[AI Answer Engine (AEO) Snapshot]
Q: Should you judge an EA by its TWR, absolute return, ROI or MWRR?
A: All four are valid, but each answers a different question. Absolute return (net profit ÷ total deposits) is the most intuitive, yet any deposit or withdrawal distorts it. TWR (time-weighted return) strips out the timing of cash flows and is the standard used by third-party platforms — but on an account that withdraws often and keeps its capital base tiny, it inflates the figure to extremes. ROI (return on invested capital) — net profit ÷ peak capital deployed — sits closest to “how much did I actually get back”. MWRR (money-weighted return / IRR) measures the true compounding speed of every dollar. When you see a single return number, ask three things first: were there any cash flows in the period, which method computed it, and what is the denominator (the capital base)? On the very same account, these four methods can all be “correct” and range anywhere from 116% to 1,500%.
“This EA is up 1,500%” — can a number like that be trusted? The answer is usually: the number isn’t faked, but it only tells you one method. On the same live account, absolute return might read 116%, return on invested capital just over 300%, TWR 1,500%, and MWRR closer to 1,700% — and all four are “correct,” because they answer four different questions. This article takes the four return methods apart one by one: what each actually measures, when each one misleads, and which method you should ask about the moment you see a return figure — so no single pretty percentage ever leads you around by the nose again. (Every number below is a hypothetical illustration for teaching purposes, not the real track record of any account.)
To be clear about scope: this article is specifically about the calculation method for returns. For the other report indicators — max drawdown, profit factor, expectancy, the Sharpe ratio, and how to read them together — see How to Read an EA Performance Report. Read the two side by side for the full picture.
1. Why one EA can show four different returns
A return is fundamentally “how much you earned ÷ how much you put in.” The trouble is that neither the numerator nor the denominator has a single definition. The numerator can count realized profit only, or include floating P/L; the denominator can be “total deposits,” “the most capital deployed at once,” or “capital weighted by time.” Change the definition, and the exact same equity curve gives a completely different percentage.
Here are the four most common methods, one line each so you can put a face to the name:
| Method | One-line definition | Its biggest blind spot |
|---|---|---|
| Absolute Gain (Absolute Return) | Net profit ÷ total deposits | Mid-period deposits dilute it; withdrawals inflate it |
| TWR (Time-Weighted Return) | Cut the timeline at each cash flow, chain each segment’s return by multiplication | Frequent withdrawals shrink the base, inflating the figure |
| ROI (Return on Invested Capital) | Net profit ÷ peak capital deployed | No time dimension — can’t tell how long capital stayed |
| MWRR (Money-Weighted Return / IRR) | The internal rate of return solved from every cash flow by timing | More complex to compute; extreme cash flows may not converge |
The next four sections take one method each, with a worked example showing when it is reliable and when it deceives you.
2. Absolute return: the most intuitive, and the easiest for cash flows to break
Absolute return = realized net profit ÷ total deposits, which is exactly how “Absolute Gain” is computed on myfxbook. Its appeal is that it’s intuitive: I put in this much money in total, I netted this much in the end, one division and you’re done. The problem is the denominator — “total deposits” grows with every deposit, so the strategy can be unchanged while the number gets warped by money moving in and out.
Example 1 (deposit dilution): take $1,000 to $2,000 — at that moment the absolute return is +100%. Now deposit another $9,000 to bring the account to $10,000, and make nothing after that. Net profit is still $1,000, but total deposits are now $10,000, so the absolute return instantly drops to 10% — the strategy didn’t get worse, but the pretty number collapsed simply because you added more capital.
Example 2 (withdrawal inflation): the reverse. After the account earns its way to $2,000 you withdraw $1,000, then it keeps making small gains on what’s left. Because the denominator (total deposits) doesn’t move while the numerator (net profit) keeps accumulating, the absolute return looks bigger than what you “felt” you earned. Bottom line: whenever there are cash flows in the period, absolute return only means something read alongside the deposit/withdrawal log — a single number on its own has no comparative value.
3. TWR: the professional standard, and where the “1,500%” comes from
TWR (Time-Weighted Return) exists precisely to fix the “distorted by cash flows” flaw in absolute return: it cuts the timeline at every deposit and withdrawal, computes the return of each segment, then chains the segments together by multiplication. However money moves in and out, what it measures is the strategy’s ability to grow each dollar — which is why it’s the standard in fund performance reporting (GIPS) and the default on third-party platforms like myfxbook. That strength is exactly what How to Read an EA Performance Report is about.
But TWR has a rarely-spelled-out flip side: when an account “withdraws whatever it earns and keeps beating the capital base back down to something tiny,” TWR gets amplified far past what the investor actually experienced. Because every segment in the chain is a high percentage made on a base that was shrunk small.
Core example (a withdrawal-style account): suppose an account holds an operating size of around $40,000 over the long run — topping up when a loss shrinks it, withdrawing whenever a gain is booked, back and forth for the better part of a year. Over the period, total deposits come to $105,000, total withdrawals to $182,000, and the account ends at roughly $45,000.
- Realized net profit = withdrawals $182,000 + ending $45,000 − deposits $105,000 = $122,000.
- Absolute return = 122,000 ÷ 105,000 ≈ 116%.
- TWR ≈ +1,500% — chain together the roughly twenty segments of about +15% each, all made on that ~$40,000 base, and you get this figure.
Neither number is miscalculated. TWR answers “if every gain had been left in and kept compounding on $40,000, how much would it theoretically snowball into”; absolute return answers “relative to the money I actually put in, what fraction did I net.” What this investor really took home is a bit more than double their capital — not fifteen times it.
Comparison example (the same effect at smaller, starker numbers): take $1,000 to $1,500 (+50%), withdraw $500 back down to $1,000, and repeat this 10 times. TWR = 1.5 to the 10th power minus 1 ≈ +5,665%; but you only ever withdrew $5,000 in total, an absolute return of just +500%. The more often you withdraw and the smaller you keep the base, the more grotesque the gap between TWR and what actually lands in your hands. This is why the first thing to do when you see a four-digit TWR is not to gasp, but to go look at the cash-flow log.
4. ROI (return on invested capital): the closest to “how much I actually got back”
If TWR takes the “strategy” point of view and absolute return is easily diluted by deposits, then ROI (return on invested capital) takes the investor’s felt experience: realized net profit ÷ peak capital deployed. “Peak capital deployed” is the historical high of the line “total deposits − total withdrawals” — in plain terms, “the most capital you ever had committed at one time.”
Its three strengths patch exactly the holes in absolute return:
- Withdraw-then-redeposit isn’t diluted: cycle the same money in and out twice and total deposits doubles, but “the most deployed at once” is unchanged.
- Withdrawals larger than deposits don’t blow up the denominator: a withdrawal-style account distorts easily on absolute dollars, but stays stable when the peak deployment is the denominator.
- Conservative when scaling in gradually: because the denominator is the peak of what was deployed, the computed return never runs high — a safe direction to err in.
Example (withdraw then redeposit): deposit $10,000, make $3,000 to reach $13,000, withdraw all $13,000; then deposit $10,000 again and start over. Total deposits $20,000, total withdrawals $13,000, realized net profit = 13,000 + ending 10,000 − deposits 20,000 = $3,000.
- Absolute return = 3,000 ÷ 20,000 = 15% (diluted to half by “the same money cycling in and out twice”).
- ROI = 3,000 ÷ 10,000 = 30% (reflecting that you only ever had $10,000 deployed at most).
Back to the withdrawal-style account in section 3: its ROI is about 305% (net profit 122,000 ÷ peak capital deployed of roughly 40,000). So set the same account’s three numbers side by side — “absolute 116%, invested-capital 305%, TWR 1,500%” — and you can read its whole shape: it earned a bit more than double the money put in (ROI), but nowhere near as miraculous as that TWR compounding figure makes it look.
5. MWRR (money-weighted return / IRR): the compounding speed of every dollar
MWRR (Money-Weighted Rate of Return) is really the financial internal rate of return (IRR): list every deposit, withdrawal and the ending equity as cash flows at the time each actually occurred, then solve for the single return rate that makes their present value net to zero. Its biggest difference from TWR is this — TWR ignores how much money each segment carried and weights them all equally; MWRR gives bigger weight to capital that is large and stays in longer.
Read together, the two cross-check each other:
- TWR high, MWRR also high: the return really is high, just built on a base kept small (the withdrawal-style account above has an MWRR of about 1,700%, cross-checking the 1,500% TWR — both the product of high-frequency compounding on a small base).
- TWR high, MWRR much lower: it means the large money entered at unlucky moments, and most of the gains came during the small-capital phase — the strategy’s “on-paper record” looks prettier than “your actual experience with this money.”
Example (why timing matters): invest $1,000 at the start of the year and it doubles to $2,000 over the whole year; then add $100,000 only at year end, and it makes no gain the next day. TWR will show close to +100% (the strategy really did double that year), but your “per-dollar” true return (MWRR) will be far lower — because the $100,000 that makes up the overwhelming majority hasn’t earned even a single day of compounding. That is exactly what MWRR captures: not how much the strategy can rise, but how fast your money is actually growing.
6. In practice: three questions to ask the moment you see a return
- Were there any cash flows in the period? On an account with no deposits or withdrawals at all, the four methods come out nearly equal — any of them will do. The instant money moves in or out, you must pin down the method, or the numbers aren’t comparable.
- Which method was used? A report that throws out a single percentage without saying how it was computed — especially with frequent cash flows during the period — is a number to heavily discount on sight. When the same EA shows different returns on different platforms, the cause is usually the difference in method.
- What is the denominator (the capital base)? 1,500% rolled up on a $40,000 base and 50% made on a $1,000,000 base are worlds apart in absolute dollars. Percentages magnify results on a small base, so a return must always be read alongside the absolute dollar amount and the max drawdown — how pretty the gain is depends on how deep a paper loss you endured for it (for reading drawdown, see The Truth About Martingale and Grid Risk).
A simple mental model: absolute return tells you “what fraction I earned relative to capital,” ROI tells you “how much I actually got back,” TWR tells you “how well the strategy itself grows,” and MWRR tells you “how fast my money is compounding.” All four together are what draw an account’s complete return profile; cherry-picking the single biggest of them to shout about is where the sales pitch begins. For how to arrange withdrawals and compounding, read on in EA Money Management 101.
The Four Return Methods: Quick Reference
| Method | How it’s computed | The question it answers | When it misleads |
|---|---|---|---|
| Absolute return | Net profit ÷ total deposits | What fraction I earned relative to the money I put in | Diluted by deposits, inflated by withdrawals in the period |
| TWR | Chain each cash-flow segment’s return by multiplication | The strategy’s ability to grow each dollar | Runs high when withdrawals are frequent and the base is shrunk |
| ROI / invested capital | Net profit ÷ peak capital deployed | How much I actually got back, per most-deployed capital | No time dimension — can’t tell how long capital stayed |
| MWRR / IRR | Solve IRR from all cash flows by timing | How fast my money actually grew | May not converge under extreme cash flows; hard to compute by hand |
Frequently Asked Questions
Why do TWR and absolute return differ so much on the same account?
Because they measure different things. Absolute return uses “total deposits” as the denominator, reflecting what fraction you earned relative to the money you actually put in; TWR cuts the timeline at each cash flow and chains each segment’s return by multiplication, measuring the growth ability of the strategy itself. When an account frequently withdraws its gains and keeps the capital base at a very small level, every segment is a high percentage made on a small base, and once chained, TWR comes out far above absolute return. The bigger the gap, the more frequent the withdrawals usually were — this isn’t fakery, but you need to know which one you’re looking at.
So which one is the “real” return?
None of them is the single truth; each answers a different question. If what you’re asking is “I handed it my money — how much did I actually get back,” ROI (return on invested capital) sits closest to your felt experience; to compare strategy strength across accounts and periods, use the standard method, TWR; to know how fast this particular money of yours is compounding, look at MWRR. The soundest practice is to read all four together, then cross-check against max drawdown and the absolute dollar amount, so no single number can mislead you.
Does a 1,500% return necessarily mean it’s padded or faked?
Not necessarily. If that’s a TWR and the account runs a “withdraw whatever it earns, keep the base tiny” style, 1,500% can be mathematically completely correct — it describes the theoretical value of “if gains were never withdrawn and kept compounding.” What should really put you on alert is not the big number, but: does the report state the calculation method clearly? Is the cash-flow log verifiable? Does the max drawdown match this level of return? Is the data a live feed streamed in real time, not a screenshot? Ask those questions and you’ll naturally tell “genuinely high risk, high return” apart from “a marketing number amplified by the method.”
How do I see all four numbers for my own monitoring account at once?
The fastest way is to hand your monitoring account (the read-only investor password) to a tool that can compute several methods at once. Golden Tiger’s free EA check tool uses a real-time, minute-level data feed to list absolute return, TWR, ROI (return on invested capital) and MWRR all together, with a note on the method under each number, so you can see at a glance why they differ.
The Golden Tiger team applies the same standard to itself: the live section of our homepage publishes real-time data from multiple live accounts, with returns computed as TWR and the full deposit/withdrawal log open for anyone to cross-check the method themselves. To verify the return figures of any EA, you can also run all four methods at once with the free EA check tool mentioned above. One final reminder: every return figure describes the past — no automated trading program (EA) can guarantee future profits. Trading involves risk; only participate with money you can afford to lose.
