What Is Time-Weighted Return (TWR)?

There's one troublesome thing when measuring returns: the case where 'money is put in or taken out midway.'

For example, the fund manager just kept managing the fund, but if investors happen to put in a large sum right before the market rises, the account balance swells greatly. Conversely, if they put it in right before a decline, the loss amount grows.

This difference arises not from the manager's skill but from the 'timing of putting money in.' So to fairly compare only the manager's pure management skill, you need a way to strip out the effect of these contributions and withdrawals.

That's exactly what the Time-Weighted Return (TWR) does. Regardless of when and how much investment came in, it measures only how well the asset itself performed.

TWR's core idea is simple. You divide the whole period into several segments based on 'when money came in or went out,' then compute each segment's return separately and chain them together by multiplication.

The formula looks like this.

TWR = [(1 + R1) × (1 + R2) × ... × (1 + Rn)] − 1

Here R1, R2 ... are each segment's return. Why multiply instead of add? Because returns accumulate through compounding. The next segment's return applies again on top of the money earned in the first segment, so instead of simply adding, you must multiply (1+return) in sequence for accuracy.

Doing this makes 'which segment the large sum was in' disappear from the calculation, and only the performance of each segment is connected purely.

Let's follow along with actual numbers. (We present this overseas example in its original units rather than won.)

First, you start investing with $100,000 in starting capital.

Second, six months later the assets have grown to $110,000. First-segment return R1 = ($110,000 − $100,000) ÷ $100,000 = +10%.

Third, right after that, the investor adds $50,000, bringing the balance to $160,000.

Fourth, at year-end the assets have become $152,000. Second-segment return R2 = ($152,000 − $160,000) ÷ $160,000 = −5%.

Now put it into the formula.

TWR = (1 + 0.10) × (1 − 0.05) − 1 = 1.10 × 0.95 − 1 = +4.5%

The large sum (the added $50,000) happened to be exposed to the −5% loss segment, but TWR ignores that timing and looks only at the performance itself—'the asset rose +10% and then fell −5%.' So the result is +4.5%.

In the example above, the return from the manager's perspective (TWR) was +4.5%. But if you calculate 'the return the investor actually earned,' it comes out much lower, at roughly +1.6%. Why the difference?

Because the investor put the large sum ($50,000) in right before a loss segment. A return that reflects even 'when and how much was put in and taken out' is called the Money-Weighted Return (MWR), and it's the same concept as the internal rate of return (IRR).

To summarize, you can divide it this way.

First, TWR (time-weighted) = the pure performance of the asset/manager. Independent of contribution timing. Suitable for comparing skill across funds.

Second, MWR (money-weighted, IRR) = the result you actually earned. It reflects even whether your contribution timing was good or bad. Closer to 'your account's report card.'

It's not that one is right and the other wrong—they serve different purposes. The reason a fund's marketing return feels different from your account's return is usually this difference.

'The Return of Almost Everything' is a site we operate that directly shows the results of buying good assets for a long time, steadily.

Putting a lump sum in all at once versus contributing monthly in installments have completely different 'timing of exposure to the market.' So even for the same asset over the same period, the felt return differs. That's precisely the principle that creates the difference between TWR and MWR.

You can directly compare how the two methods differ in the lump-sum and recurring-investing simulators. Also, we don't show returns only prettily—we display the maximum drawdown, loss duration, recovery period, fees, and exchange-rate effects together. Because there's no asset that only goes up.

This article is not one that tells you to buy some stock or predicts future prices. Please think of it as a tool that teaches you 'how to read the number called return without misunderstanding it.'