What’s changed

Three new calculators

All three run entirely in your browser. Nothing you type is sent anywhere, nothing is stored, and none of them asks for an account.

• TFSA and RRSP contribution room, which deposits count against which tax year. The RRSP part is the one people get wrong: a contribution in the first 60 days of a year can be applied to the previous year's deduction limit, which is when most Canadians contribute. Put your deposits in and it shows where each one lands, including the ones that belong to a year you were not measuring.
• US withholding tax, what US dividends lose before they reach you, in each account, and whether anything recovers it. The input most calculators leave out is how you hold the shares, and it changes the answer more than the account does: the treaty exemption applies to a dividend paid to the plan, so it survives a US-listed fund bought inside an RRSP and does not survive a Canadian-listed fund holding the US shares on your behalf.
• US dollar capital gain, each side of a trade converted at the rate on its own date, with the security half separated from the currency half. A US-listed holding whose price never moved can still produce a real, reportable Canadian-dollar gain, and this shows you that on your own figures.

The first two are checked against the product's own routines on every change, so a number here and the number inside Luum cannot drift apart. The withholding calculator has no counterpart inside the product to be checked against; its page says so rather than leaving you to assume otherwise, and its rates carry a review date that fails the build when it lapses.

Two screens from the product

The tour gained two stops. One is the transaction ledger, with Canadian-dollar and US-dollar trades sitting in the same table in their own currency, which is what a cost base is actually built from. The other is X-ray, which resolves every fund you hold to the companies inside it and counts how many separate ways you own each name. Two index funds can hold the same companies without either statement saying so.

X-ray states what it cannot see before it shows you anything. Some funds publish only their largest holdings and some publish none at all, so the part that cannot be seen is counted on its own rather than spread across the names that can, and every figure is a floor rather than a precise number.

The comparison pages describe the product we have now

The seven Luum-versus pages had gone stale, and measurably so: eight things Luum does appeared on none of them, and all eight were already documented elsewhere on this site. Contribution room tracking, the tax export, watchlists and alerts, the IPO calendar and market movers, the portfolio diagnostics, peer benchmarking, and the projection work were all missing. They are there now.

What has not changed is that each page still says where the other product is stronger than ours. Every one of those concessions survives. We removed four sentences that went further than conceding and told you to go and buy the other product, which is not something a comparison page owes anyone, but the honest account of where we are behind is the reason the pages are worth reading.

An automated check now ties each capability those pages claim to the documentation page that describes it, so a feature that goes away cannot keep being advertised on seven pages.

Smaller things

Two pages that could only be reached from deep inside the Learn and Compare sections now have a route from those section indexes. The Tools index described six calculators when there were ten. A blank figure in the return calculator used to say no rate solves your flows, which was not true for a bad year, and it now says which of the three reasons applies.

A correction first

Two pages, the FAQ and the brokerage sync guide, claimed Snaptrade reaches RBC Direct Investing and Scotia iTRADE, and it does not reach either one, nor National Bank Direct Brokerage, Qtrade or Desjardins. That is two of the Big Five, wrongly listed on the pages people read before signing up. Both pages are corrected, and they now name the seven Canadian institutions that are reachable and state the absent ones plainly.

If your brokerage is one of the five, the account is still trackable: every plan accepts a CSV or spreadsheet upload. We would rather tell you that here than have you find out during signup. An automated check now derives the supported list from the same source the home page uses, so the two cannot drift apart again.

Six guides on how people actually invest

The Resources section covered what you owe and which account holds it, and nothing about how a portfolio is actually run. That gap is now filled, in a new Investing approaches group:

• The couch potato portfolio, the structural argument without the evangelism, the three mechanical rebalancing rules and how each one fails, and the two jobs the method quietly hands back to you.
• One-ticker ETF portfolios, what the management fee actually buys, and the two mistakes that cancel the reason for holding one.
• Asset location across your accounts, which account should hold which holding, and the consequence usually left out: doing it properly makes every individual account unbalanced by design.
• Dividend investing in Canada, the ex-dividend mechanic, and why a dividend is a transfer of value rather than free money.
• Norbert's Gambit, step by step, the currency conversion technique, the four ways it goes wrong, and the cost-base records it leaves behind.
• Covered call ETFs, explained, what the fund is selling, why the headline yield is not a return, and what the distribution mix does to your cost base.

None of these recommends anything, and none of them quotes a fund fee, a distribution yield or a tax rate. Every real number is left to the source that owns it, so none of these pages can go stale or contradict a figure published elsewhere on this site.

Two guides on measuring performance

• How to benchmark your portfolio, a benchmark is a counterfactual, not a scoreboard. The four ways a comparison misleads, and three yardsticks that answer better questions than a market index.
• CAGR vs average annual return, averaging yearly returns overstates what you earned, always in the same direction, and the start-and-end formula applied to an account you contribute to is not a return at all.

Two more free calculators

Both run entirely in your browser, like the four already there.

• CAGR calculator, the compound annual growth rate between two balances, and then, once you declare your deposits, the rate your money actually earned. Most CAGR calculators report only the first and mention the problem in a footnote; this one shows both and labels the gap between them as your contributions.
• Dividend calculator, income a year, current yield and yield on cost side by side, because those are three different claims that get treated as one. If you reinvest, it also shows the adjusted cost base that reinvestment is quietly building, the figure people meet years later when they sell.

Smaller fixes

Two accessibility defects were corrected: seven comparison tables opened with an empty corner header, leaving the first column unlabelled for a screen reader, and a decorative brokerage logo was announcing its name twice. The methodology page for portfolio returns also now describes the benchmark picker accurately, it previously named a default that was never used and a setting that does not exist.

Six pages went live in the Resources section.

A glossary of the terms Canadian statements and slips actually use. Book value against adjusted cost base, identical property, superficial loss, disposition, pension adjustment, look-through, each with the thing it is most often confused with. It deliberately carries no figures at all: every limit and rate links to the page that owns it, so there is one place to correct when a number changes and no second copy to disagree.

Adjusted cost base across two brokerages. Canada pools identical property across every account one taxpayer holds, which means the same ETF at two brokers is a single cost base and neither broker can compute it. Worked in both directions, with the reconciliation check.

Companion guides for tracking accounts at specific brokerages, plus a roundup of the portfolio-tracker category organised by the job each tool does rather than by rank.

Two smaller corrections shipped the same day. The RRSP contribution deadline is now stated on the page people already reach when searching for it. And FAQ sections on Resources and Comparison pages are now published as structured data, so a search engine can read the question and the answer as a pair.

One page was also made less confident: the US accounts guide previously implied support Luum does not offer, and now says what it can and cannot do.

Four free calculators are live, and each one runs entirely in your browser: nothing you type is sent anywhere, nothing is stored, and none of them asks for an account.

• Adjusted cost base calculator, one security across buys, sells and commissions, showing what the cost base became after each transaction.
• T5008 and cost base checker, box 20 from your slips beside your own records, with the Schedule 3 totals each implies.
• Time-weighted vs money-weighted return, the return of the holdings, and the return your money actually earned.
• Investment fee drag calculator, what a management expense ratio adds up to over a long holding period.

Three of them are pinned to the product's own routines by automated checks that run on every change, so a calculator here and the figure inside Luum cannot drift apart. A public tool that disagrees with the product would be worse than no tool: a reader with two numbers has no way to tell which one to act on.

Tools also moved into the main navigation, and the four now have an index at /tools/ rather than living as pages you could only reach if you already knew the URL.

A contact form. Previously the only route was an email address, and the form that replaced it had a real defect worth naming: messages were being sent to a mailbox that had never been created, and the delivery service returned success anyway. Both the mailbox and the error handling are fixed, and the form now reports a failure instead of silently accepting a message nobody would read.