Here at Pear Tree Property, we offer a $25 a month discount to new residents if they sign a two year lease. If the resident leaves before the lease is up, then that $25 a month is added to their account balance when we get a judgement. It sounds like a simple win-win scenario – the resident pays less and we have a lower turnover.
How does it pan out in practice? Does providing an incentive to sign a two year lease make sense? This post is letting out a bit of my inner stats geek, but I hope you find it interesting.
Buildium Reporting Fun
First things first, get the data. We use a system called Buildium for all of our property management and accounting. It’s got a lot of nice features: it’s easy to update, has a nice user interface, and is entirely web based so our investors can check up on us whenever they want.
When it comes to reporting, Buildium is a bit of a let down. To take a first pass at today’s problem all I really want is to see all of our leases, when they were supposed to end, and when the tenant actually moved out. Simple right? Negative Ghost Rider.
Without going into too much detail, I had to pull raw event history and then write a python script to parse out the relevant lease transactions. This “give me somewhat useful lease information” script is just the most recent in my growing arsenal of scripts to work around short comings in Buildium’s reporting. (feel free to drop me a line if you want the script).
That said, my background probably predisposes me to want to access data in a format that the less geeky user couldn’t care less about, so I shouldn’t knock Buildium too hard. It was actually sort nice to have a call back to my finance days.
OK, moving on… I’ve got my data, now what?
The first thing I did was chuck the leases that are still active. Using this approach yielded the following high level numbers:
Those numbers are, quite frankly, scary. Only 14% of two year leases are fulfilled? Huh?
Turns out these numbers come from my crappy methodology. Here’s why: let’s say a we only have two residents, both signed on August 1, 2012. Today is July 6th, 2013. Let’s also say we had to evict one of the residents on October 1, 2012. Since I’m throwing out all current leases, our above table would show none of the leases “fulfilled” and 60 as the “Average Days As Tenant.” This effect is more pronounced for two year leases.
It’s even worse considering we’re always buying more homes and so have proportionally more current leases than former leases.
This is what stats folks like to call a selection bias.
So how do we get around this selection bias? Honestly, I’ve no fricking clue. If you’ve got a recommendation, drop me a comment. What I can do is go to the other extreme and include all current leases. Unfortunately, doing this makes our “Average number of days” useless (if we just signed a lease it would bring down the average).
Here are the high level numbers after some crunching:
Conclusion on Two Year Lease Incentives
What can we conclude? I can safely say we don’t have much data. Interestingly enough, it’s pretty hard to come draw a conclusions without enough data.
That said, I’ll make some general observations. In the second table, the “percent fulfilled or current” is roughly the same for one and two year leases. Even with a huge bias in the first table, the “average days as tenant” is still higher for a two year lease. I don’t think we’re shooting ourselves in the foot by offering a slight discount to sign a two year lease.
The close reader will note that what we’re calling the “percent filled or current” is actually the inverse of our eviction rate, which would be 31%! That’s worrying.
It’s not all doom and gloom. Our occupancy is always above 90%, which means our property manager, Suzy, is doing a good job getting new residents. Also, our eviction rate is steadily improving. If we restrict our sample to leases starting in 2011 or later, it’s 27%. If we look at only 2012 and beyond, it’s 22%.
Feed Kenny Feedback!
This is my first blog post digging into the statistics and exposing some of our internal data. Please let me know what you think. Do you want more stuff like this? Did your eyes gloss over? Are you not even reading this sentence because you saw the word statistics and closed your browser?
Latest posts by Kenny Estes (see all)
- The Surprising Truth About Interest Rates and Real Estate Values - 26 December, 2013
- The future of real estate lending - 19 December, 2013
- Is This a Bitcoin Bubble? - 13 December, 2013
- Cynical investing - 5 November, 2013
- Swing dance around the world - 20 October, 2013