Measuring your actual returns (you're probably doing it wrong)

Ask most investors their annual return and they'll do something like: balance now, minus balance last year, divided by balance last year. If you added or withdrew money during the year — which is everyone — that number is wrong, sometimes wildly. And without an honest return number compared against an honest benchmark, you literally cannot know whether anything you're doing is working.

The two kinds of return

Time-weighted return (TWR) measures how well your INVESTMENTS performed, ignoring the timing of your deposits — it's what fund fact sheets report and the right yardstick for comparing against an index. Money-weighted return (IRR) measures how well your DOLLARS performed, including the timing of your contributions — it's your personal experience. They diverge whenever you add money before a rally or a crash. Both are legitimate; confusing them is how people fool themselves in both directions.

Good-enough math for regular people

1. Simple approximation: return ≈ (ending balance − starting balance − net contributions) ÷ (starting balance + half of net contributions). Accurate within a fraction of a percent for steady monthly investing.
2. Better: use your brokerage's 'personal rate of return' figure (they compute real money-weighted returns) — most major brokerages display it, and almost nobody looks.
3. Best, if you like spreadsheets: the XIRR function with every dated cash flow gives your exact money-weighted annual return.
4. Whatever method you pick, use the same one every year, in a note or spreadsheet you keep. Memory is the least reliable financial instrument you own.

Benchmarking without self-deception

• Compare against a blended benchmark matching YOUR allocation — a 70/30 portfolio should be measured against 70% total market / 30% bond index, not against the S&P 500. Trailing an all-stock index in a year your bonds did their job is not underperformance.
• Include everything: the experimental account, the old 401k, the individual stocks. Counting only winners is called cherry-picking when fund managers do it; it has the same name at home.
• Measure over 3–5 year windows. One year is noise; five years starts to be signal.
• The question that matters: 'Did my active decisions beat what the boring three-fund version of me would have earned?' If the answer is no for five straight years, the boring version of you is the better investor — hire them.

The bottom line

You can't improve — or trust — what you don't measure honestly. Compute one true return number a year (your brokerage will do the money-weighted math for you), compare it to a benchmark that matches your actual allocation, and judge over five-year windows. Most people who do this discover the boring index portfolio in them was the star performer all along. That discovery is worth more than most stock tips: it's the one that ends the expensive experiments.

A worked example: the flattering number vs. the true one

Say your account showed $100,000 on January 1 and $130,000 on December 31, and you contributed $20,000 during the year. The naive calculation — up $30,000 on $100,000, a 30% return — is the one your brain wants to keep, but $20,000 of the increase was simply your own deposits. The money-weighted truth depends on timing: if the $20,000 arrived mid-year, your investments earned roughly 9%, not 30%. The same illusion works in reverse to hide failure: an account that ends the year flat despite $20,000 of contributions actually lost about $20,000 of market value, a fact the ending balance politely conceals. This is why 'my account is up' is nearly meaningless without separating flows from growth — and why brokerages were historically happy to show balances instead of returns. Most major platforms now bury a true time-weighted or money-weighted return figure somewhere in the performance tab; finding it once a year tells you more than twelve months of balance-watching ever will.

1. 1
   Locate the real number

   In your brokerage's performance section, find 'personal rate of return' (money-weighted) or 'time-weighted return' — not the change in balance. Note both annual and since-inception figures.

2. 2
   Compare against a fair benchmark

   A 70/30 portfolio should be judged against a 70/30 index blend, not the S&P 500. Vanguard publishes index blend returns; or use a target-date fund with your allocation as a ready-made yardstick.

3. 3
   Check after fees, after inflation

   Subtract roughly 3% (2025 inflation is near the Fed's 2-3% range) to see real growth, and confirm advisory fees are reflected in the reported figure — they often are not.

4. 4
   Write down one line per year

   Date, balance, contributions, return, benchmark. Ten minutes annually builds the only performance record that can ever tell you whether your strategy — or your tinkering — is working.

The habit pays for itself the first time it catches a gap: investors who discover they have trailed their own benchmark by two points a year rarely need a lecture about tinkering — the number delivers it.

You manage what you measure, and in investing the honest measurement takes ten minutes a year.

Start the log this year even if past records are messy — the since-inception numbers your brokerage already stores are good enough to seed it, and future-you will inherit the one document that separates strategy from story.