Whilst unorthodox to box EUR vs CAD, trade may be a reasonable candidate to take profits for those with EUR steepeners (Trades like reds/greens EONIA steepeners), and take advantage of a potential reversal in the consensus short in the paid front-end, CAD rates trade (BA 2nd-5th contract curve is now the steepest in G10 at 20bp). You can more clearly look at how stretched CAD front-end is pricing in the CB dated OIS, where the next fully priced 25 bp hike for the BoC is in July 2018 whereas for the Fed this is in Nov (a USD vs CAD box is for another argument altogether, but this is just for a metric of 'strecthed-ness'!). I think after the 2nd hike in Sept, people assume too quickly of a regime change. Yes, there are real merits of Canada's improving economy, but to assume they would be on their first few hikes faster than the Fed's first few hikes. ? I doubt that. Some leading indicators have peaked as well, and the BoC has changed their tone to more neutral. On the EUR side, we may also be in a phase where steepening may lose steam with the EUR curve steepest in 2 years in spot 2s5s, with the trade now consensus, and with the ECB being very careful with its tapering communication. With the box at 10y wides, it seems tactically good risk-reward wise, to fade the EUR steepening, and CAD steepening safely.
Roll of the box through time shows 1y fwd point of the box to be best to play the trade.
Trade and what it rolls to in grey
If you look at the outright EUR vs CAD spread, you can see that the box proves a more roll-optimal, and entry-optimal trade compared to the outrights, where if you Rec CAD 2y1y for e.g. vs Pay the same in EUR, you would be incurring -12 bp of roll per annum.
The box is closely related to the front-end spread over the past 2 years (R^2 of 0.78). Over a longer out history it has periods of being disconnected and linked but weaker. So one could argue that it is a reasonable substitute vs front end trade.