Completely accurate, which is why I state "I am basing this entire analysis on a number of large assumptions. Firstly, that my current outstanding loans will be repaid with the same reliability as my completed loans", and a number of other assumptions in addition. Nor do I know the extent to which previous loans, which may now be 100% repaid, may have dipped into default and subsequently recovered, which also incur a marginal cost in terms of opportunity cost of capital etc. My analysis is very much "cash based", which is limited, but this is the data I have available. A quick look through some of the 100% repaid loans suggests that at certain points in the cycle these were also overdue. Perhaps I could obtain the individual repayments versus due dates for each loan and work this out with greater accuracy, but I haven't bothered so far. To what extent are my current loans likely to default at the 18.65% rate you suggest? I have no idea. To what extent did my previous loans default to this extent? I have no idea. This is why I simply take a bird's eye perspective and look at net cash flows. However, as you accurately imply, for me to genuinely do this with precision I should wait until all current loans have been repaid, which either means I have idle funds on the platform, or I have to complicate the analysis by considering my average outstanding balance, which I have assumed for simplicity is $1000, which it has been to date roughly.

However, are you considering the gross interest income I will earn on the loans that don't default over the forthcoming period? This will partially offset some of the defaults, I can't say with precision to what extent. I count only principal outstanding, not interest income due. So, to some extent this will lessen the impact of possible defaults.

Finally, I have done a couple of new bids since my cut-off date of April 6th when I downloaded the data, so the numbers might not add up 100%. Oh, and Jessica is my wife, she started on Zidisha first, before I took over!

At the end of the day I am hesistant to say resolutely that Zidisha is a break-even venture. It has been so far, but subject to certain assumptions which I hope I have stated clearly. 18 months is a decent trial-period, but it's not a perfect analysis. And if I do subsequently lose a few percent on $1000 that is tolerable. Would I put $10.000 on the platform? No. And comments warning that Zidisha needs to tighten delinquency are completely accurate, and to an extent I am gambling on their ability to do just that - progress seems good so far (in 2014). What intrigues me is the innovation in the business model. It is disruptive. It is a first-mover in this space. Does that mean the model is perfectly refined and cannot be improved? I doubt it, and I look forward to seeing how they develop. But I think it is worth giving them a chance, which is what I have done (to a modest extent), and waiting to see what happens. I will update my blog periodically when more data comes in.