June 6 also happens to be D Day.....and since a Saturday launch is likely designed to be a first day on a less busy day of week there are likely gonna be some empty seats.....would be a nice PR move to allow Military Vets to ride for free on launch day (IMO)
Don't forget the classic "novelty surge" (aka, the surge of users at the introduction of a fancy new service).
Better to have 80%+ full trains rather than standing people, so they really had to choose a quiet day. Then later make a big splash of popularity claims when 80% full trains become too packed during peak. The curve of daily Pearson traffic and surges from big planes, will cause near-empty trains and packed trains, so both naysayers and Metrolinx will be fighting for years -- mark my words. (But at the end of the day, it should eventually become profitable at average seat revenue). Tourists who pay $27.50 to be standees will be vocal with negative social networking. Metrolinx seems aware of this. Better to target nearly full, rather than risk overfilling the trains too early.
I am pretty sure, however, sales of $6 Presto cards at Pearson, will probably be accounted as part of UPX's revenue, inflating a $19 seat sale to a $25 seat sale (or $22 average revenue per roundtrip tourist). This is probably a very reasonable accounting rule, as long as cost of presto ongoing operating/manufacturing overhead (e.g. $1) is subtracted ($5 presto "profit") since, obviously, captive audience will prefer to pay $25 to get a Presto, over a $27.50 non-Presto, and it also gives tourists an extra incentive to use the other transportation services (TTC, GO) at Union or to pre-fill it for a return trip so they decide to take UPX back instead of a taxi. Repeat tourist visitors will sometimes remember to bring their Presto, and have a considerable incentive to use UPX again (even if they catch a taxi at Union to go to a second destination near downtown, like Liberty or Exhibition Place -- which will eventually have RER options, too and reduce the need for a taxi too).
Under this rule, I'd guesstimate $23 per passenger average revenue (combination of $19 residents, $19+Presto purchases for visitors, and $27.50 non-Presto). Operating cost is mentioned at $68-$70M per annum, and I expect UPX to surpass projections to about 3 million people within 5 years. This covers operating revenue. Assumng average of 150 seats per train (60 per car, 120 for two-car, 180 for three-car), each full train without standees pulls an average of $3450 revenue. Including standees which would sometimes happen during A380/747 landing events, this might be more. But let's use $3450 per train as a "maximum possible average" with the current fleet. UPX runs 15 mins from 5am to 1am, with 140 trips, yielding a theoretical maximum of non-standee $483,000 revenue per day. If you could do this 365 days a year, you get $176 million annually for 365 days! So $68M versus $176M.
So you only need to utilize 68/176ths capacity to break even.... 38%.
Therefore, UPX is profitable if trains are more than 38% full on average
I point out Metrolinx forecast about 2 to 2.5 million people within 5 years, while theoretical UPX capacity would be 7.7 million people. That's only 30%, but Metrolinx may be using slightly different numbers (e.g. using a higher predicted average revenue per passenger, I predict lower because of the Presto bundled purchases + prefills for return trip). I predict 3 million people in less than 5 years, partly thanks to Presto prefills, and I would not be surprised to see it exceed 3 million especially given Pearson's expected slight acceleration in growth brought upon by UPX, as I know I would now sometimes use Pearson instead of Billy Bishop for short trips like Ottawa, now that UPX exists. I'm sure many others agree too. So I'd forecast slightly lower average revenue per person, but higher passenger traffic within 5 years, based on common-sense stuff. It's always good to lowball the projections for a controversial route, and that has been done frequently (by precedent).
So they can afford a few empty trains in between full trains. That doesn't even include standees during 747/A380 surges. This doesn't even include an upgrade to an all-3-car fleet (which would be worthwhile to absorb surge 747/A380 moments, even if the extra cars run empty-ish offpeak). An all-3-car fleet would be $3680/train (or $4500 with standees), or about $1M per day at 7.5 minute headways, if they were inclined to do so (operating costs would go up, but would not double). This doesn't even include the potential ability to ramp service up to 12 minute intervals during events (e.g. Pan Am), if they use all trains they have.
Although I'd prefer slotting fast-accelerating all-stop GO RER trains between the UPX trains instead to make more efficient use of the corridor capacity. From a people-moving volume perspective, UPX is a waste of corridor capacity (compared to, say, a frequent bi-level GO RER train). However, on an operating profit perspective, I think UPX actually will do better than Toronto's population expects.
Whether this is good or bad is always a debate, but many seem to falsely think UPX is going to be a loss-making endeavour -- considering the captive airport audience with worse "cheap" public transit alternatives to downtown, than similiar expensive-train cities like London/Tokyo/HongKong. Many of those cities have cheaper appealing train alternatives (e.g. a single subway line) while at Pearson there's only a bus under an ugly arrivals viaduct that takes more than an hour to Union with a Kipling transfer. In addition to the typical precedent of lowballing of projections of a controversial transit projects worldwide, this is another reason precisely why I think UPX will be incredibly popular with Pearson travellers, and passengers will exceed Metrolinx projections significantly.