How Tradune calculates your real return

Money-weighted return vs a broker’s simple return

Most brokerage apps show a simple return: current value divided by cost basis, minus one. That number ignores when money went in. If you added a large deposit the week before a market drop, a simple return blends that new money in as if it had been invested the whole time, distorting the picture of how your decisions actually performed.

Tradune instead computes XIRR — the annualized rate that makes the net present value of your dated cash flows equal zero, using actual/365 day-count and solved numerically. In the app, the solver is a bisection on net present value, which refuses to return a number at all when no sign change brackets a root. The free beat-the-market calculator on this site runs the same solver in your browser — the same bisection and day count, checked against the app on a shared set of test cases — on figures you type in rather than on recorded transactions. Where more than one rate fits a set of cash flows, the calculator says so instead of showing one. XIRR reflects both the size and the timing of every flow, which is why two investors with the same trades but different deposit timing can show different money-weighted returns even on the same stock.

Two modes: investment return and account return

Tradune reports exactly one of two modes for a given return figure — they are never mixed.

• Investment return (trade flow). Each buy is a cash outflow, each sell is a cash inflow, and your current holdings value is treated as the final inflow. This measures the timing and sizing of your trades on the securities themselves. Idle, uninvested cash sitting in the account is not counted.
• Account return. When you record deposits and withdrawals, those become the cash flows, and the terminal value is your full account value, including cash. This measures how the whole account performed, cash drag included.

"The same cash in SPY": the benchmark twin

To answer "how did I do compared to the market," Tradune builds a benchmark twin: it takes your exact dated cash flows — the same amounts, on the same days — and invests them in SPY at that date’s dividend-adjusted (total-return) closing price instead of your actual trades. The twin’s terminal value is constructed the same way as yours, and the same XIRR solver is run on it. The twin is modeled without trading costs, so any commissions you recorded weigh only on your side of the comparison — a deliberately conservative choice that can narrow, never inflate, your edge.

We use the dividend-adjusted price series, not a raw index-price overlay, because a price-only comparison would silently flatter your return in years SPY paid meaningful dividends, or penalize it in ways that have nothing to do with your decisions. The dividend-adjusted twin answers a specific, honest question: what would this same money, moved on these same days, have earned in SPY with dividends reinvested?

Why we call it "excess return," not "alpha"

Your excess return (also shown as "your edge") is your money-weighted return minus the SPY twin’s money-weighted return, expressed in percentage points. It is a direct arithmetic difference between two XIRR figures.

We deliberately avoid the word alpha for this number. Alpha, in its standard definition, is a beta-adjusted figure produced by regressing returns against the market to isolate the portion not explained by market exposure — a different calculation with different assumptions. Calling a simple difference "alpha" would overstate its precision, so Tradune does not use that term for this figure.

Costs, periods, and annualization

Returns are shown net of the commissions you recorded on your trades. They exclude taxes, advisory fees, and any cost you have not entered into Tradune — if it is not in your data, it is not in the number.

For periods under one year, Tradune shows the plain cumulative return over that period — a 5% gain over three weeks is shown as 5%, never annualized into a triple-digit rate that no one earned. For periods of one year or longer, the return is annualized, consistent with how XIRR is defined. See also how we define CAGR and max drawdown elsewhere on the site — related but distinct measures.

When we show nothing — or flag it — instead of a wrong number

Tradune suppresses the return figure rather than display one it cannot stand behind. That happens when:

• Your recorded history is incomplete — for example, a holding with no recorded original purchase, so the true starting cash flow is unknown.
• The cash-flow pattern is unusual enough that the XIRR solver cannot resolve it — we say so rather than force a number.
• The prices underlying a figure are stale — the figure is flagged rather than shown as current.

We would rather show you nothing than a wrong number about your own money.

Your data, your privacy

Your return is computed entirely from your own transactions and daily portfolio snapshots — the trades and deposits you enter, plus any positions you import by CSV (each recorded as a buy funded by a matching deposit). Tradune never links to or logs into a brokerage account to obtain this data; you provide it directly. The SPY series used for the benchmark twin is standard, publicly available market price data, not anything drawn from your own account.

Illustrative example

The figures below are illustrative only — round, made-up numbers to show the mechanism, not a real account, a real result, or a performance claim.

Say an investor deposits $10,000 on January 1 and a further $5,000 on July 1, and the account is worth $16,500 on December 31 of the same year — a period under one year, so the return is shown cumulatively rather than annualized. Tradune’s benchmark twin takes the identical two cash flows, $10,000 on January 1 and $5,000 on July 1, and prices them into SPY’s dividend-adjusted close on those two dates, then values the hypothetical SPY position on December 31. If that hypothetical SPY position had grown to $16,000, the investor’s cumulative return and the SPY twin’s cumulative return are each computed by the same method on the same dates — and the gap between them, in percentage points, is the illustrative excess return.

Frequently asked questions