Kiva voodoo

I lend money through Kiva. I like the idea of P2P lending, not so much because traditional finance might not be available to the borrower, but because traditional finance is nothing short of legalized crime.

This said, I see a financial hole in my Kiva account:

Do you see it too?

If you don’t know how Kiva works, you make a deposit to create a balance, and then make loans from your balance. As people repay their loans, it falls back into your balance and you can re-lend the money. Kiva recommend a $3.75/$25 donation to them, but I wind that to zero, and usually just donate my left-over balance at the end of the year.

So, I’ve injected $275 of deposits. In theory, my total outstanding loans, plus balance, plus donations should equal $275. Any difference should be due to defaulted (unpaid) loans.

My outstanding loans total $209.91. My balance is $0.15. My donations are not explicitly stated, but $1,450 / $25 gives 58, which at $0.90/$25 means I’ve donated a total of $52.20.

$275 minus $209.91 minus $0.15 minus $52.20 should therefore equal zero.

But it doesn’t. $12.74 is missing.

Ah yes! Defaulted loans! These are third-world borrowers to whom normal financial institutions don’t lend. A 4.6% default rate sounds low even! Except I have no defaulted loans.

The answer turns out to be due to currency fluctuations, whereby Kiva essentially deduct the interim difference between the loan going out of $US to some other currency, and when it is paid back (and converted back to $US). And this annoys me greatly (though not as much as having to pay a traditional bank a fee to make a $US payment into Kiva in the first place!):

This alternative account summary shows the currency loss, but also compares me to other lenders.

I’ve made 45 loans with a total value 4 times that of the average lender. I have a zero delinquency rate compared to almost 9% for the average lender, and a zero default rate compared to 1.37% on average. So, if I’m out performing the lending market so well, why the hell am I losing 2.5x more to currency loss?!

(Yes, I realise this is about only $13, but that’s half of a loan to somebody!)

Maybe the “average” loan goes to a US (or $US borrower) more often (mine never do since they have access to traditional finance); but if so, then those delinquency and default rates suggest those with access to traditional finance are worse borrowers who I want to avoid.

It seems no matter what, you’ll lose money along the way: 0.9% in my case to currency fluctuations, and 1.4% on average to bad borrowers.

That must be why a bank charges 19.9% APR. Oh wait! That doesn’t add up either…


I moved to a new house, and want my (admittedly not too) humble abode to be nice and secure.

I – perhaps naively with hindsight – thought I might simply call a local alarm company and they’d come and install an alarm. It turns out though that, much like the electricians in the area, it’s really quite hard to get somebody to come and do some simple work in exchange for money.

Apparently I’m in the wrong business then. Electricians two years out of their snot-nosed apprenticeships and their ilk don’t just turn down work; they don’t even respond to enquiries! At least I had one alarm company respond. They quoted silly money to do the job though, so no cigar there…

So, I figure if I can design and build billion-dollar technical estates I can deal with an alarm system. And, pleasantly, it turns out my 2 year-old could probably do so too. Remember that the next time you need an alarm system.

Pro tips:

  • Avoid Yale systems. They’re apparently simple, and trivial to jam and defeat.
  • Avoid “cloud” alarms. I work in this industry, and can tell you first hand that it’s a BAD idea to place this stuff on the Internet. Our neighbour’s complaint that he couldn’t set his alarm when his broadband went down only proves my point!
  • Be wary of monitored alarms that have high monthly costs. Rules are fairly strict around what alarm conditions can generate a police response… essentially not many, so what’s the point?!
  • Avoid shock sensors. They sound great in theory, but a bit of reading suggests they provide a good way for an attacker to probe the responsiveness of any monitoring company before breaking in for real.
  • Weigh up the technical security of a wired system against the pain-in-the-ass of wiring it.

If you go wireless:

  • Select a system with bi-directional encrypted communications and anti-jamming tamper detection.
  • Draw a floor plan with all your sensors mapped out, and number them on plan and physically on each device.
  • Learn (sync/pair/etc.) and configure each sensor at the table all in one go.
  • Don’t forget that tamper circuits trigger outside of engineer mode: 120dB at 30cm is nothing short of painful!

I ended up opting for Pyronix Enforcer. I toyed with a Texecom Premier Elite too, but was advised it has unresolved firmware issues presently.

The end result was about 3 hours to read the manual and configure everything, and another 2-3 hours screwing sensors to doors and walls.

That’s about £160/hour then. Like I say, I’m in the wrong business.