You don't become a Palm Beach Post reporter because you're good at the math of annual forecasting, so everything about how the numbers are reported is silly.
(OT: just like most of the reporting on Amtrak Virginia...including VADRPT's early powerpoints...always just compared year-to-date traffic versus annual goals and failed to notice the month-over-month growths. Halfway through a year of compound growth, you DO NOT expect your passenger counts to be half of the annual goal, but more like 1/3rd). As it turned out, Amtrak Virginia blew through its Annual numbers by a wide margin by the end of its first year, even though, for most of the year, it appeared "below goal")
Brightline is clearly below forecast. I can't spin the numbers (below) to show that it is any better than 77% of forecast. But calling Q1 + Q2
...ridership for the first half of the year was 180,870, just 16 percent of the 1.1 million passengers Brightline told investors in December that it expected for the full year of 2018.
is just about the stupidest way of reporting it that I can imagine.
So here are 5 more reasonable "how are they doing" estimates from passenger data so far:
FORECASTS for CALENDAR YEAR 2018
1) Calling H1 just 16% of the annual forecast is meaningless. If they'd straight-line extrapolated H1 into H2:
180k+180k =
360k = 33% of calendar 2018 forecast
2) Better, straight-line forecast Q2 as the number for Q3 and Q4
74k+106k+106k+106k =
392k = 36% of calendar 2018 forecast
3) Better, consider Q1/Q2 growth of 30% is likely to continue (as happens in "new mode" introductions, like Light Rail or Amtrak Virginia)
74k+106k+140k+180k =
500k = 45% of calendar 2018 forecast
4) Better, consider that 74/106 is actually 43% quarterly growth (in very early quarters of "new mode" intros this is possible)
74k+106k+150k+210k =
540k = 49% of calendar 2018 forecast
5) Better, consider that going from 11 tpdew to 16 tpdew may capture both time-of-day and S-curve (frequency-convenience) growth in Q3 & Q4
74k+106k+180k+250k =
610k = 55% of calendar 2018 forecast
NOTE that given the assumptions in (5), above, it is possible that BL will have 250k passengers in Q4 2018, which would be (roughly) consistent with expecting 1.1m passengers per year "soon"
Finally, consider that the "Calendar 2018" forecast could also have been considered "Quarters 2 through 5 of operations." True, the bondholders need stuff to happen on time, but as far as passengers are concerned the ramp happens only when the service is available and has good word of mouth. So if Q1 of service (which had the 74k of pax) had happened as Q4 2017,
everything would have been shifted a quarter, which matters a lot for initial passenger forecasts when Q/Q growth can be 30% ~ 45%
FORECAST for Q2 through Q5 of OPERATIONS (as the 1100k forecast assumed service start in Q4 2017)
6) Taking (3) and shifting it 1 quarter you'd expect (by the end of March 2019)
106k+140k+180k+240k =
600k vs 1100k 54% of forecast for Q2 through Q5 of operations
7) Taking (4) and shifting it 1 quarter, you'd expect
106k+150k+210k+290k =
680k vs 1100k = 62% of forecast for Q2 through Q5 of operations
8) Taking (5) and shifting it 1 quarter, you'd expect
106k+180k+250k+320k =
850k vs 1100k = 77% of forecast for Q2 through Q5 of operations
The real growth will be a very different story than "16% of a forecasted 1.1m"
Is there someplace that BL actually made quarterly forecasts? There's NO WAY that 1.1m for 2018 assumed a flat 253k per quarter. Any forecast that got to 1.1m for 2018 would have assumed a quarterly ramp of about 30%. The Forecast for 1.1m in 2018 probably looked something like:
180k + 234k + 304k + 395k = 1.1m (ish)
And then you'd forecast at least 400k per quarter in 2019 and some more growth for a number in the 1.6m ~ 2.0m range.
There's no doubt that 3 bad things are "at play"
1) A late service start...you can't start compound growth until you've got that first quarter
2) A start at just 11/16ths of full schedule
3) Forecasting error. they do seem to have been 30% to 40% too optimistic (based on performance so far)