Thanks to dioxide45 for providing me an Excel export of the data in the ROFR.net database so I could do a little number crunching on the pass/fail metrics. Fortunately, that export contained the DC point value for most of the weeks in the data, but I used Steven Ting's online Google doc to fill in the few gaps. For now, I limited my analysis to the last 12 months, but might go back later to do another year or two. Not sure it would be worthwhile to go back farther than that, though, especially since what the data shows so far isn't particularly revealing.
The image below summarizes 203 ROFR.net entries over the last 12 months. I calculated the price per DC Point for each and looked at whether they passed or failed. The logic of this is as I mentioned in a post above -- if MVC is using ROFR to add inventory to the Trust, they would look at what exercising ROFR would cost them per point added to the Trust. That might not be the ONLY consideration in the decision, but it should be a key factor. I looked at pass/fail rates in roughly $0.50/point increments between $2/point and $4/point and the rates above and below those levels.
Here are the key data points as I see them:
- Overall, 52% of ROFR.net submissions passed ROFR
- Almost two-thirds of the submissions were at prices that resulted in a price/point of under $2/point. Only 46% of those passed.
- There were only 21 submissions reported at $4/point or more and 62% of those passed
- Between $2/point and $4/point, the pass rate was also 62%.
- The $0.50 increment breakdown results in small sample sizes in each category, so I'm not sure the variations seen in those increments really mean anything other than random variation.
View attachment 8053
Before I started this, based on the preliminary work I did a week or two ago looking at Hawaii ROFR, I would have theorized that somewhere around $3/point was the sweet spot for having a decent shot at getting past ROFR. But when we look at all 203 entries in ROFR.net over the last year, it appears you have just as much of a shot with an offer of $2/point or more.
One significant item from the detail data can be seen in this next screen shot. Of the
ROFR Failures at prices over $3/point,
all but one were points resales. So, all but one of the
fifteen weeks that were submitted at a price that resulted in a per point cost of $3/point or more, passed ROFR. That's a pass rate of 93% for weeks.
I didn't do a detailed isolation of weeks versus points below $3/point, but just eye-balling the data, the VAST majority of the ROFR submissions below $3/point are weeks, so the pass rate below $3/point for weeks would likely be very close to the 49% overall average noted in the first image above. So the passage rate for weeks of 93% above $3/point versus about 50% or so below that, is a significant variation.
There were 25 DC points resales submitted above $3/point, of those, 11 passed and 14 failed, a pass rate of only 44%. Above $4/point, the were 15 entries, and 7 passed, for a pass rate of only 47%. There were only two at $4.50/point or above - one failed and one passed. Points seem to fail at a much higher price than weeks. Not sure why if they all go back into the Trust.
View attachment 8055
So, based on all of this, I would conclude the following:
- ROFR does still appear somewhat random, even when looking at the data very organically.
- If you want the best chances to pass for a resale week, always make sure your offer is more than $2/point based on the DC point value of the week.
- If you are after a hard-to-find week that you don't want to risk failing, offering over $3/point may be prudent.
- For DC Points resales, you probably need to be at $4/point or more to have a good chance of passage, and even then, the odds appear to be 50-50 at best.
As always, we need to remember though that these results are only as good as the underlying data, and ROFR.net is still a user-sourced database, so the data entered may not be consistent.
I'm glad I didn't have to spend more than a couple hours doing this, because it didn't provide as conclusive an answer as I thought it might.