpart; the P&L series is absolute dollars. For judging a strategy's pure performance, TWR is the one to read.

Related metrics

Further reading